J. Adhesion Sci. Technol., Vol. 20, No. 13, pp. 1463– 1474 (2006)  VSP 2006.

Also available online - www.brill.nl/jast

Relation between the cleanability of bare or polysiloxane-coated stainless steels and their water contact angle hysteresis LAURENCE BOULANGÉ-PETERMANN 1,∗ , CHRISTELLE GABET 1 and BERNARD BAROUX 2 1 Ugine & Alz, Research Center, BP 15, 62330 Isbergues, France 2 Institut National Polytechnique de Grenoble, Laboratoire de Thermodynamique

et Physico-Chimie

Métallurgique, 38402 Saint Martin d’Hères Cedex, France Received in final form 19 July 2006 Abstract—Cleaning of bare or coated stainless steel surfaces is investigated using some specific techniques for both particulate soil and oil removal. Particulate soil is removed from the surface by a water drop sliding, whereas oil is eliminated by shear flow of a commercial detergent. The cleanability performance is found to depend both on surface energy and topography. In general, the water contact angle hysteresis, which itself is related to the advancing contact angle and the surface roughness, is found to be an appropriate criterion for characterizing the cleaning performance. This finding is discussed in terms of retention and removal forces during the cleaning process and could provide in the future a criterion for material selection for industrial use of stainless steel surfaces. Keywords: Contact angle hysteresis; drop sliding; retention force; sliding force; soil removal.

1. INTRODUCTION

Soiling of stainless steel surfaces is commonplace in the catering industry [1], medical appliances [2, 3], food industry [4] and wall panels of buildings [5, 6]. This surface soiling can be a favorable environment for further bacterial growth. Thus, cleaning is of major importance in the performance of materials. Under a shear flow on horizontal surfaces or under gravity effect on inclined surfaces, liquid drops in contact with solid surfaces can slide on or lift from solid surfaces. The retention force of a drop on a solid surface depends on both advancing ∗ To whom correspondence should be addressed. Current address: BDMedical Pharmaceutical Systems, 11, rue Aristide Bergès, 38800 Le Pont de Claix, France. Tel.: (33-4) 7668-9421. Fax: (33-4) 7668-3773. E-mail: [email protected]

1464

L. Boulangé-Petermann et al.

and receding contact angles, as well as on the drop shape [7, 8], whereas the sliding force is a function of numerous parameters such as the velocity of the shearing fluid [9], the tilt angle necessary to achieve drop sliding on an inclined surface [10], the surface roughness and the liquid viscosity [11]. To assess the cleanability of a material, different empirical tests can be used, involving natural exposure for a long term [5], carbon black soiling for simulating soiling of buildings in urban environments [12], oils [13, 14], and removal of pathogenic microorganisms in food industry [15, 16]. On a flat stainless steel, the ease of cleaning is generally discussed in terms of surface composition [17] or topography or polarity [18, 19]. Here, we have investigated the cleanability of bare and coated materials under controlled flow geometry. A laminar flow system was initially described by Pratt-Terpstra et al. [20] and adapted by Boulangé-Petermann et al. [21] to study the oil removal by a shear flow containing surfactants. For particulate soil removal, we have developed a new method involving measuring the number of particles contaminating the surface, as well as the minimal angle to achieve water drop sliding on soiled surfaces. The purpose of this work was to assess the role of the contact angle hysteresis in the cleaning kinetics of soiled steel surfaces. To this end, we selected different stainless steel surfaces and polysiloxane coatings. Advancing and receding contact angles, as well as topographic measurements were performed on bare and coated surfaces. The ease of cleaning was evaluated by assessing the oil droplet removal under a shear flow and particle soil removal by water droplet sliding.

2. MATERIALS AND METHODS

2.1. Selection of materials SS30400 (Unified Number System)-type 1-mm-thick stainless steel sheets with different surface finishes, referred to as Ss1 to Ss16 , were studied (Table 1). The final surface condition depends on the finishing process. Cold-rolled stainless steel sheets were generally heat-treated to attain suitable mechanical properties. Bright annealed condition (Ss1 ) means that the final annealing is performed in a hydrogen containing atmosphere and does not need any subsequent chemical pickling. However, the water content in the atmosphere is sufficient to form a passive film [22]. In contrast, when such annealing is performed in an oxidizing atmosphere, a final pickling is carried out in order to remove the oxide scale (Ss2 ), formed during annealing. Then, the film formation is completed by rinsing with water and further exposure to an ambient atmosphere (relative humidity 30%). Mechanically ground samples (Ss3 to Ss7 ) were obtained from an Ss2 sheet processed in laboratory conditions (Ss4 , Ss5 ) or in industrial conditions (Ss3 , Ss6 , Ss7 ) using abrasive strips with different sizes of carbide particles (120 and 320 µm). Textured surfaces (Ss8 to Ss15 ) were obtained using textured work rolls. (Ss8 ) and (Ss9 ) surfaces were obtained from an Ss2 sheet processed with a polished roll.

Relation between cleanability of stainless steel surfaces and water contact angle

1465

Table 1. Advancing (θa ) and receding (θ r ) water contact angles (in degrees), angle of tilt necessary to achieve water sliding (α), contact angle hysteresis (H ) and surface topography parameters (in µm) Sample designation

Ss1 Ss2 Ss3 Ss4 Ss5 Ss6 Ss7 Ss8 Ss9 Ss10 Ss11 Ss12 Ss13 Ss14 Ss15 Ss16 C1 C2 C3 C6 C7 C8 C9 C16

Preparation procedure

Bright annealed Pickled Mechanically ground Mechanically ground Mechanically ground Mechanically ground Mechanically ground Textured Textured Textured Textured Textured Textured Textured Textured Chemically attacked Bright annealed + coated Pickled + coated Mechanically ground + coated Mechanically ground + coated Mechanically ground + coated Textured + coated Textured + coated Chemically attacked + coated

Topography parameters

After cleaning with solvents

After cleaning with alkaline detergent

Sa

St

α

θa

θr

H

θa

θr

H

0.05 0.3 0.16 1.5 0.15 0.4 0.07 1.3 0.7 0.7 0.16 3.4 1.7 0.6 0.3 3.9 0.06

2.3 2.4 1.2 10.2 1.9 2.0 0.5 9.1 4.9 4.6 1.4 15.3 8.2 4.9 1.8 17.2 0.4

45 55 38 — — 49 37 54 55 — — — — — 39 60 15

79 95 95 — — 86 97 99 100 — — — — — 100 118 104

35 42 44 — — 54 59 49 62 — — — — — 58 73 87

0.63 0.83 0.78 — — 0.52 0.64 0.81 0.62 — — — — — 0.70 0.76 0.30

57 89 93 60 92 — — 52 88 65 50 78 62 93 — — 103

16 25 34 17 35 — — 14 19 16 9 20 9 26 — — 71

0.41 0.89 0.88 0.45 0.86 — — 0.35 0.90 0.50 0.34 0.73 0.52 0.95 — — 0.55

0.20 0.13

3.9 2.6

— —

— —

— —

— —

107 103

73 70

0.58 0.57

0.1

0.8

25

106

83

0.40

—

—

—

0.05

0.3

17

100

79

0.36

—

—

—

1.0 0.9 3.1

5.8 4.7 12.4

30 — 35

113 — 129

83 — 81

0.52 — 0.79

— 106 144

— 78 83

— 0.48 0.93

Sa and St correspond, respectively, to the arithmetic average roughness and the maximum peak-tovalley height.

Other textured surfaces were obtained from an Ss1 sheet processed respectively with a polished roll where the average size of carbide particles was between 180 and 220 µm for (Ss10 ) and (Ss11 ), a final roll (shot-blasted) for (Ss12 ), (Ss13 ) and (Ss14 ), with a textile around the roll for (Ss15 ). The chemically attacked sample (Ss16 ) was obtained from an Ss2 sheet by immersing for 10 min in a 60% HNO3 solution at a current density of 120 mA/cm2 . Polysiloxane coatings were obtained using Plasma-Assisted Chemical Vapor Deposition (PACVD) with HMDSO (hexamethyldisiloxane) monomer in the reactor chamber on Ss1 , Ss2 , Ss3 , Ss6 , Ss7 , Ss8 , Ss9 and Ss16 finishes and are referred to as

1466

L. Boulangé-Petermann et al.

C1 , C2 , C3 , C6 , C7 , C8 , C9 and C16 respectively. This particular coating material was selected in order to obtain hydrophobic surfaces [19]. Prior to any testing, the test specimens were first soaked in an ethanol/acetone (50:50 v/v) mixture for 5 min. Surfaces tested in a shear flow were also cleaned in an alkaline detergent (RBS 35, Traitements Chimiques de Surfaces, Frelinghem, France) at 50◦ C for 5 min at a concentration of 2% (v/v) (pH 10.8). This commercial formulation contained both nonionic and anionic surfactants. Lastly, the surfaces were rinsed five times with distilled water at 50◦ C and then five times at room temperature. 2.2. Contact angle measurements Advancing (θa ) and receding (θr ) contact angles (in degrees) using a Krüss G-10 goniometer were measured by the sessile drop method where the volume of water droplet was increased or decreased until the contact line moved over the solid surface. The measurements were repeated 20 times on each sample. Rather than the usual contact angle hysteresis θa −θr , we prefer to consider the hysteresis parameter, H , as H = cos θr − cos θa ,

(1)

because as mentioned in previous papers [7, 8, 10, 11], the retention force of a drop on a solid surface depends on cosines of both advancing and receding contact angles. 2.3. Surface topography The selected parameters [23] were the arithmetic average roughness (Sa ) and the maximum peak-to-valley height (St ), both expressed in µm. These parameters were deduced with a profilometer (Microsurf 3D, Fogale, Montpellier, France) (scan area 100 × 100 µm2 ) using the Surfvision software. The measurements were repeated three times per sample. 2.4. Particulate soil removal from solid surfaces by water droplet sliding Surfaces were soiled with carbon black particles with a diameter less than 45 µm. A water droplet with a volume of 60 µl was deposited on the sample inclined at 60◦ with respect to the horizontal. The minimal angle (α) to achieve a water droplet sliding was measured by tilting the sample until the water drop started to move. α is expressed in degrees. The percentage of particulate soil removed by the water droplet D = Dt /D0 ×100 was measured by an image analysis system, where D0 corresponds to the initial number of particles on the surface and Dt to the number of particles removed by the water droplet from the surface and the sample surface was observed with an optical microscope (Olympus AX60) at a magnification of 100, coupled with an

Relation between cleanability of stainless steel surfaces and water contact angle

1467

image analysis (Analysis version 5.0). The number of soil particles was determined by image analysis. A particle removal rate D larger than 90% (and α lower than 35◦ ) was considered as revealing a good cleaning performance. The experiments were done in triplicate on each sample. 2.5. Oil droplet detachment from solid surfaces by shear flow Prior to cleaning, the surfaces were soiled with nutritious sunflower oil (Lesieur, France), mainly composed of fatty acids (13%), monounsaturated fatty acid (oleic acid) (22%) and polyunsaturated acids (65%), by spraying the oil. Using this technique, the surfaces were covered by a large number of small droplets with a diameter not exceeding 1 µm. Surface cleaning was then performed using a laminar flow cell, which was presented elsewhere [21]. This cell consisted of two flat plates between which a commercial detergent ‘Mr. Propre™’ (Procter & Gamble, Neuilly sur Seine, France), pH 8.2 at 20◦ C, at a final concentration of 1.2% (v/v) circulated [23]. This detergent contains anionic (less than 5%) and non-ionic (between 5 and 15 wt%) surfactants. The critical micelle concentration (cmc) of the commercial detergent determined by the du Noüy ring method was 0.1 wt% [23]. Based on the weight difference, the percentage of oil removed after 20 s of cleaning, R = 100(W2 −W3 )/(W2 −W1 ), was determined, with W1 , W2 and W3 , respectively, being the weights of the initial sample, after soiling with oil and after cleaning in the laminar flow cell (and air drying). R values more than 90% characterize a good cleaning performance, R between 50 and 90% an intermediate performance and R values less than 50% a poor performance in cleaning. The experiments were done in triplicate on each sample.

3. RESULTS

3.1. Water contact angles Advancing and receding water contact angles (θa and θ r ) are shown in Table 1. After solvent cleaning, the stainless steel surfaces are hydrophobic (θa varies from 79◦ on Ss1 to 118◦ on Ss16 ) and the hysteresis parameter (H ) ranges from 0.52 to 0.83. After cleaning with alkaline detergent, the stainless steel surfaces are slightly less hydrophobic (θa varies from 57 to 93◦ ). Whatever the cleaning treatment, polysiloxane coatings deposited on stainless steels are hydrophobic (θa from 100 to 144◦ ) but the hysteresis (H ) and the topographic parameter St are smaller than for the bare metal. 3.2. Cleaning performance and contact angle hysteresis On polysiloxane coatings, the minimal angle (α) to achieve water drop sliding ranges between 15◦ (sin α = 0.26) and 35◦ (sin α = 0.57) whereas higher values of

1468

L. Boulangé-Petermann et al.

Figure 1. Minimal angle (α) to achieve water drop sliding as a function of the contact angle hysteresis on bare stainless steels (Ss1 , Ss2 , Ss3 , Ss6 , Ss7 , Ss8 , Ss9 and Ss15 ) and polysiloxane coatings (C1 , C6 , C7 , C8 and C16 ). (F) and (E) Materials where particulate soil removal is more than 90%; (Q) materials where particulate soil removal is less than 90%. Filled symbols correspond to bare surfaces, whereas open symbols to polysiloxane coatings.

minimal angle were obtained on bare stainless steel surfaces (Fig. 1). In the latter case, α varies between 37 (sin α = 0.60) and 60◦ (sin α = 0.87). On bare stainless steel surfaces, (Table 1), α is higher than 35◦ and no clear relation between α and H was evidenced. The minimal angle α is proportional to the average roughness (Sa ) (Fig. 2). For the same range of Sa , the minimal angle is lower on polysiloxane than on bare surfaces. Irrespective of the soiling and the removal methods used, a decreasing relation between the cosine of the advancing contact angle and hysteresis parameter is evidenced (Figs 3 and 4). When particulate soil is removed from solid surfaces by water droplet sliding, it is possible to discriminate bare stainless steel materials with intermediate (F), or poor cleaning performance (Q) from polysiloxane coatings (E) with good cleaning performance (Fig. 3). Coated surfaces C6 and C7 are materials with high cleaning performance (D > 90% and α < 35◦ ), whereas uncoated stainless steel surfaces Ss6 and Ss7 have intermediate performance in cleaning (D > 90% and α > 35◦ ). Bare surfaces such as Ss1 and Ss8 are materials with low cleaning performance (D < 90% and α > 35◦ ), but after polysiloxane coating (C1 and C8 ), their cleaning performance is improved (D > 90% and α < 35◦ ). When an oil droplet is detached from solid surfaces by a shear flow, it is possible to discriminate stainless steel materials (Q and F) from polysiloxane coatings (E)

Relation between cleanability of stainless steel surfaces and water contact angle

1469

Figure 2. sin α as a function of the arithmetic average roughness (Sa , in µm). (F) and (E) Materials where particulate soil removal is more than 90%, whereas (Q) materials where particulate soil removal is less than 90%. Filled symbols correspond to bare surfaces whereas open symbols to polysiloxane coatings.

(Fig. 4). Only bare stainless steel surfaces show high cleaning performance (Ss1 , Ss3 , Ss4 , Ss5 , Ss8 , Ss10 and Ss11 ). After polysiloxane coatings on Ss1 and Ss3 , their cleaning performance is strongly lowered (C1 and C3 ). However, experimental points corresponding to good (R > 90%) and intermediate cleaning performances (50 < R < 90%) for bare stainless steel surfaces are on the same line. Materials with good cleaning performance (Ss3 and Ss5 ) or intermediate performance (Ss2 , Ss9 and Ss14 ) display similar contact angle hysteresis and advancing contact angle. Materials with good performance have an average roughness less than 0.2 µm (Ss3 , Sa = 0.16 µm; Ss5 , Sa = 0.15 µm), which is smaller than those of the intermediate group (Ss2 , Sa = 0.3 µm; on Ss9 , Sa = 0.7 µm; Ss14 , Sa = 0.6 µm). A similar trend was observed on more hydrophilic bare surfaces, where better cleaning performance was observed on smoother surfaces Ss4 and Ss10 with Sa = 1.5 and 0.7 µm, respectively, in comparison to intermediate performing material Ss13 (Sa = 1.7 µm).

4. DISCUSSION

In this section, first the effect of water contact angles on cleanability is discussed and then the forces governing soil retention on solid surfaces and soil removal from solid surfaces will be related to the hysteresis parameter.

1470

L. Boulangé-Petermann et al.

Figure 3. Particulate soil removal: relation between contact angle hysteresis and cos θa . (F), (E) and (Q) Materials with good cleaning performance where particulate soil removal (D) is more than 90% and α less than 35◦ , materials with intermediate cleaning performance where particulate soil removal (D) is more than 90% and α more than 35◦ , and materials with poor cleaning performance where particulate soil removal (D) is less than 90% and α more than 35◦ on bare stainless steels (Ss1 , Ss2 , Ss3 , Ss6 , Ss7 , Ss8 , Ss9 and Ss15 ), as well as on polysiloxane coatings (C1 , C6 , C7 , C8 and C16 ).

4.1. Water contact angles on coated and bare stainless steels Cleaning with an alkaline detergent is known to decrease surface hydrophobicity [19, 24] which, in the case of stainless steels, is assumed to be the effect of removal of a thin carbon layer contaminating the surface [25]. Cleaning with an aqueous solution containing surfactants removes better the carbon contamination than a solvent such as an ethanol/acetone mixture. Moreover, after a detergent cleaning on bare stainless steel surfaces, the energetic characteristics, as well as the surface charge expressed by the zeta potential, are modified [24, 26], possibly due to the adsorption of some molecules from the cleaning detergent. Furthermore, coating by vacuum deposition lowers the topographic parameters obtained from profilometer analysis. Bare stainless steel surfaces display some holes that are filled by the polysiloxane coating. After coating, the decrease in contact angle hysteresis could thus be due either to some modifications in surface topography and/or at the interface between water and solid.

Relation between cleanability of stainless steel surfaces and water contact angle

1471

Figure 4. Oil removal: relation between contact angle hysteresis and cos θa . (F), (Q) and (P) Materials with good cleaning performance where oil removal is more than 90%, intermediate cleaning performance where oil removal is between 50 and 90% and low cleaning performance where oil removal is less than 50% on bare stainless steels (Ss1 , Ss2 , Ss3 , Ss4 , Ss5 , Ss8 , Ss9 , Ss10 , Ss11 , Ss12 , Ss13 and Ss14 ), as well as on polysiloxane coatings (C1 , C2 , C3 , C9 and C16 ).

4.2. Water contact angle hysteresis and cleanability An adhering drop slides along an inclined solid surface when the sliding force (FS ) FS = mg sin α,

(2)

overcomes the retention force (FR ) FR =

4Rγ (cos θr − cos θa ), π

(3)

where m, α, R, θ r and θa denote, respectively, the mass of the drop, the angle of tilt necessary to produce sliding [7, 8], the average radius of the drop (≈3 mm) and the receding and advancing contact angles [9]. γ is the interfacial energy between the liquid and the soil. Equating the sliding force and the retention force for a drop sliding along a solid surface yields 4Rγ (4) (cos θr − cos θa ) . π Equations (3) and (4) are intended for liquid drops that retain a circular contact line up to the point of sliding [10]. However, irrespective of the solid under consideration, there is a relation between H and cos θa indicating that the sliding force (FS ) is dictated by the advancing mg sin α =

1472

L. Boulangé-Petermann et al.

contact angle [27]. This can be related to the physical characteristics of the surface, i.e., the solid surface energy (γs ) and topography. 4.2.1. Particulate soil removal. On polysiloxane coatings, the minimal angle (α) to achieve water drop sliding increases with water contact angle hysteresis, as illustrated in Fig. 1. This trend was observed only on five experimental points, but it could be considered as a linear relation with a slope of 1.17 and a regression coefficient R 2 = 0.8144. However, no clear relation was obtained for bare stainless steels. On polysiloxane coatings, the water drop is not deformed before sliding whereas on bare stainless steel surfaces, the water droplet first deforms before sliding (Fig. 5). It is assumed in the latter case that the water droplet first deforms before sliding. The contact line of elongated drops can be approximated as an ellipse, and theoretically, the retention force increases as drops elongate in the direction of the inclination [10]. This drop deformation can be explained by (i) some variations in surface topography expressed by the average roughness [11] or (ii) the surface energy of solid surface. When a particulate soil is removed from solid surfaces by a water droplet (Fig. 3), three behaviors can be distinguished, corresponding to polysiloxane coatings, stainless steel surfaces with (D > 90%, α < 35◦ ), (D > 90%, α > 35◦ ) and (D < 90%, α > 35◦ , respectively). The main difference between the polysiloxane coatings and bare stainless steel surfaces is believed to be an effect of surface chemistry leading to low γs values (around 20 mJ/m2 ) compared with bare stainless steel surfaces (between 30 and 50 mJ/m2 , values obtained on stainless steel surfaces contaminated by a carbon layer) [19]. The main difference between the two groups of bare surfaces discriminated by D values could be attributed to a topographic effect. In the case of particulate soil, the retention force of the particles on solid surfaces is dictated by water contact angle hysteresis. The sliding force directly involved in the particle removal depends on the solid surface energy and topography. In our case, a low energy and an average roughness less than 0.1 µm leads to the best performance in particle cleanability.

(a)

(b)

Figure 5. Water drop sliding on (a). C16 -coated stainless steel and (b) Ss16 bare stainless steel.

Relation between cleanability of stainless steel surfaces and water contact angle

1473

4.2.2. Oil removal. When an oil drop is removed from a solid surface by a shear flow with surfactants, the same decreasing relation between H and cos θa is found (Fig. 4) with different behaviors for polysiloxane coatings and bare stainless steels. The main difference in cleaning performance was observed on bare stainless steel surfaces (with R > 90% or 50% < R < 90%) and coated surfaces (with R < 50%). The difference observed between coated and bare surfaces could be attributed to an effect of the surface energy. In this case, surfaces soiled by oil were cleaned with a shear flow containing surfactants, then, on can suggest favorable interactions between polar sites on bare stainless steel surfaces and the head group of surfactants. On polysiloxane coatings, these interactions are not possible due the lack of polar sites [19]. In this case, the surface energetics governs the cleaning mechanisms. Difference in stainless steels exhibiting intermediate or good cleaning performance can be explained by a topographic effect. Ss2 and Ss9 are surfaces with many holes caused by chemically pickling, and oil impregnation occurs in the metal grain boundaries and this is not accessible to the detergent flow [19]. As previously, the retention force of oil soil depends on the water contact angle hysteresis whereas the force of soil removal is dependent on both surface energy and topography. Last, only droplet sliding and no rolling motion was observed in this work. For droplet-assisted particulate soil removal, it is generally known that whenever optimal cleaning or self-cleaning conditions are satisfied, it is the rolling motion of the drop that takes place, collecting the soil particles [28]. This could be explained in terms of topographic effect. Our surfaces display various topographies at a macroscale (typically 1 to 10 µm), whereas self-cleaning surfaces have fine textured topography (10 to 100 nm) or large-scale topography (100 µm) [29, 30]. Future work will be devoted to drop deformation under a shear flow and detachment of drops from model high and low energy solid surfaces by a shear flow of pure water or with surfactants.

5. CONCLUSIONS

The cleaning performance depends both on the surface energy of the material and its topography. In this respect, coatings can strongly modify the initial properties of bare materials, and also soil affinity and interaction with the cleaning agent. A better understanding of these phenomena will help to develop new materials. From a practical standpoint, there is no universal surface with ‘good cleaning performance’. However, the contact angle hysteresis is a relevant parameter for assessing the general cleanability of a surface. Cleanability of a material depends both on retention forces of the soil on the solid and removal forces induced by water drop sliding or a shear flow containing surfactants. After particulate or oil liquid soiling, the retention forces are directly linked to the water contact angle hysteresis parameter. Removing of particles by a water drop sliding or oil by a shear flow of surfactants depends on contact angle hysteresis.

1474

L. Boulangé-Petermann et al.

In order to design new surface treatments for improving the cleanability, or simply with the aim to use a general criterion for material selection, one might expect in the future that the hysteresis parameter could be used as a relevant cleaning performance index.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.

J. T. Holah and R. H. Thorpe, J. Appl. Bacteriol. 69, 599 (1990). R. O. Darouiche, Clin. Infect. Dis. 33, 1567 (2001). R. M. Donlan and J. W. Costerton, Clin. Microbiol. Rev. 15, 167 (2002). S. K. Hood and E. A. Zottola, Int. J. Food Microbiol. 37, 145 (1995). H. A. Viles and A. A. Gorbushina, Building Environ. 38, 1217 (2003). C. A. Pio, M. M. Ramos and A. C. Duarte, Atmospheric Environ. 32, 1979 (1998). C. W. Extrand and A. N. Gent, J. Colloid Interface Sci. 138, 431 (1990). R. E. Johnson Jr. and R. H. Dettre, in: Surface and Colloid Science, E. Matijevic (Ed.), Vol. 2, pp. 85–151. Wiley-Interscience, New York, NY (1969). S. Basu, K. Nandakumar and J. H. Masliyah, J. Colloid Interface Sci. 190, 253 (1997). C. W. Extrand and Y. Kumagai, J. Colloid Interface Sci. 170, 515 (1995). A. Carre and M. E. R. Shanahan, J. Adhesion 49, 177 (1995). C. M. Rossi, R. M. Esbert, F. Diaz-Pache and F. J. Alonso, Building Environ. 38, 147 (2003). I. Redsven, R. Kuisma, L. Laitala, E. Pesonen-Leinonen, R. Mahlberg, H. R. Kymäläinen, M. Hautala and A. M. Sjöberg, Tensides Surfactants Detergents 40, 346 (2003). H. K. Tikka, M. Suvanto and T. A. Pakkanen, J. Colloid Interface Sci. 273, 388 (2004). R. Foschino, C. Picozzi, E. Giorgi and A. Bontempi, Ann. Microbiol. 53, 253 (2003). R. Foschino, C. Picozzi, A. Civardo, M. Bandini and P. Faroldi J. Food Eng. 60, 375 (2003). L. Boulangé-Petermann, C. Jullien, P. E. Dubois, T. Benezech and C. Faille, Biofouling 20, 25 (2004). C. Jullien, T. Benezech, V. Lebret and C. Faille, J. Food Eng. 56, 77 (2003). L. Boulangé-Petermann, C. Gabet and B. Baroux, Colloids Surfaces A 272, 56 (2006). I. H. Pratt-Terpstra, A. H. Weerkamp and H. J. Busscher, J. Gen. Microbiol. 133, 3199 (1987). L. Boulangé-Petermann, C. Debacq, P. Poiret and B. Cromières, in: Contact Angle, Wettability and Adhesion, K. L. Mittal (Ed.), Vol. 3, pp. 501–519. VSP, Utrecht (2003). D. Gorse, J. C. Joud and B. Baroux, Corrosion Sci. 33, 1455 (1992). E. S. Adelmawla, M. M. Koura, T. M. A. Maksoud, I. M. Elewa and H. H. Soliman, J. Mater. Process Technol. 123, 133 (2002). L. Boulangé-Petermann, B. Baroux and M. N. Bellon-Fontaine, J. Adhesion Sci. Technol. 7, 221 (1993). M. Mantel and J. P. Wightman, Surf. Interface Anal. 34, 164 (2002). L. Boulangé-Petermann, A. Doren, B. Baroux and M. N. Bellon-Fontaine, J Colloid Interface Sci. 171, 179 (1995). D. Quéré, M. J. Azzopardi and L. Delattre, Langmuir 14, 2213 (1998). W. Barthlott and C. Neinhuis, Planta 202, 1 (1997) T. J. McCarthy and D. Oner, Langmuir 16, 7777 (2000). Z. Yoshimitu, A. Nakajima, T. Watanabe and K. Hashimoto, Langmuir 18, 5818 (2002).