130 doi:10.1017/S1431927605051068

Microsc Microanal 11 (supp 3), 2005 Copyright 2005, LASPM

Adhesion Forces for Mica and Silicon Oxide Surfaces Studied by Atomic Force Spectroscopy (AFS) F. L. Leite*,**, E. C. Ziemath***, O. N. Oliveira Jr.**, and P. S. P. Herrmann* *Embrapa Agricultural Instrumentation; Rua XV de Novembro, 1452, PO Box 741, São Carlos, SP, 13560-970, Brasil. e-mail: [email protected] **Institute of Physics of São Carlos, Universidade de São Paulo (USP), PO Box 369, São Carlos, SP, 13560-970, Brasil. ***Institute of Geosciences and Exact Sciences, Universidade de São Paulo (USP), PO Box 178, Rio Claro, SP, 13550-970, Brasil. Keywords: adhesion forces, atomic force spectroscopy, roughness, van der Waals forces, silicon, mica, atomic force microscopy The possibility of analyzing surfaces at the nanoscale provided by atomic force microscopy [1] (AFM) has been explored for various materials, including polymers [2], biological materials [3] and clays [4]. Further uses of AFMs involved nanomanipulation [5] and measurements of interaction forces, where the latter has been referred to as atomic force spectroscopy (AFS) [6]. Measurements of surface-surface interactions at the nanoscale are important because many materials have their properties changed at this range [7]. For samples in air, the interactions with the tip are a superimposition of van der Waals, electrostatic and capillary forces. A number of surface features can now be monitored with AFS, such as adsorption processes and contamination from the environment. Many implications exist for soil sciences and other areas, because quantitative knowledge of particle adhesion is vital for understanding technological processes, including particle aggregation in mineral processing, quality of ceramics and adhesives. In this paper, we employ AFS to measure adhesion (pull-off force) between the AFM tip and two types of substrate. Adhesion maps are used to illustrate sample regions that had been contaminated with organic compounds. Muscovite mica and silicon wafers were used as samples. Mica can be easily cleaved to yield an atomically planar surface (surface roughness ≈ 0.1 nm) (see Figure 1a). The mica used here was kindly supplied by Dr. Jane Frommer from IBM Almaden Research Center, San Jose, CA (USA). Silicon substrates were cleaned with a piranha solution. Because the measurements were carried out in air, the samples are actually covered by a thin-film of silicon oxide. All measurements were carried out on a Topometrix TMX 2010 Discoverer AFM, operating in the contact mode. The silicon cantilevers have a spring constant of k = 0.11±0.02 N/m with tip curvature radius, Rt = 30±5 nm. The force curves were obtained by measuring the vertical displacement of the sample – driven by the piezoscanner - and the deflection of the cantilever with respect to its position at rest. The curves were acquired under ambient conditions of 46±3% relative humidity and at 25±1oC. They were obtained at 100 points equally spaced from each other over the sample scanned area. Each force curve comprised a row of a maximum of 250 data points collected during the vertical movements of approach and retraction of the cantilever. Statistica software (StatSoft, 1999 version) was used to create the adhesion maps. The adhesion force between an AFM tip and mica or silicon oxide substrates was measured in air. Adhesion depends strongly on both the surface roughness and type of material. Figure 1b shows

Microsc Microanal 11 (supp 3), 2005

131

histograms obtained from force curve measurements performed several times at the same point for the two types of substrates, where the cantilever deflection (i.e. proportional to the adhesion force) changes considerably from mica to silicon oxide. For mica, with a roughness ∼ 0.1 nm for a scanned area of 1µm x 1µm, the variability was at most 1.7 %, while for silicon oxide which had a roughness ∼1 nm this variability was ≥ 5.0 %. Therefore, roughness and surface energy affected the magnitude of pull-off forces. The adhesion force, Fad, can be estimated from the cantilever deflection using Hooke’s law, Fad = kδ max where k and δmax are the elastic constant and the maximum deflection of the cantilever, respectively. Average values of the adhesion force between the silicon tip and the substrates were 26.6±0.4 nN and 19.0±1.7 nN for mica and silicon oxide substrates, respectively. As expected the dispersion is higher for the rougher substrate. The values of adhesion forces varied when different spots were analyzed, as shown in the adhesion maps. Part of this variability was associated with the roughness of the substrate, as commented upon above. Indeed, it is known that surface properties in the nanoscale, such as adhesion, are affected by the surface topography [8], with lowering roughness leading to a decrease in adhesion force [9]. It should also be stressed that the changes in the measured adhesion forces can be due to heterogeneity in the contact area caused by the geometry of the tip and surface roughness, in addition to capillary effects that depend on the meniscus radius. For a surface with nm roughness cannot be considered as a flat plane because the radius of the tip is also in the nanoscale. For example, when the tip approaches a peak region of the substrate, the measured adhesion is artificially lowered because the contact area between the tip and surface is small. The homogeneity of sample surfaces, as featured in adhesion maps, can be further decreased if the surface is contaminated. We have proven this statement by exposing the mica and silicon oxide substrates to ambient air before the force curve measurements were taken. Figure 2 shows two adhesion maps from distinct regions of mica and silicon oxide after 2 hours of air exposure with constant humidity. Table 1 summarizes the findings in terms of average adhesion forces and surface roughness. The average adhesion forces vary considerably from one region to the other for both substrates, even though the roughness does not change. For mica, in particular, changes in adhesion due to changes in roughness should be at most 17% [10], whereas a change of 33% was observed (Table 1). In subsidiary x-ray photoelectron spectroscopy experiments (results not shown), the contamination was identified as arising from organic compounds. Contamination is not homogeneous, as indicated by the adhesion maps, and the regions with lower adhesion forces are those that were probably most contaminated, since mica becomes hydrophobic upon prolonged exposure to air [11]. The tip is used as a nanometer-scale adhesion tester with which one measures the force required to remove the tip from intimate contact with the surface. The pull-off force is thus proportional to the local adhesion energy. If the surfaces of both substrate (mica and silicon oxide) and AFM tip were hydrophilic, a high compatibility and larger water meniscus could be formed around the tip, resulting in a strong adhesion force. In contrast, if the substrate surface is hydrophobic the adhesion force is minimized due to the high interfacial tension for the substrate-tip system, which is indicative of low compatibility. Thus the compatibility degree between an AFM tip and mica or silicon plate can be evaluated through force spectroscopy. In order to do this, we performed two pieces of experiment: the first with silicon oxide plates hydrophobized by deposited hydrocarbons and the second with plates that were previously cleaned in H2SO4/H2O2, 7:3 v/v solutions for 1h, followed by extensive washing in ultra-pure water. The results in the figure showed a considerable increase in the adhesion force after cleaning. Therefore, the large adhesion found for

132 doi:10.1017/S1431927605051068

Microsc Microanal 11 (supp 3), 2005 Copyright 2005, LASPM

the clean silicon plates is due to the low interfacial tension, which may provide a quantitative measurement of substrate cleanliness. [12] References [1] G. Binnig et al., Phys. Rev. Lett. 56 (1986) 930. [2] R. F. Lobo et al., Nanotechnology 14, (2003) 101. [3] D. Leckband and J. Israelachvili, Quartely Rev. Biophys. 34 (2001) 105. [4] B. R. Bickmore et al., Am. Mineral. 87 (2002) 780. [5] Q. Tang et al., J. Nanosci. Nanotechnol. 4 (2004) 948. [6] F. L. Leite and P. S. P Herrmann., J. Adhesion Sci. Technol. 19 (2005) 365. [7] R. García and R. Pérez, Surf. Sci. Rep. 47 (2002) 197. [8] K. L. Johnson, Tribology International 31 (1998) 413. [9] K. N. G. Fuller and D. Tabor, Proc. Roy. Soc. London A 345 (1975) 327. [10] F. L. Leite et al., J Adhesio Sci. Technol. 17 (2003) 2141. [11] J. Hu et al., Surface Science 344 (1995) 221. [12] This work was supported by FAPESP, CNPq and CT-Hidro (Brazil). TABLE 1. Adhesion force and roughness values. Adhesion force (nN) Material Region 1 Region 2 Mica 16±3 23±3 Silicon oxide 12±3 23±13

Roughness (nm) Region 1 Region 2 0.13 0.11 1.09 1.21

10

Mica Silicon

9 8

(233±4) nm

Frequence (%)

7 6 5 4

(178±11) nm 3 2 1 0

(a)

140

160

180

200

220

240

260

cantilever deflection (nm)

(b) Fig. 1. (a) AFM image of mica and (b) histogram illustrating how the pull-off force measurements taken at the same point vary for mica and silicon oxide substrates. The values of adhesion force were 26.6±0.4 nN and 19.0±1.7 nN for mica and silicon, respectively.

Microsc Microanal 11 (supp 3), 2005

133

Po sit

ion (nm )

ition Pos

(nm

Adhesion force (nN) 22 23 24 above

g

adhesion force (nN) 15 16 17 above

Po

)

sit ion

(n

m)

(i)

n itio pos

) (nm

(ii)

(a) g

Adhesion force (nN)

Adhesion force (nN)

9,9 14,9 19,9 24,9 29,9 34,9 above

9,9 10,9 11,9 above

Po

sit ion

Po sit io

(n m)

n (n itio Pos

(i)

m)

m) n (n itio Po s

n( nm )

(ii)

(b) Fig. 2. Adhesion maps onto mica (a): (i) region 1, (ii) region 2; and onto silicon (b): (i) region 1, (ii) region 2. Each adhesion map corresponds to a scanning area of 1 µm2.