3616
Langmuir 2004, 20, 3616-3622
Preparing Contamination-free Mica Substrates for Surface Characterization, Force Measurements, and Imaging Jacob N. Israelachvili,*,† Norma A. Alcantar,†,‡ Nobuo Maeda,† Thomas E. Mates,† and Marina Ruths§ Departments of Chemical Engineering and Materials Science and Materials Research Laboratory, University of California, Santa Barbara, California 93106, Department of Chemical Engineering, University of South Florida, Tampa, Florida 33620, and Department of Physical Chemistry, Åbo Akademi University, Porthansgatan 3-5, FIN-20500 Åbo, Finland Received July 17, 2003. In Final Form: January 21, 2004 Due to its perfect cleavage that provides large areas of molecularly smooth, chemically inert surfaces, mica is the most commonly used natural substrate in measurements with the surface forces apparatus (SFA), in atomic force microscopy (AFM), and in many adsorption studies. However, preparing mica surfaces that are truly clean is not easy since mica is a high-energy surface that readily adsorbs water, organic contaminants, and gases from the atmosphere. Mica can also become charged on cleaving, which makes it prone to picking up oppositely charged particles or mica flakes from the surroundings. High refractive index particles, such as metals, will adhere to mica through van der Waals forces. Recent articles have demonstrated that particle contamination is obtained when inappropriate cutting and handling procedures for the mica are used. In this paper, we show that both particle and other critical contamination is easy to detect and provide proper steps to take during the sample preparation process.
Introduction This paper discusses the issue of contamination of mica surfaces, which are commonly used in the surface forces apparatus (SFA) and related instruments, and procedures to detect and avoid various types of contamination. Monolayer and Submonolayer Contamination in SFA and AFM Experiments. In the case of contaminant monolayers, these can be detected down to fractions as low as 0.001 to 0.01 (0.1-1%) of a monolayer by X-ray photoelectron spectroscopy (XPS), electron spectroscopy for chemical analysis (ESCA) or secondary-ion mass spectrometry (SIMS). In SFA measurements using fringes of equal chromatic order (FECO) to measure surface separations,1 one can directly measure both the thickness and refractive index of such adsorbed films2 as well as their effects on the long-range colloidal and short-range adhesion forces.3 Previous tests4-6 have shown that, on prolonged exposure to laboratory air, a 0.3-0.4 nm fluid layer of water and common airborne organics adsorb to the mica surface from the time it is cleaved until it is installed in the SFA, atomic force microscopy (AFM), or other surface characterization chamber. Depending on the specific laboratory conditions, this layer may slowly build up in time, and it has even been seen to form 2-nmhigh fluid nanolenses on the mica surface.5,6 This layer is * To whom correspondence should be addressed: telephone (805) 893-8407; fax (805) 893-7870; e-mail
[email protected]. † University of California, Santa Barbara. ‡ University of South Florida. § Åbo Akademi University. (1) Tolansky, S. An introduction to interferometry, 2nd ed.; Longman, Green & Co.: New York, 1973. (2) Israelachvili, J. N. J. Colloid Interface Sci. 1973, 44, 259-272. (3) Israelachvili, J. N. Intermolecular and Surface Forces, 2nd ed.; Academic Press: San Diego, 1991. (4) Winterton, R. S. H. Thesis; University of Cambridge: Cambridge, UK, 1968. (5) Israelachvili, J. N.; Adams, G. E. J. Chem. Soc. Faraday Trans. 1 1978, 74, 975-1001. (6) Israelachvili, J. N. Surface Forces, Thesis; University of Cambridge: Cambridge, UK, 1971.
dissolved away when the surfaces are immersed in pure water and in some organic liquids,7-11 which can be readily ascertained from an inward shift of the contact position (D ) 0) by 0.5-0.8 nm in the liquid relative to the value measured in air.5,6 Additionally, since water has a low contact angle on mica both in air and in organic liquids, a water bridge forms between two mica surfaces when a droplet is injected between them. This water dissolves away the surface layer, which allows for the true contact position (D ) 0) to be determined, usually to within 0.10.2 nm, even when the (anhydrous) oil itself does not dissolve the contaminant layer. If adsorbed layers are present in AFM experiments, they will not only affect the measured adhesion force but also shift the apparent location of the contact position (D ) 0) to larger separations by an amount that is not easy to establish in the same way as in SFA experiments by the FECO optical technique. Particulate Contamination in SFA and AFM Experiments. Mica can become charged on cleaving,12,13 which makes it prone to picking up oppositely charged particles, ions, or mica flakes from the surroundings. High refractive index metals will adhere to the mica surfaces through van der Waals forces, either as lattice atoms that bind to the 50% of the potassium ion sites vacated on cleaving or as nanoparticles. Particles can come from at least four different sources (see below). They can be hard or soft, strongly or weakly adhering, isolated or densely packed, and of uniform or nonuniform size. This potential variety in the source, coverage, and chemical and physical properties of adsorbed contaminants can make it difficult to identify their existence in any experiment unless special (7) Christenson, H. K.; Israelachvili, J. N. J. Colloid Interface Sci. 1987, 117, 576-577. (8) Christenson, H. K. J. Phys. Chem. 1993, 97, 12034-41. (9) Ohnishi, S.; Hato, M.; Tamada, K.; Christenson, H. K. Langmuir 1999, 15, 3312-3316. (10) Raviv, U.; Laurat, P.; Klein, J. Nature 2001, 413, 51-54. (11) Raviv, U.; Klein, J. Science 2002, 297, 1540-1543. (12) Metsik, M. S. J. Adhesion 1972, 3, 307-314. (13) Mu¨ller, K.; Chang, C. G. Surf. Sci. 1969, 14, 39-51.
10.1021/la0352974 CCC: $27.50 © 2004 American Chemical Society Published on Web 03/23/2004
Preparing Contamination-free Mica Substrates
Langmuir, Vol. 20, No. 9, 2004 3617
Figure 1. Geometry of two mica sheets in adhesive contact with a particle of height h trapped between them. For typical values in SFA measurements, d . h, as given by eq 1 and measured experimentally.19
measurements or additional characterization techniques are taken, which is not always feasible. Their prevention and identification is therefore a matter of general interest and concern. In the case of hard particles, their presence can usually be readily identified during an AFM scan.9,14-16 In SFA experiments using the FECO optical technique, an isolated hard particle will distort the fringes from their straight, flattened shape over lateral distances of d > 10 µm even when the height of the trapped particle is only h ) 5-10 nm. This deformation can be readily seen as humps, spikes, or discontinuities in the FECO fringes (requiring that the measurements be moved to another, clean contact position). For a particle of height h trapped between two adhering mica surfaces (Figure 1) of adhesion or surface energy γ, thickness T, and elastic modulus E (55-65 GPa for mica17,18), the surfaces deform over a diameter given by19
d ) 2(Eh2T 3/γ)1/4
(1)
Since the lateral resolution (in the surface plane) in the FECO technique is ∼1 µm, a particle with a height of, say, 10 nm is easily identified from the ∼10 µm wide distorted shapes of the FECO fringes. In addition, in the presence of one or more particles between the surfaces, the measured colloidal and adhesion forces will be variable and differentsoften by orders of magnitudesfrom the thermodynamically expected values for molecularly smooth surfaces.20 All of these effects are readily measured in any routine SFA experiment, as exemplified in some recent measurements of the interactions between mica surfaces coated with dense layers of gold and platinum nanoparticles, whose forces and deformations were totally different from those of particlefree surfaces.21 FECO fringes can also be used to detect the presence of light-absorbing monolayers confined between the surfaces and give the twist angle between the crystallographic axes of the two mica sheets, which can affect the forces between them. These features are discussed further in connection with Figure 7. Hot-Wire Cutting of Mica Sheets for Use as Substrates. The possible deposition of platinum or molten (14) Kohonen, M. M.; Meldrum, F. C.; Christenson, H. K. Langmuir 2003, 19, 975-76. (15) Lin, Z.; Granick, S. Langmuir 2003, 19, 7061-7070. (16) Heuberger, M.; Za¨ch, M. Langmuir 2003, 19, 1943-1947. (17) McNeil, L. E.; Grimsditch, M. J. Phys. Condens. Matter 1993, 5, 1681-1690. (18) Vaughan, M. T.; Guggenheim, S. J. Geophys. Res. 1986, 91, 4657. (19) Stengl, R.; Mitani, K.; Lehmann, V.; Gosele, U. In Proc. IEEE 1989 SOS/SOI Technology Conference, 1989; pp 123-124. (20) Israelachvili, J.; Giasson, S.; Kuhl, T.; Drummond, C.; Berman, A.; Luengo, G.; Pan, J.-M.; Heuberger, M.; Ducker, W.; Alcantar, N. Proceedings of the 26th Leeds-Lyon Symposium, Tribology Series 2000, 38, 3-12. (21) Alcantar, N.; Park, C.; Pan, J.-M.; Israelachvili, J. Acta Materialia 2003, 51, 31-47.
mica particles on nearby mica surfaces during hot-wire cutting has recently been suggested to arise in all cases where mica sheets (substrates) are prepared in this way.9,14-16 The first systematic observations9,14 indicated that particles with diameter 30-150 nm and height 2-10 nm were formed on mica sheets during cutting. Their surface density (0.1-10 particles/µm2) was highest near the edges and also increased when thicker mica sheets were cut (because of the increased exposure time), but many of the particles were removed in water.9,14 In a couple of recent reports,15,16 particles were found to cover the whole hot-wire cut mica sheet, and to remain on the mica when rinsed by water. In one case the contamination was severe enough to give a reduction of the intensities of the even-order fringes (Figure 7a).15 Perhaps because of this severity, the particle density in this particular case did not depend on the mica thickness or on the position on the surface relative to the platinum wire.15 Hot-wire cutting was developed more than 35 years ago4 as a means of shaping thin (1-3 µm thick) mica sheets before they are placed on another, large backing sheet, where their contacting surfaces remain protected in adhesive contact until the sheets are peeled away for later use (while on the backing sheet, the exposed surfaces can be coated with a layer of silver for future FECO measurements). The molten edges allow for easy tweezer insertion prior to the peeling. Scissor cutting was rejected early on as a viable cutting method for three reasons: First, the surfaces were found to have small mica flakes on them, which are due to the charge exchange occurring during cutting that results in small flakes being attracted to and adhering electrostatically to the surfaces. Second, scissor cutting results in mica edges that are sharp and therefore not amenable to tweezer insertion and pick-up after the cut sheets are placed on a backing sheet. Third, mica is brittle and cleaves along its natural cleavage planes, so that unwanted cracks usually develop along directions that are different from the cutting direction. Hot-wire cutting serves as a very convenient method for preparing thin mica sheets of any given shape, which can then be stored in a desiccator in a clean vapor atmosphere or under vacuum more or less indefinitely for future use (the vacuum prevents contaminant buildup in capillary condensed water and organic vapors at mica edges that get dragged in during peeling). Given the suitability of mica for many surface studies and applications, and its already widespread use, it is therefore worth considering how wire-cut mica can be prepared to provide contaminant-free, especially particle-free, surfaces. Ohnishi et al.9 and Kohonen et al.14 recently reviewed and summarized the best procedure for obtaining particlefree mica sheets by hot-wire cutting. These procedures (see also refs 5 and 6) include using thin mica sheets (1-3 µm thick), cut to large areas (> 100-200 mm2), with thin Pt wire (diameter ∼0.25 mm or less), at low current and without overheating and breakage of the wire during cutting. The recommended wire length is 8 mm, so that the hot central region is no longer than 1-2 mm (avoiding excessive wire lengths such as the 60 mm used in ref 15). The mica cutting speed should be rapid (∼30 s/sheet), and each sheet should be turned over after cutting before being placed on the freshly cleaved backing sheet so that the upward-facing surface (cf. Figure 2B) adheres to the backing sheet. Many researchers now prefer to cut a very large piece (∼10 cm2, cutting time ∼1 min), which is placed on the backing sheet, from which a number of smaller pieces are later cut by scalpel. Most importantly, the mica mounting geometry during cutting must allow for free horizontal laminar flow of clean air at a high velocity (cf.
3618
Langmuir, Vol. 20, No. 9, 2004
Israelachvili et al.
Figure 2. (A) Inset: Incorrect way of mounting mica sheets for cutting, allowing for particles to become deposited on the top surface (main figure) due to obstructed air flow, and on the lower surface during hot-wire cutting. Reproduced with permission from ref 16, Figure 1A. Copyright 2003 American Chemical Society. (B) One of the recommended arrangements (schematic) for mounting a mica sheet for hot-wire cutting. The air flow should be from the back and parallel to the aligned metal blocks. Other conditions for producing particle-free mica sheets have been described in various previous publications14 and SFA Users Manuals.22,23
Figure 2), but not so high that the thin flexible sheets cannot be handled because of flutter. All thick sheets, used for cleaving or as backing sheets, must be handled only at the edges, and all items that touch the mica such as the supporting blocks and tweezers (see Figure 2B) should be degreased with ethanol before use. It is recommended that laminar flow hoods be checked every year. Free flow of air above and below the sheet is particularly important for preventing particles from settling on the surfaces, and exposure to stationary unfiltered air for only a few seconds can result in the settling of hundreds of particles per square millimeter. Figure 2B shows how to mount mica sheets to ensure free horizontal laminar flow of air, i.e., without obstructions, both above and below the sheets. If the air flow is obstructed, as shown in Figure 2A, a large number of various contaminants from the air, including molten mica and platinum particles, will land on both the upper and lower surfaces. Deposition from below during hot-wire cutting occurs because the hot wire heats the air locally, which therefore moves upward toward the mica surface, carrying contaminants and particles with it; hence the desirability of turning the sheets over before placing them on the backing sheet. Experimental Procedures Measurements of Particulate Contamination of HotWire Cut Mica Surfaces. Four of us have independently prepared and placed on four different large backing sheets a total of 28 sheets of hot-wire cut mica following the procedure each of us normally uses, which is close to that described in Figure 2B and laid down in various theses,4,6 manuals,22,23 and publications.5,9,14 In these tests, a variety of different cuts were made, including slow (>1 min) and fast (∼20 s) cuts, turned over and not turned over sheets, large (>2 cm2) and small (
