- Email: [email protected]
AFM adhesion force measurements on conversion-coated EN AW-6082-T6 aluminium
ARTICLE IN PRESS International Journal of Adhesion & Adhesives 29 (2009) 471–477
Contents lists available at ScienceDirect
International Journal of Adhesion & Adhesives journal homepage: www.elsevier.com/locate/ijadhadh
AFM adhesion force measurements on conversion-coated EN AW-6082-T6 aluminium B.S. Tanem a, O. Lunder a,, A. Borg b, J. Ma˚rdalen c a b c
SINTEF Materials and Chemistry, NO-7465 Trondheim, Norway Department of Physics, Norwegian University of Science and Technology, NO-7491 Trondheim, Norway SINTEF Petroleum Research, NO-7465 Trondheim, Norway
a r t i c l e in fo
abstract
Article history: Accepted 18 September 2008 Available online 10 October 2008
Atomic force microscopy (AFM) has been used to measure the adhesion force between functionalised AFM tips and smooth surfaces of an EN AW-6082-T6 aluminium alloy, both before and after application of different conversion coatings. In addition, the surface of a sapphire sample was studied as a model aluminium surface. The results obtained for the sapphire surface were highly reproducible, and were used as a mean to establish proper routines for the more complex industrial surfaces. The adhesion force between a chromate conversion-coated (CCC) EN AW-6082-T6 aluminium alloy and a COOH functionalised tip was significantly increased compared to the uncoated surface, probably as a result of strong hydrogen bonding. However, the adhesion force decreased with time during the first 24 h after treatment due to aging of the CCC. Chromate-free Ti–Zr-based treatment also increased the adhesion, but the adhesion force varied significantly due to non-uniform deposition and composition of the conversion coating. The measured AFM adhesion forces correlated qualitatively with macroscopic adhesion test results obtained previously for these specific conversion coatings. The AFM technique may thus provide useful information on the adhesion behaviour of heterogeneous conversion-coated aluminium surfaces. & 2008 Elsevier Ltd. All rights reserved.
Keywords: Aluminium and alloys Atomic force microscopy Chemical conversion
1. Introduction Chromate conversion coatings (CCCs) on aluminium have been successfully used for decades to provide corrosion resistance and adhesion to paints and adhesives. Alternative pre-treatments are being developed for environmental reasons and some of them are already in use. However, their efficiency and reliability are still a matter of discussion. The formation and properties of conversion coatings on aluminium may be rather complex due to the inherent heterogeneous microstructure of commercial alloys. In particular, surface treatment processes are significantly influenced by presence of intermetallic particles exhibiting electro-chemical properties different from the aluminium matrix. Commercially available chromate-free conversion coatings are often less robust than the CCCs [1] and are in general more affected by the composition and microstructure of the substrate. Lateral variations in the surface chemistry resulting from the subsurface microstructure may occur on conversion-coated surfaces [2]. In order to assess the influence of such variations on local adhesion to an organic coating, adhesion measurements
Corresponding author. Tel.: +47 9823 0423; fax: +47 7359 6892.
E-mail address: [email protected] (O. Lunder). 0143-7496/$ - see front matter & 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijadhadh.2008.09.005
with high lateral resolution are required. Atomic force microscopy (AFM) is particularly useful for such studies. The measured adhesion force (Fad) necessary to pull a functionalised AFM tip free from the surface is interpreted in terms of the specific interaction between molecules on the AFM tip and those on the surface. In general, Fad is the sum of several forces [3]: F ad ¼ F el þ F vdW þ F cap þ F chem
(1)
including electrostatic forces (Fel), van der Waals forces (FvdW), meniscus or capillary forces (Fcap) as well as forces due to chemical bonds or acid–base interactions (Fchem). A significant electrostatic contribution may occur on insulators and at low humidity, when charge dissipation is ineffective. The van der Waals forces are always present, and the sum of these forces is in most cases attractive. At ambient conditions, a water meniscus forms between the AFM tip and the substrate due to condensation and adsorption of a water film on the surfaces. The magnitude of this interaction depends both on relative humidity and the hydrophilic nature of the tip and the sample. Depending on the chemical end groups present on the tip and the substrate, chemical bonds may form during contact. Quantitative AFM adhesion force measurements require that measurements are performed under conditions where the
ARTICLE IN PRESS B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
electrostatic and capillary forces are minimized or can be corrected for. However, quantification is also limited by adsorption of contaminants on high-energy solid surfaces. Furthermore, surface roughness [4] and tip wear must be considered. In principle, the exact contact geometry needs to be known and procedures involving SEM or TEM are generally used to determine the tip geometry. Despite these complications, AFM force– distance curves have become an important method for studying adhesion properties with a lateral resolution at the nanometer level [3]. In the present work, we have used AFM to measure the adhesion force between functionalised AFM tips and smooth surfaces of an EN AW-6082-T6 aluminium alloy, both before and after conventional chromate and Ti–Zr-based pre-treatments. A chromated surface was examined because it represents a benchmark system for pre-treatment of aluminium. Furthermore, fundamental knowledge of the adhesion mechanisms of CCCs may be useful to develop more environmentally friendly alternatives. The Ti–Zr-based pre-treatment was chosen as a representative of a chrome-free alternative, which seems to have gained widespread acceptance industrially. Commercially available AFM tips with carboxyl (COOH) and methyl (CH3) functional groups were used. These groups represent typical hydrophilic (COOH) and hydrophobic (CH3) groups associated with the adhesion of organic coatings to oxidised aluminium. The AFM measurements were conducted on specimens prepared by ultramicrotomy prior to conversion coating to minimize surface roughness effects. Complementary measurements on a sapphire single-crystal surface were conducted to estimate effects of tip wear and relative humidity. The results are discussed in light of the wedge adhesion test results obtained previously for these specific conversion coatings [5].
2.2. AFM adhesion force measurements 2.2.1. General principle The basic principles of AFM adhesion force measurements have been described and discussed in the literature [3,7]. In short, the AFM measures the deflection of a cantilever with a functionalised tip (radius about 10–50 nm) as a function of displacement of an interacting surface as sketched in Fig. 1. Initially, the distance between the tip and the surface is large (A) and there is no interaction between the two. As the surface approaches the tip, van der Waals forces start to deflect the cantilever. At a certain point, when the attractive forces become large enough, the tip jumps onto the sample surface (B). The continued movement of the surface in contact with the tip causes a deflection of the cantilever in upwards direction, represented by the diagonal line in the deflection–displacement curve. When the movement is reversed (C), the cantilever tip adheres to the surface past the point of zero deflection (D) due to adhesion forces. Eventually, the spring jumps off the surface (E) and loses contact with the surface (F). The force required to pull the tip off the surface corresponds to the adhesion force (Fad) and is calculated from Hooke’s law: F ad ¼ kDx
(2)
where k is the spring constant of the cantilever and Dx is the maximum deflection of the cantilever (difference between E and F in Fig. 1).
2. Experimental 2.1. Materials, surface preparation and characterisation The substrate materials used included a single-crystal sapphire supplied by Goodfellow, and commercially extruded EN AW-6082T6 aluminium. Sapphire (Al2O3) represents an idealized aluminium surface for AFM adhesion force measurements due to its hardness and chemical stability. The RMS roughness of the sapphire was 0.089 nm, as determined from the tapping mode AFM image of a 100 100 nm2 area. Clean and smooth surfaces of the aluminium alloy (composition by weight 0.63% Mg, 1.0% Si, 0.53% Mn, 0.17% Fe, 0.0034% Cu, 0.016% Zn, 0.013% Ti, Al balance) were prepared by using a Reichert–Jung ultramicrotome and a diamond knife from MicroStar [6]. The ultramicrotomed surfaces were oriented parallel with the extrusion direction and generated close to the surface of the extrudate, i.e. within the 500 mm thick recrystallized layer covering the fibrous microstructure of the bulk material. The freshly generated EN AW-6082-T6 surfaces were subsequently treated according to the following conversion coating procedures:
Ti–Zr-based pre-treatment: deoxidation in 4% AlfideoxTM 73 for s
described above to enable an assessment of the change in adhesion force resulting from deposition of the conversion coatings. The composition of intermetallic particles present at the surface was determined by using a JXA-8900 Superprobe electron probe microanalyzer (EPMA) operated at 15 kV. The presence of conversion coatings was confirmed by using a Hitachi S-4300 SE field emission SEM equipped with an energy dispersive X-ray analysis system (EDS).
30 s, tap water rinsing, immersion in 4% Gardobond X4707 (H2TiF6–H2ZrF6 solution of pH 2.9, 20 1C) for 1 min, rinsing in distilled water and air drying. Chromating: deoxidation in 4% AlfideoxTM 73 for 30 s, tap water rinsing, immersion in 15 ml/l Alodines C6100 for 3 min, rinsing in distilled water and air drying.
An uncoated ultramicrotomed specimen was included as a reference. The reference surface was deoxidised and rinsed as
2.2.2. Instrumentation and tip preparation Adhesion force measurements in this work were performed by using a Nanoscope IIIa MultimodeTM AFM with commercially available carboxyl (COOH) and methyl (CH3) functionalised tips from Bioforce Nanoscience. These probes are made by coating Si3N4 AFM tips with a thin gold film, which is modified with a selfassembled monolayer of thiols with desired functionality (X–(CH2)n–SH, where SH is connected to the gold surface and X is the functional end group). The spring constant of the asreceived cantilevers was calibrated according to the method described by Cleveland [8], which involves attaching a known
C Cantilever deflection
472
0
F
D
A
B
E Displacement Fig. 1. Schematic curve of cantilever deflection versus displacement (distance to surface) in AFM adhesion force measurements.
ARTICLE IN PRESS B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
mass to the end of the cantilever (in this case a spherical tungsten particle) and measuring the resulting change in resonance frequency. The calculated spring constants for the COOH and CH3 functionalised tips were found to be 0.057 and 0.062 N/m, respectively. 2.2.3. Sapphire surface Measurements on sapphire were conducted to estimate effects of tip wear and relative humidity. No significant change in the adhesion force with time was observed during several hundred consecutive measurements, suggesting negligible wear of the tip surfaces. A gradual increase in the measured adhesion force would be expected due to an increase in the contacting area if significant tip wear had occurred. In general, Au-coated AFM tips functionalised with thiols are considered quite stable due to the inert nature of Au and its strong interaction with the –SH end group. Our experiments confirmed this behaviour. The relative humidity of ambient air had only a moderate effect on the measured adhesion force, Fad. By increasing the relative humidity from 15% to 80% the measured adhesion force increased by only about 5 nN for both the COOH and CH3 functionalised AFM tips, in agreement with the experience that the magnitude of capillary forces increases with relative humidity of the ambient air. 2.2.4. EN AW-6082-T6 aluminium surfaces Multiple sets of force–displacement curves with 5 s interval between each cycle were acquired at a scan rate of 5 mm/s, with a ramp delay time of 1 s (corresponding to position C in Fig. 1). These parameters were empirically determined to provide good reproducibility of the adhesion force measurements within a reasonable time. Maximum applied force, corresponding to the minimum sample-distance separation in the force–distance curves, was kept as constant and small as possible to minimize wear of the contacting surfaces. Adhesion force measurements on EN AW-6082-T6 aluminium were made on the matrix and intermetallic a-Al(Mn,Fe)Si particles at the ultramicrotomed surface, before and after conversion coating. The adhesion force (Fad) reported is the average value obtained from about 50 force–displacement curves, measured on a few intermetallic particles and locations on the matrix of each sample surface, respectively. The same functionalised tip was employed for the uncoated, chromated and Ti–Zrtreated surfaces to avoid effects of any variation in tip geometry from one tip to another. Adhesion force measurements are often performed in a liquid medium to avoid the capillary effect. This approach was not attempted in this study since hydration of the conversion-coated surfaces would occur in the presence of water, which would also influence the adhesion force measurements. The relative humidity during measurements was maintained at 13–15% by keeping the AFM in a closed chamber with silica gel as a drying agent. Since CCCs are known to undergo an aging process during storage in ambient air, all measurements on the EN AW-6082-T6 specimens were performed within 2 h after surface preparation, unless otherwise stated.
3. Results and discussion 3.1. Surface formation and characterisation Fig. 2 shows SEM images and EDS spectra of the ultramicrotomed (Fig. 2a) and conversion-coated aluminium surfaces in areas including a constituent particle. The particles were typically a few microns in size. EPMA showed that they consisted of
473
63–73% Al, 11–13% Mn, 10–16% Fe and 7–9% Si, corresponding to cubic a-Al(Fe,Mn)Si [9]. The main purpose of the conversion coatings is to promote adhesion of paints and adhesives to the aluminium substrate. Formation and characterisation of the presently applied chromate and Ti–Zr-based conversion coatings are described in detail elsewhere [2,10]. The CCC, formed by a redox reaction between chromate ions and aluminium, essentially consisted of a hydrated Cr(III)–Cr(VI) mixed oxide. The characteristic cracking of the coating was observed in SEM (Fig. 2b) and is due to dehydration of the coating. EDS analysis also revealed the presence of Cr on the a-Al(Fe,Mn)Si particles, in agreement with previous observations [10]. The Ti–Zr treatment resulted in deposition of a very thin (o20 nm) film which was barely visible in SEM. By X-ray EDS analysis of the Ti–Zr-treated surface (Fig. 2c) the conversion layer was detected only on the a-Al(Fe,Mn)Si particles. This is consistent with earlier observations of preferential deposition of the conversion layer on and around cathodic particles due to high pH developed locally at these sites during treatment [2]. Auger electron spectroscopy has furthermore shown that the continuous conversion layer covering the particles essentially consists of a hydrated Ti and Zr oxide mixture with a Ti/Zr ratio of about 5, which explains the apparent absence of the Zr Ka peak at 2.04 keV in the EDS spectrum in Fig. 2c. The presence of Al (oxy-) hydroxide and F-containing compounds was predominant on the Al matrix away from the particles, where a particulate type of deposit with low coverage was observed [2]. AFM topographic images of the aluminium surface before and after application of the conversion coatings are shown in Fig. 3. The images were acquired from areas away from the a-Al(Fe,Mn)Si particles to exclude the topographic features formed by cutting through hard and brittle particles embedded in the aluminium matrix. As expected from previous experience [6], the uncoated ultramicrotomed surface was very smooth, giving an indistinct appearance (Fig. 3a). Measurements of more than ten 300 300 nm2 areas on the ultramicrotomed surface showed an average RMS roughness (Rq) of 0.17 nm. The surface roughness increased considerably as a result of both conversion coating procedures due to oxide deposition as well as etching of the aluminium substrate. For the chromated (Fig. 3b) and Ti–Zr-treated (Fig. 3c) surfaces, average RMS roughness values of 16 and 5.4 nm were obtained, respectively. The particulate type of deposit typically formed by this Ti–Zr treatment [2] is clearly reflected in Fig. 3c.
3.2. Adhesion forces at freshly formed conversion coatings Fig. 4 shows examples of typical force–displacement curves obtained for the uncoated aluminium matrix interacting with the COOH and CH3 functionalised tips. The curves were constructed from measured cantilever deflection versus displacement data and the spring constant of the tips. A linear relationship was generally observed in the contact regime, and the approaching (dashed line) and retracting (solid line) part of the force curves were nearly identical, indicating negligible plastic or viscoelastic deformation of the samples [3]. A small oscillation of the cantilever at the pull-off point was sometimes observed (Fig. 4a). The magnitude of the observed adhesion forces indicates that many chemical bonds were involved in the contacting process. AFM adhesion forces measured by using the COOH functionalised tip on the aluminium alloy surfaces are summarised in Fig. 5. The same tip was used in all measurements to enable quantitative comparison between the surfaces [11]. For the uncoated ultramicrotomed surface an adhesion force of about 20 nN was
ARTICLE IN PRESS 474
B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
Fig. 2. SEM images of EN AW-6082-T6 aluminium surface and EDS analysis of intermetallic particles shown; (a) uncoated ultramicrotomed, (b) after chromate treatment and (c) after Ti–Zr-based treatment.
obtained, independent of whether the measurements were made on the a-Al(Fe,Mn)Si particles or on the aluminium matrix. After chromate treatment, the adhesion force increased by a factor of
three, both on the matrix and the particles. A significant increase in adhesion force was also observed as a result of Ti–Zr treatment, however, only on the aluminium matrix. The adhesion force
ARTICLE IN PRESS B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
475
Fig. 3. AFM topographic images of (a) uncoated ultramicrotomed (Rq ¼ 0.1770.03 nm), (b) chromated (Rq ¼ 15.873.2 nm) and (c) Ti–Zr-treated (Rq ¼ 5.470.5 nm) EN AW6082-T6 aluminium.
70 COOH
Adhesion force (nN)
60 50 40 30 20 10 0
UM (p)
UM (m) CCC (p) CCC (m) Ti-Zr (p) Ti-Zr (m) Surface treatment
Fig. 5. Average adhesion force7std. dev. measured on the a-Al(Fe,Mn)Si particles (p) and aluminium matrix (m) of the uncoated ultramicrotomed (UM), chromated (CCC) and Ti–Zr-treated (Ti–Zr) EN AW-6082-T6 surfaces. Measurements were performed within 2 h after surface treatment, using a COOH functionalised tip.
Fig. 4. Typical force–displacement curves obtained for the (a) COOH and (b) CH3 functionalised tip interacting with the uncoated aluminium matrix of EN AW-6082T6. The y-axis was calibrated with respect to the spring constant k of each tip.
measured on the a-Al(Fe,Mn)Si particles, corresponding to the locations with a thicker conversion layer, was practically the same as for the uncoated ultramicrotomed surface. The observed ranking in terms of local adhesion force (CCC 4Ti–Zr 4untreated) is in qualitative agreement with results obtained from macroscopic wedge adhesion tests performed on epoxy-bonded AA6060-T6 aluminium joints where the same conversion coatings were applied [5]. In general, comparing nanoscale observations with macroscopic behaviour should be considered with great caution. However, it is interesting to note that the lateral variations in adhesion force measured on the Ti–Zr-treated specimen were related to the local changes in conversion layer chemistry resulting from the presence of a-Al(Fe,Mn)Si particles on the surface. Conversely, a homogeneous
composition was obtained over the entire surface after Alodine treatment [10], consistent with the constant and reproducible adhesion force presently measured on both matrix and particles. The significant difference in adhesion force measured between particles and aluminium matrix after Ti–Zr-based treatment may be attributed to several factors. For a COOH functionalised tip, the adhesion force is primarily associated with hydrogen bond interactions [11]. Thus, local differences in the measured adhesion force may be attributed to variations in the density of surface hydroxyls capable of forming bonds with the COOH tip. However, the possible influence of capillary forces cannot be excluded. It has been shown [12,13], based on contact angle measurements and XPS analyses, that treatment of aluminium by immersion in hexafluorozirconate solution results in the formation of an Al–Zr–O–F-based hydrated layer that strongly increases the hydrophilicity of the surface. The increased adhesion force measured on the aluminium matrix as a result of Ti–Zr treatment (Fig. 5) may therefore partly be attributed to increased capillary forces generated by the water meniscus formed between the (hydrophilic) COOH tip and the hydrophilic aluminium surface [14]. Adhesion forces measured by using the CH3 functionalised tip are shown in Fig. 6. For all substrates, the adhesion forces were significantly smaller than those obtained with the COOH tip. Although quantitative comparison between the two tip functionalities requires that the exact contact area of the tips is known, the lower values obtained for the CH3 functionalised tip are not
ARTICLE IN PRESS 476
B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
20
Adhesion force (nN)
CH3 15
10
5
0 UM (p)
UM (m) CCC (p) CCC (m) Ti-Zr (p) Ti-Zr (m) Surface treatment
Fig. 6. Adhesion forces obtained by using a CH3 functionalised tip on the surfaces described in Fig. 5 (note change of scale).
unexpected since only van der Waals interactions are believed to contribute in this case [11]. These forces are much weaker than the hydrogen bonds. Fig. 6 further shows that there were quite small differences in adhesion force between the various treatments and locations on the aluminium surface. The only exception was the relatively low adhesion force on the aluminium matrix of the Ti–Zr-treated specimen, which contrasts the high adhesion force measured with the COOH tip (Fig. 5). Again, this observation indicates that a highly hydrophilic surface is formed by the Ti–Zr treatment, because the presence of a water film adsorbed on the hydrophilic surface should exert a repulsive force towards the hydrophobic CH3 functionalised tip [14]. The AFM adhesion force results provide direct evidence of a strong interaction between the CCC surface and the COOH tip, in accordance with the well-known adhesion-promoting capabilities of CCCs towards organic coatings in practice. CCCs are considered to be hydrophobic in nature [15], at least in comparison with the Ti–Zr-based treatment involving use of hexafluorozirconate. Therefore, the significantly higher adhesion forces measured on the CCC relative to the uncoated and Ti–Zr-treated surfaces cannot be attributed to stronger capillary force effects. Surface roughness on a scale smaller than the AFM tip (typical radius of the order 10–100 nm) is another factor that influences the adhesion force measured by AFM. Theoretical models describing the influence of roughness have been proposed, as reviewed by Butt [3]. The main effect of roughness is to avoid a closer contact between the tip and the specimen surface, thus reducing the adhesion force. Ultramicrotomy was presently used to minimize roughness and chemical modification of the aluminium surface during preparation. However, the surface roughness increased significantly as a result of conversion coating. Therefore, to the extent that roughness is influencing the measurements, the adhesion forces measured for the conversion-coated surfaces (Figs. 5 and 6) are probably underestimated. On a macroscopic scale, roughness and porosity of conversion coatings should generally increase the adhesion to organic coatings due to increased surface area and mechanical interlocking [16].
3.3. Effect of aging on the CCC The composition of a freshly formed CCC is often considered to comprise one or more of the phases Cr2O3 nH2O, CrOOH or Cr(OH)3 [17]. Thus, hydrogen bond interactions will be the
dominating forces providing adhesion, either to a COOH functionalised AFM tip [11] or to organic coatings [18]. During aging of CCCs in ambient air dehydration of the coating takes place, leading to significant losses in corrosion resistance and shrinkage cracking of the coatings. Relative humidity strongly affects the rate of dehydration, but the most significant changes typically occur within a few days in ambient atmosphere [17,19]. Dehydration and losses in corrosion resistance occur in a matter of minutes upon exposure to temperatures above 60 1C [20]. Dehydration of the CCC implies a reduction of possible hydrogen bond interactions with the COOH functionalised tip, suggesting a simultaneous loss of adhesion with time. Measurements performed after ambient aging at 40–50% relative humidity (18–22 1C) indeed showed that the adhesion force decreased with time. The major changes occurred during the first 24 h after treatment, when the adhesion force decreased by about 15 nN relative to the initial value obtained for a freshly formed coating. No significant change in the adhesion force was observed during continued storage for up to 300 h. This time scale corresponds reasonably well with XPS analyses [17], indicating no significant changes in the chemistry of the CCC (Alodines 1200S) for aging times longer than 40 h. Earlier [17,19] and present results demonstrate that CCCs are highly dynamic coatings, with corrosion and adhesion properties that may vary considerably in both short and long term. Industrial practice usually involves drying of CCCs and chromate-free alternatives at elevated temperatures (o100 1C) before painting. However, the influence of drying conditions on dehydration and adhesion of various conversion coatings to organic coatings is little known. AFM adhesion force measurements performed either in non-aqueous solution or inert atmosphere [21] to eliminate capillary effects between tip and sample may prove to be a useful technique for such studies.
4. Conclusions AFM with functionalised probe tips has been used to measure local adhesion forces towards EN AW-6082-T6 aluminium and sapphire (Al2O3), the latter used as a smooth and stable reference surface. Chromate treatment of the aluminium alloy surface significantly increased the adhesion to a COOH functionalised tip by hydrogen bonding. The pull-off adhesion force increased by a factor of three relative to the untreated surface, independent of the subsurface microstructure. Application of a chromate-free Ti–Zr-based treatment also increased adhesion, but the adhesion force varied significantly due to the presence of a-Al(Fe,Mn)Si particles causing non-uniform deposition of the conversion coating. The measured AFM adhesion forces correlated qualitatively with the macroscopic adhesion test results obtained previously for these specific conversion coatings. The technique may thus provide useful information about the adhesion behaviour of heterogeneous conversion-coated aluminium surfaces. However, strict control of the factors contributing to the measured adhesion force is required. In addition, aging of the conversion coating itself must be considered.
Acknowledgements M. Saxegaard and A. H. Berge, NTNU are acknowledged for measurements on sapphire surface. This work was part of a Norwegian national research program entitled Light Metal Surface Science, financed by The Norwegian Research Council, Hydro Aluminium, Profillakkering AS, Norsk Industrilakkering AS, Fundo Wheels AS, Jotun Powder Coatings AS and DuPont Powder Coating.
ARTICLE IN PRESS B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
References [1] Sheasby PG, Pinner R. The surface treatment and finishing of aluminium and its alloys. Ohio: ASM International, Materials Park; 2001. [2] Lunder O, Simensen C, Yu Y, Nisancioglu K. Surf Coat Technol 2004;184:278. [3] Butt HJ, Cappella B, Kappl M. Surf Sci Rep 2005;59:1. [4] Thoreson EJ, Martin J, Burnham NA. J Colloid Interface Sci 2006;298:94. [5] Lunder O, Lapique F, Johnsen B, Nisancioglu K. Int J Adhes Adhes 2004;24:107. [6] Tanem BS, Svenningsen G, Ma˚rdalen J. Corros Sci 2005;47:1506. [7] Drelich J, Tormoen GW, Beach ER. J Colloid Interface Sci 2004;280:484. [8] Cleveland JP, Manne S, Bocek D, Hansma PK. Rev Sci Instrum 1993;64:403. [9] Mondolfo LF. Aluminium alloys: their structure and properties. London: Butterworths; 1976. [10] Lunder O, Walmsley JC, Mack P, Nisancioglu K. Corros Sci 2005;47:1604.
[11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]
477
van der Vegte EW, Hadziioannou G. Langmuir 1997;13:4357. Chidambaram D, Clayton DR, Halada GP. J Electrochem Soc 2004;151:B151. Chidambaram D, Clayton DR, Halada GP. Electrochim Acta 2005;51:2862. Kaibara Y, Sugata K, Tachiki M, Umezawa M, Kawarada H. Diamond Relat Mater 2003;12:560. Kendig M, Jeanjaquet S, Addison R, Waldrop J. Surf Coat Technol 2001;140:58. van den Brand J, Van Gils S, Beentjes PCJ, Terryn H, Sivel V, de Wit JHW. Prog Org Coat 2004;51:339. Hughes AE, Taylor RJ, Hinton BRW. Surf Interface Anal 1997;25:223. Maeda S, Yamamoto M. Prog Org Coat 1998;33:83. Zhang W, Buchheit RG. Corrosion 2003;59:356. Laget V, Jeffcoate CS, Isaacs HS, Buchheit RG. J Electrochem Soc 2003;150:B425. Lubarsky GV, Davidson MR, Bradley RH. Surf Sci 2004;558:135.
Contents lists available at ScienceDirect
International Journal of Adhesion & Adhesives journal homepage: www.elsevier.com/locate/ijadhadh
AFM adhesion force measurements on conversion-coated EN AW-6082-T6 aluminium B.S. Tanem a, O. Lunder a,, A. Borg b, J. Ma˚rdalen c a b c
SINTEF Materials and Chemistry, NO-7465 Trondheim, Norway Department of Physics, Norwegian University of Science and Technology, NO-7491 Trondheim, Norway SINTEF Petroleum Research, NO-7465 Trondheim, Norway
a r t i c l e in fo
abstract
Article history: Accepted 18 September 2008 Available online 10 October 2008
Atomic force microscopy (AFM) has been used to measure the adhesion force between functionalised AFM tips and smooth surfaces of an EN AW-6082-T6 aluminium alloy, both before and after application of different conversion coatings. In addition, the surface of a sapphire sample was studied as a model aluminium surface. The results obtained for the sapphire surface were highly reproducible, and were used as a mean to establish proper routines for the more complex industrial surfaces. The adhesion force between a chromate conversion-coated (CCC) EN AW-6082-T6 aluminium alloy and a COOH functionalised tip was significantly increased compared to the uncoated surface, probably as a result of strong hydrogen bonding. However, the adhesion force decreased with time during the first 24 h after treatment due to aging of the CCC. Chromate-free Ti–Zr-based treatment also increased the adhesion, but the adhesion force varied significantly due to non-uniform deposition and composition of the conversion coating. The measured AFM adhesion forces correlated qualitatively with macroscopic adhesion test results obtained previously for these specific conversion coatings. The AFM technique may thus provide useful information on the adhesion behaviour of heterogeneous conversion-coated aluminium surfaces. & 2008 Elsevier Ltd. All rights reserved.
Keywords: Aluminium and alloys Atomic force microscopy Chemical conversion
1. Introduction Chromate conversion coatings (CCCs) on aluminium have been successfully used for decades to provide corrosion resistance and adhesion to paints and adhesives. Alternative pre-treatments are being developed for environmental reasons and some of them are already in use. However, their efficiency and reliability are still a matter of discussion. The formation and properties of conversion coatings on aluminium may be rather complex due to the inherent heterogeneous microstructure of commercial alloys. In particular, surface treatment processes are significantly influenced by presence of intermetallic particles exhibiting electro-chemical properties different from the aluminium matrix. Commercially available chromate-free conversion coatings are often less robust than the CCCs [1] and are in general more affected by the composition and microstructure of the substrate. Lateral variations in the surface chemistry resulting from the subsurface microstructure may occur on conversion-coated surfaces [2]. In order to assess the influence of such variations on local adhesion to an organic coating, adhesion measurements
Corresponding author. Tel.: +47 9823 0423; fax: +47 7359 6892.
E-mail address: [email protected] (O. Lunder). 0143-7496/$ - see front matter & 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijadhadh.2008.09.005
with high lateral resolution are required. Atomic force microscopy (AFM) is particularly useful for such studies. The measured adhesion force (Fad) necessary to pull a functionalised AFM tip free from the surface is interpreted in terms of the specific interaction between molecules on the AFM tip and those on the surface. In general, Fad is the sum of several forces [3]: F ad ¼ F el þ F vdW þ F cap þ F chem
(1)
including electrostatic forces (Fel), van der Waals forces (FvdW), meniscus or capillary forces (Fcap) as well as forces due to chemical bonds or acid–base interactions (Fchem). A significant electrostatic contribution may occur on insulators and at low humidity, when charge dissipation is ineffective. The van der Waals forces are always present, and the sum of these forces is in most cases attractive. At ambient conditions, a water meniscus forms between the AFM tip and the substrate due to condensation and adsorption of a water film on the surfaces. The magnitude of this interaction depends both on relative humidity and the hydrophilic nature of the tip and the sample. Depending on the chemical end groups present on the tip and the substrate, chemical bonds may form during contact. Quantitative AFM adhesion force measurements require that measurements are performed under conditions where the
ARTICLE IN PRESS B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
electrostatic and capillary forces are minimized or can be corrected for. However, quantification is also limited by adsorption of contaminants on high-energy solid surfaces. Furthermore, surface roughness [4] and tip wear must be considered. In principle, the exact contact geometry needs to be known and procedures involving SEM or TEM are generally used to determine the tip geometry. Despite these complications, AFM force– distance curves have become an important method for studying adhesion properties with a lateral resolution at the nanometer level [3]. In the present work, we have used AFM to measure the adhesion force between functionalised AFM tips and smooth surfaces of an EN AW-6082-T6 aluminium alloy, both before and after conventional chromate and Ti–Zr-based pre-treatments. A chromated surface was examined because it represents a benchmark system for pre-treatment of aluminium. Furthermore, fundamental knowledge of the adhesion mechanisms of CCCs may be useful to develop more environmentally friendly alternatives. The Ti–Zr-based pre-treatment was chosen as a representative of a chrome-free alternative, which seems to have gained widespread acceptance industrially. Commercially available AFM tips with carboxyl (COOH) and methyl (CH3) functional groups were used. These groups represent typical hydrophilic (COOH) and hydrophobic (CH3) groups associated with the adhesion of organic coatings to oxidised aluminium. The AFM measurements were conducted on specimens prepared by ultramicrotomy prior to conversion coating to minimize surface roughness effects. Complementary measurements on a sapphire single-crystal surface were conducted to estimate effects of tip wear and relative humidity. The results are discussed in light of the wedge adhesion test results obtained previously for these specific conversion coatings [5].
2.2. AFM adhesion force measurements 2.2.1. General principle The basic principles of AFM adhesion force measurements have been described and discussed in the literature [3,7]. In short, the AFM measures the deflection of a cantilever with a functionalised tip (radius about 10–50 nm) as a function of displacement of an interacting surface as sketched in Fig. 1. Initially, the distance between the tip and the surface is large (A) and there is no interaction between the two. As the surface approaches the tip, van der Waals forces start to deflect the cantilever. At a certain point, when the attractive forces become large enough, the tip jumps onto the sample surface (B). The continued movement of the surface in contact with the tip causes a deflection of the cantilever in upwards direction, represented by the diagonal line in the deflection–displacement curve. When the movement is reversed (C), the cantilever tip adheres to the surface past the point of zero deflection (D) due to adhesion forces. Eventually, the spring jumps off the surface (E) and loses contact with the surface (F). The force required to pull the tip off the surface corresponds to the adhesion force (Fad) and is calculated from Hooke’s law: F ad ¼ kDx
(2)
where k is the spring constant of the cantilever and Dx is the maximum deflection of the cantilever (difference between E and F in Fig. 1).
2. Experimental 2.1. Materials, surface preparation and characterisation The substrate materials used included a single-crystal sapphire supplied by Goodfellow, and commercially extruded EN AW-6082T6 aluminium. Sapphire (Al2O3) represents an idealized aluminium surface for AFM adhesion force measurements due to its hardness and chemical stability. The RMS roughness of the sapphire was 0.089 nm, as determined from the tapping mode AFM image of a 100 100 nm2 area. Clean and smooth surfaces of the aluminium alloy (composition by weight 0.63% Mg, 1.0% Si, 0.53% Mn, 0.17% Fe, 0.0034% Cu, 0.016% Zn, 0.013% Ti, Al balance) were prepared by using a Reichert–Jung ultramicrotome and a diamond knife from MicroStar [6]. The ultramicrotomed surfaces were oriented parallel with the extrusion direction and generated close to the surface of the extrudate, i.e. within the 500 mm thick recrystallized layer covering the fibrous microstructure of the bulk material. The freshly generated EN AW-6082-T6 surfaces were subsequently treated according to the following conversion coating procedures:
Ti–Zr-based pre-treatment: deoxidation in 4% AlfideoxTM 73 for s
described above to enable an assessment of the change in adhesion force resulting from deposition of the conversion coatings. The composition of intermetallic particles present at the surface was determined by using a JXA-8900 Superprobe electron probe microanalyzer (EPMA) operated at 15 kV. The presence of conversion coatings was confirmed by using a Hitachi S-4300 SE field emission SEM equipped with an energy dispersive X-ray analysis system (EDS).
30 s, tap water rinsing, immersion in 4% Gardobond X4707 (H2TiF6–H2ZrF6 solution of pH 2.9, 20 1C) for 1 min, rinsing in distilled water and air drying. Chromating: deoxidation in 4% AlfideoxTM 73 for 30 s, tap water rinsing, immersion in 15 ml/l Alodines C6100 for 3 min, rinsing in distilled water and air drying.
An uncoated ultramicrotomed specimen was included as a reference. The reference surface was deoxidised and rinsed as
2.2.2. Instrumentation and tip preparation Adhesion force measurements in this work were performed by using a Nanoscope IIIa MultimodeTM AFM with commercially available carboxyl (COOH) and methyl (CH3) functionalised tips from Bioforce Nanoscience. These probes are made by coating Si3N4 AFM tips with a thin gold film, which is modified with a selfassembled monolayer of thiols with desired functionality (X–(CH2)n–SH, where SH is connected to the gold surface and X is the functional end group). The spring constant of the asreceived cantilevers was calibrated according to the method described by Cleveland [8], which involves attaching a known
C Cantilever deflection
472
0
F
D
A
B
E Displacement Fig. 1. Schematic curve of cantilever deflection versus displacement (distance to surface) in AFM adhesion force measurements.
ARTICLE IN PRESS B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
mass to the end of the cantilever (in this case a spherical tungsten particle) and measuring the resulting change in resonance frequency. The calculated spring constants for the COOH and CH3 functionalised tips were found to be 0.057 and 0.062 N/m, respectively. 2.2.3. Sapphire surface Measurements on sapphire were conducted to estimate effects of tip wear and relative humidity. No significant change in the adhesion force with time was observed during several hundred consecutive measurements, suggesting negligible wear of the tip surfaces. A gradual increase in the measured adhesion force would be expected due to an increase in the contacting area if significant tip wear had occurred. In general, Au-coated AFM tips functionalised with thiols are considered quite stable due to the inert nature of Au and its strong interaction with the –SH end group. Our experiments confirmed this behaviour. The relative humidity of ambient air had only a moderate effect on the measured adhesion force, Fad. By increasing the relative humidity from 15% to 80% the measured adhesion force increased by only about 5 nN for both the COOH and CH3 functionalised AFM tips, in agreement with the experience that the magnitude of capillary forces increases with relative humidity of the ambient air. 2.2.4. EN AW-6082-T6 aluminium surfaces Multiple sets of force–displacement curves with 5 s interval between each cycle were acquired at a scan rate of 5 mm/s, with a ramp delay time of 1 s (corresponding to position C in Fig. 1). These parameters were empirically determined to provide good reproducibility of the adhesion force measurements within a reasonable time. Maximum applied force, corresponding to the minimum sample-distance separation in the force–distance curves, was kept as constant and small as possible to minimize wear of the contacting surfaces. Adhesion force measurements on EN AW-6082-T6 aluminium were made on the matrix and intermetallic a-Al(Mn,Fe)Si particles at the ultramicrotomed surface, before and after conversion coating. The adhesion force (Fad) reported is the average value obtained from about 50 force–displacement curves, measured on a few intermetallic particles and locations on the matrix of each sample surface, respectively. The same functionalised tip was employed for the uncoated, chromated and Ti–Zrtreated surfaces to avoid effects of any variation in tip geometry from one tip to another. Adhesion force measurements are often performed in a liquid medium to avoid the capillary effect. This approach was not attempted in this study since hydration of the conversion-coated surfaces would occur in the presence of water, which would also influence the adhesion force measurements. The relative humidity during measurements was maintained at 13–15% by keeping the AFM in a closed chamber with silica gel as a drying agent. Since CCCs are known to undergo an aging process during storage in ambient air, all measurements on the EN AW-6082-T6 specimens were performed within 2 h after surface preparation, unless otherwise stated.
3. Results and discussion 3.1. Surface formation and characterisation Fig. 2 shows SEM images and EDS spectra of the ultramicrotomed (Fig. 2a) and conversion-coated aluminium surfaces in areas including a constituent particle. The particles were typically a few microns in size. EPMA showed that they consisted of
473
63–73% Al, 11–13% Mn, 10–16% Fe and 7–9% Si, corresponding to cubic a-Al(Fe,Mn)Si [9]. The main purpose of the conversion coatings is to promote adhesion of paints and adhesives to the aluminium substrate. Formation and characterisation of the presently applied chromate and Ti–Zr-based conversion coatings are described in detail elsewhere [2,10]. The CCC, formed by a redox reaction between chromate ions and aluminium, essentially consisted of a hydrated Cr(III)–Cr(VI) mixed oxide. The characteristic cracking of the coating was observed in SEM (Fig. 2b) and is due to dehydration of the coating. EDS analysis also revealed the presence of Cr on the a-Al(Fe,Mn)Si particles, in agreement with previous observations [10]. The Ti–Zr treatment resulted in deposition of a very thin (o20 nm) film which was barely visible in SEM. By X-ray EDS analysis of the Ti–Zr-treated surface (Fig. 2c) the conversion layer was detected only on the a-Al(Fe,Mn)Si particles. This is consistent with earlier observations of preferential deposition of the conversion layer on and around cathodic particles due to high pH developed locally at these sites during treatment [2]. Auger electron spectroscopy has furthermore shown that the continuous conversion layer covering the particles essentially consists of a hydrated Ti and Zr oxide mixture with a Ti/Zr ratio of about 5, which explains the apparent absence of the Zr Ka peak at 2.04 keV in the EDS spectrum in Fig. 2c. The presence of Al (oxy-) hydroxide and F-containing compounds was predominant on the Al matrix away from the particles, where a particulate type of deposit with low coverage was observed [2]. AFM topographic images of the aluminium surface before and after application of the conversion coatings are shown in Fig. 3. The images were acquired from areas away from the a-Al(Fe,Mn)Si particles to exclude the topographic features formed by cutting through hard and brittle particles embedded in the aluminium matrix. As expected from previous experience [6], the uncoated ultramicrotomed surface was very smooth, giving an indistinct appearance (Fig. 3a). Measurements of more than ten 300 300 nm2 areas on the ultramicrotomed surface showed an average RMS roughness (Rq) of 0.17 nm. The surface roughness increased considerably as a result of both conversion coating procedures due to oxide deposition as well as etching of the aluminium substrate. For the chromated (Fig. 3b) and Ti–Zr-treated (Fig. 3c) surfaces, average RMS roughness values of 16 and 5.4 nm were obtained, respectively. The particulate type of deposit typically formed by this Ti–Zr treatment [2] is clearly reflected in Fig. 3c.
3.2. Adhesion forces at freshly formed conversion coatings Fig. 4 shows examples of typical force–displacement curves obtained for the uncoated aluminium matrix interacting with the COOH and CH3 functionalised tips. The curves were constructed from measured cantilever deflection versus displacement data and the spring constant of the tips. A linear relationship was generally observed in the contact regime, and the approaching (dashed line) and retracting (solid line) part of the force curves were nearly identical, indicating negligible plastic or viscoelastic deformation of the samples [3]. A small oscillation of the cantilever at the pull-off point was sometimes observed (Fig. 4a). The magnitude of the observed adhesion forces indicates that many chemical bonds were involved in the contacting process. AFM adhesion forces measured by using the COOH functionalised tip on the aluminium alloy surfaces are summarised in Fig. 5. The same tip was used in all measurements to enable quantitative comparison between the surfaces [11]. For the uncoated ultramicrotomed surface an adhesion force of about 20 nN was
ARTICLE IN PRESS 474
B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
Fig. 2. SEM images of EN AW-6082-T6 aluminium surface and EDS analysis of intermetallic particles shown; (a) uncoated ultramicrotomed, (b) after chromate treatment and (c) after Ti–Zr-based treatment.
obtained, independent of whether the measurements were made on the a-Al(Fe,Mn)Si particles or on the aluminium matrix. After chromate treatment, the adhesion force increased by a factor of
three, both on the matrix and the particles. A significant increase in adhesion force was also observed as a result of Ti–Zr treatment, however, only on the aluminium matrix. The adhesion force
ARTICLE IN PRESS B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
475
Fig. 3. AFM topographic images of (a) uncoated ultramicrotomed (Rq ¼ 0.1770.03 nm), (b) chromated (Rq ¼ 15.873.2 nm) and (c) Ti–Zr-treated (Rq ¼ 5.470.5 nm) EN AW6082-T6 aluminium.
70 COOH
Adhesion force (nN)
60 50 40 30 20 10 0
UM (p)
UM (m) CCC (p) CCC (m) Ti-Zr (p) Ti-Zr (m) Surface treatment
Fig. 5. Average adhesion force7std. dev. measured on the a-Al(Fe,Mn)Si particles (p) and aluminium matrix (m) of the uncoated ultramicrotomed (UM), chromated (CCC) and Ti–Zr-treated (Ti–Zr) EN AW-6082-T6 surfaces. Measurements were performed within 2 h after surface treatment, using a COOH functionalised tip.
Fig. 4. Typical force–displacement curves obtained for the (a) COOH and (b) CH3 functionalised tip interacting with the uncoated aluminium matrix of EN AW-6082T6. The y-axis was calibrated with respect to the spring constant k of each tip.
measured on the a-Al(Fe,Mn)Si particles, corresponding to the locations with a thicker conversion layer, was practically the same as for the uncoated ultramicrotomed surface. The observed ranking in terms of local adhesion force (CCC 4Ti–Zr 4untreated) is in qualitative agreement with results obtained from macroscopic wedge adhesion tests performed on epoxy-bonded AA6060-T6 aluminium joints where the same conversion coatings were applied [5]. In general, comparing nanoscale observations with macroscopic behaviour should be considered with great caution. However, it is interesting to note that the lateral variations in adhesion force measured on the Ti–Zr-treated specimen were related to the local changes in conversion layer chemistry resulting from the presence of a-Al(Fe,Mn)Si particles on the surface. Conversely, a homogeneous
composition was obtained over the entire surface after Alodine treatment [10], consistent with the constant and reproducible adhesion force presently measured on both matrix and particles. The significant difference in adhesion force measured between particles and aluminium matrix after Ti–Zr-based treatment may be attributed to several factors. For a COOH functionalised tip, the adhesion force is primarily associated with hydrogen bond interactions [11]. Thus, local differences in the measured adhesion force may be attributed to variations in the density of surface hydroxyls capable of forming bonds with the COOH tip. However, the possible influence of capillary forces cannot be excluded. It has been shown [12,13], based on contact angle measurements and XPS analyses, that treatment of aluminium by immersion in hexafluorozirconate solution results in the formation of an Al–Zr–O–F-based hydrated layer that strongly increases the hydrophilicity of the surface. The increased adhesion force measured on the aluminium matrix as a result of Ti–Zr treatment (Fig. 5) may therefore partly be attributed to increased capillary forces generated by the water meniscus formed between the (hydrophilic) COOH tip and the hydrophilic aluminium surface [14]. Adhesion forces measured by using the CH3 functionalised tip are shown in Fig. 6. For all substrates, the adhesion forces were significantly smaller than those obtained with the COOH tip. Although quantitative comparison between the two tip functionalities requires that the exact contact area of the tips is known, the lower values obtained for the CH3 functionalised tip are not
ARTICLE IN PRESS 476
B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
20
Adhesion force (nN)
CH3 15
10
5
0 UM (p)
UM (m) CCC (p) CCC (m) Ti-Zr (p) Ti-Zr (m) Surface treatment
Fig. 6. Adhesion forces obtained by using a CH3 functionalised tip on the surfaces described in Fig. 5 (note change of scale).
unexpected since only van der Waals interactions are believed to contribute in this case [11]. These forces are much weaker than the hydrogen bonds. Fig. 6 further shows that there were quite small differences in adhesion force between the various treatments and locations on the aluminium surface. The only exception was the relatively low adhesion force on the aluminium matrix of the Ti–Zr-treated specimen, which contrasts the high adhesion force measured with the COOH tip (Fig. 5). Again, this observation indicates that a highly hydrophilic surface is formed by the Ti–Zr treatment, because the presence of a water film adsorbed on the hydrophilic surface should exert a repulsive force towards the hydrophobic CH3 functionalised tip [14]. The AFM adhesion force results provide direct evidence of a strong interaction between the CCC surface and the COOH tip, in accordance with the well-known adhesion-promoting capabilities of CCCs towards organic coatings in practice. CCCs are considered to be hydrophobic in nature [15], at least in comparison with the Ti–Zr-based treatment involving use of hexafluorozirconate. Therefore, the significantly higher adhesion forces measured on the CCC relative to the uncoated and Ti–Zr-treated surfaces cannot be attributed to stronger capillary force effects. Surface roughness on a scale smaller than the AFM tip (typical radius of the order 10–100 nm) is another factor that influences the adhesion force measured by AFM. Theoretical models describing the influence of roughness have been proposed, as reviewed by Butt [3]. The main effect of roughness is to avoid a closer contact between the tip and the specimen surface, thus reducing the adhesion force. Ultramicrotomy was presently used to minimize roughness and chemical modification of the aluminium surface during preparation. However, the surface roughness increased significantly as a result of conversion coating. Therefore, to the extent that roughness is influencing the measurements, the adhesion forces measured for the conversion-coated surfaces (Figs. 5 and 6) are probably underestimated. On a macroscopic scale, roughness and porosity of conversion coatings should generally increase the adhesion to organic coatings due to increased surface area and mechanical interlocking [16].
3.3. Effect of aging on the CCC The composition of a freshly formed CCC is often considered to comprise one or more of the phases Cr2O3 nH2O, CrOOH or Cr(OH)3 [17]. Thus, hydrogen bond interactions will be the
dominating forces providing adhesion, either to a COOH functionalised AFM tip [11] or to organic coatings [18]. During aging of CCCs in ambient air dehydration of the coating takes place, leading to significant losses in corrosion resistance and shrinkage cracking of the coatings. Relative humidity strongly affects the rate of dehydration, but the most significant changes typically occur within a few days in ambient atmosphere [17,19]. Dehydration and losses in corrosion resistance occur in a matter of minutes upon exposure to temperatures above 60 1C [20]. Dehydration of the CCC implies a reduction of possible hydrogen bond interactions with the COOH functionalised tip, suggesting a simultaneous loss of adhesion with time. Measurements performed after ambient aging at 40–50% relative humidity (18–22 1C) indeed showed that the adhesion force decreased with time. The major changes occurred during the first 24 h after treatment, when the adhesion force decreased by about 15 nN relative to the initial value obtained for a freshly formed coating. No significant change in the adhesion force was observed during continued storage for up to 300 h. This time scale corresponds reasonably well with XPS analyses [17], indicating no significant changes in the chemistry of the CCC (Alodines 1200S) for aging times longer than 40 h. Earlier [17,19] and present results demonstrate that CCCs are highly dynamic coatings, with corrosion and adhesion properties that may vary considerably in both short and long term. Industrial practice usually involves drying of CCCs and chromate-free alternatives at elevated temperatures (o100 1C) before painting. However, the influence of drying conditions on dehydration and adhesion of various conversion coatings to organic coatings is little known. AFM adhesion force measurements performed either in non-aqueous solution or inert atmosphere [21] to eliminate capillary effects between tip and sample may prove to be a useful technique for such studies.
4. Conclusions AFM with functionalised probe tips has been used to measure local adhesion forces towards EN AW-6082-T6 aluminium and sapphire (Al2O3), the latter used as a smooth and stable reference surface. Chromate treatment of the aluminium alloy surface significantly increased the adhesion to a COOH functionalised tip by hydrogen bonding. The pull-off adhesion force increased by a factor of three relative to the untreated surface, independent of the subsurface microstructure. Application of a chromate-free Ti–Zr-based treatment also increased adhesion, but the adhesion force varied significantly due to the presence of a-Al(Fe,Mn)Si particles causing non-uniform deposition of the conversion coating. The measured AFM adhesion forces correlated qualitatively with the macroscopic adhesion test results obtained previously for these specific conversion coatings. The technique may thus provide useful information about the adhesion behaviour of heterogeneous conversion-coated aluminium surfaces. However, strict control of the factors contributing to the measured adhesion force is required. In addition, aging of the conversion coating itself must be considered.
Acknowledgements M. Saxegaard and A. H. Berge, NTNU are acknowledged for measurements on sapphire surface. This work was part of a Norwegian national research program entitled Light Metal Surface Science, financed by The Norwegian Research Council, Hydro Aluminium, Profillakkering AS, Norsk Industrilakkering AS, Fundo Wheels AS, Jotun Powder Coatings AS and DuPont Powder Coating.
ARTICLE IN PRESS B.S. Tanem et al. / International Journal of Adhesion & Adhesives 29 (2009) 471–477
References [1] Sheasby PG, Pinner R. The surface treatment and finishing of aluminium and its alloys. Ohio: ASM International, Materials Park; 2001. [2] Lunder O, Simensen C, Yu Y, Nisancioglu K. Surf Coat Technol 2004;184:278. [3] Butt HJ, Cappella B, Kappl M. Surf Sci Rep 2005;59:1. [4] Thoreson EJ, Martin J, Burnham NA. J Colloid Interface Sci 2006;298:94. [5] Lunder O, Lapique F, Johnsen B, Nisancioglu K. Int J Adhes Adhes 2004;24:107. [6] Tanem BS, Svenningsen G, Ma˚rdalen J. Corros Sci 2005;47:1506. [7] Drelich J, Tormoen GW, Beach ER. J Colloid Interface Sci 2004;280:484. [8] Cleveland JP, Manne S, Bocek D, Hansma PK. Rev Sci Instrum 1993;64:403. [9] Mondolfo LF. Aluminium alloys: their structure and properties. London: Butterworths; 1976. [10] Lunder O, Walmsley JC, Mack P, Nisancioglu K. Corros Sci 2005;47:1604.
[11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]
477
van der Vegte EW, Hadziioannou G. Langmuir 1997;13:4357. Chidambaram D, Clayton DR, Halada GP. J Electrochem Soc 2004;151:B151. Chidambaram D, Clayton DR, Halada GP. Electrochim Acta 2005;51:2862. Kaibara Y, Sugata K, Tachiki M, Umezawa M, Kawarada H. Diamond Relat Mater 2003;12:560. Kendig M, Jeanjaquet S, Addison R, Waldrop J. Surf Coat Technol 2001;140:58. van den Brand J, Van Gils S, Beentjes PCJ, Terryn H, Sivel V, de Wit JHW. Prog Org Coat 2004;51:339. Hughes AE, Taylor RJ, Hinton BRW. Surf Interface Anal 1997;25:223. Maeda S, Yamamoto M. Prog Org Coat 1998;33:83. Zhang W, Buchheit RG. Corrosion 2003;59:356. Laget V, Jeffcoate CS, Isaacs HS, Buchheit RG. J Electrochem Soc 2003;150:B425. Lubarsky GV, Davidson MR, Bradley RH. Surf Sci 2004;558:135.