How do lubricants really bond to hard drives?
Drs. Wilfred T. Tysoe & Nicholas D. Spencer | TLT Cutting Edge June 2012
Experiments reveal perfluoropolyether magnetic hard-drive lubricants react with dangling bonds on the surface of crevices in the carbon overcoat.
www.canstockphoto.com
THE REMARKABLE PROPERTIES OF MAGNETIC HARD DISK DRIVES rely on the presence of a thin (1-2 nm) layer of lubricant that is chemically bonded to the surface. These perfluoropolyether (PFPE) lubricants invariably comprise a backbone containing low-surface-energy tetrafluoroethylene oxide (-CF
2-CF
2-O-) and difluoromethylene oxide (-CF
2-O-) units that provide lubricity. Their hydrophobicity also protects against corrosion. The molecules are terminated at both ends by hydroxyl (-OH) groups that form weak hydrogen bonds with the surface.
Stronger bonds are also formed by a chemical reaction between the hydrogen atoms of the -OH group with so-called dangling bonds present on the surface of the carbon overcoat. The hydroxyl hydrogen atoms react with one dangling bond, and the resulting reactive terminal oxygen atom reacts with another to form a strong chemical bond with the surface.
The dangling bonds in the carbon overcoat can be detected using a technique known as electron spin resonance (esr), and it has been found that they disappear when the overcoat is exposed to the lubricant. It has also been found that one type of PFPE lubricant, Z-tetraol, which contains two hydroxyls on each end group, bonds to the carbon overcoat better than Z-dol, which contains only one, confirming the importance of hydroxyl groups in the surface-anchoring reaction.
It would appear, therefore, that lubricant bonding to carbon overcoats on magnetic hard drives is quite well understood. The conundrum, however, is that the dangling bonds on or near the surface also react quite quickly with water vapor or oxygen in the atmosphere. This implies that the dangling bonds required for chemical bonding are present only at a deeper level of the granule composite.
In order to test this hypothesis, Drs. Paul Kasai and Tsuyoshi Shimizu of Moresco Corp. in Japan measured the reactivity of a series of PFPE lubricants with similar backbones but with terminal -OH groups that were tethered via -CH
2- groups with differing lengths. The diameter of the hydrocarbon tethering groups (~0.27 nm) is less than that of the fluorocarbon backbone (~0.44 nm). Thus, if the thinner tethering groups can move into crevices between granules to reach a dangling bond, longer anchoring groups should be able to access them more easily.
Accordingly, they measured the rate at which the strongly bonded layer formed when using molecules with -OH groups separated from the PFPE backbone by one, four or six oxygen or -CH
2- groups and found that the strongly bonded layer formed the fastest with six groups and the most slowly with one, in accord with their postulate.
They also tested a molecule with five groups, one of which was -CH(OH)- group in addition to the terminal -OH group. If all dangling bonds were accessible, this should react more rapidly than molecules with a single OH terminus, while the steric hindrance afforded by the additional -OH group should cause it to react more slowly with dangling bonds in crevices. It was found to react more rapidly than the molecule with only one -CH
2- group but more slowly than the other molecules, further confirming the notion that reaction occurs in the crevices.
To rationalize the observations, they noted that the sputtered carbon overcoat is typically made up from granules that are about 3 nm in diameter. A simple geometrical model with three contacting spheres of 3 nm in diameter would create openings about 0.5 nm across. This is small enough to accommodate the thinner chains made of -CH
2- and -O- groups (~0.27 nm) but is about the same size as the diameter of the PFPE backbone (~0.44 nm).
While this geometrical model is clearly approximate, it does offer an intriguing rationale for their observations. They also note that such a model applies only to thermally cured films and not to those that are cured by ultraviolet light, where the photogenerated electrons interact directly with the PFPE chains, allowing them to bond to the surface.
FOR FURTHER READING:
Kasai, P.H., and Shimizu, T. (2012) “Bonding of Hard Disk Lubricants with OH-Bearing End Groups,”
Tribology Letters,
46 (1), pp. 43-47.
Eddy Tysoe is a Distinguished Professor of Physical Chemistry at the University of Wisconsin-Milwaukee. You can reach him at wtt@uwm.edu.
Nic Spencer is professor of surface science and technology at the ETH Zurich, Switzerland. Both serve as editors-in-chief of STLE-affiliated Tribology Letters journal. You can reach him at nspencer@ethz.ch.