Anti-Stiction Coatings in MEMS Devices
| Technology | Drift]
have been many exciting predictions that the future of micromachines or
microelectromechanical systems (MEMS) is just "around the corner",
but this future has proven to be slow in coming. Despite the demonstration
of numerous MEMS devices and product concepts each year, a very small
number have actually succeeded in the market place. Two prominent examples
are the digital mirror device (DMD) of Texas Instruments, Inc. and the
MEMS accelerometer of Analog Devices, Inc. While these two products are
vastly different, one can identify a common ingredient in their formula
of success: both companies successfully developed special surface coating
technologies [1, 2]. Indeed, the difficulty in controlling surface forces
is a critical impediment to the fabrication and operation of many MEMS
devices [3, 4]. This is a consequence of the scaling law: the surface-to-volume
ratio scales with the inverse of device dimension and surface forces dominate
at length scales < 1mm.
A well-known problem in the fabrication of MEMS devices from surface micromachining
is stiction, which occurs when surface adhesion forces are higher than
the mechanical restoring force of the micro-structure. When a device is
removed from the aqueous solution after wet etching of an underlying sacrificial
layer, the liquid meniscus formed on hydrophilic surfaces pulls the microstructure
towards the substrate and stiction occurs. While this release-stiction
problem may be alleviated by dry HF etching or supercritical CO2 drying,
a more difficult problem is in-use stiction which occurs during operation
when microstructures come into contact (intentionally or accidentally).
In-use stiction may be caused by capillary forces, electrostatic attraction,
and direct chemical bonding. To circumvent the stiction problem, many
MEMS developers are forced to switch to bulk micro-machining, which is
less capable and versatile than surface micromachining in terms of device
function. Even for devices from bulk-micromachining, in-use stiction is
still of concern, perhaps to a lesser extent.
One attractive approach to tackle the stiction problem is to provide low-energy
surface coating in the form of an organic passivation layer on the inorganic
surface [3, 4]. Such a coating can not only eliminate or reduce capillary
forces and direct chemical bonding, but also reduce electro-static forces
if the thin organic layer is directly applied to the semiconducting substrate,
without the intervening oxide layer [3,4 ]. Texas Instruments uses a fluorinated
fatty acid self-assembled monolayer (SAM) on the aluminum oxide surface
in their DMD , while Analog Devices coats the surfaces of their inertia
sensors using thermal evaporation of silicone polymeric materials at the
packaging stage after the device is completely released. Another much
advocated approach is the formation of siloxane self-assembled monolayers
(SAMs) on the oxide terminated surface, but the difficulty of this chemistry
and the poor reproducibility put significant limitations on its practical
In view of the critical importance of anti-stiction coatings in MEMS products,
we are developing a number of chemical processes that possess the following
attributes: (1) the chemistry is simple and reproducible; (2) the coatings
are of monolayer nature and are covalently bonded to the substrate; (3)
the coating processes are compatible with dry or aqueous etching processes;
(4) the monolayers are chemically and mechanically stable under conditions
of processing and operation. The design principle of the coating process
is illustrated in scheme 1. A key component in the coating chemical is
a molecule containing two major parts, R & X. The R group is
selected to provide low surface energy, i.e., "wax" or "Teflon"
like, while the X group is chosen to selectively react with the solid
surface of interest for covalent linkage. The attachment of these molecules
to the solid surface is a specially designed process which provides kinetic
control; the reaction self-terminates after a saturated monolayer coverage
is reached. This is very different from self-assembled monolayers (SAMs).
Such a selective and kinetically controlled reaction ensures that the
coating is uniform and conformal with solid surfaces in a MEMS device.
The thickness of the coating is chosen to be 1-2 nm.
Scheme 1. Design of low energy surface coating for MEMS devices. R
represents a molecular group to give low surface energy and X is a functional
group for selective attachment to the solid surface.
1 shows a group of poly-Si cantilever beams completed released by the
above coating process. The cantilever beam arrays (CBAs) are fabricated
by the SUMMiT process of Sandia National Laboratory. A cross section of
the structure is shown schematically in the upper part of the figure.
The poly-Si beams, with thickness of 2.25 mm, are anchored at one end.
The sacrificial PSG layer under each beam is ~2 mm thick. The bending
of p-Si beams on each CBA sample is characterized by interference microscopy.
Out-of-plane bending of the beams is shown as optical fringes; each fringe
corresponds to the bending of l/2 or 311 nm. The lower part of figure
1 shows typical images of beam arrays (2 mm in length) with alkoxyl coating.
All beams are completely released. The fringes reflect (downward) bending
due to residual stress. The maximum bending is ~1.8 mm. Therefore, the
above monolayer coating process eliminates release stiction. Electro-static
actuation of these coated beams also reveals little in-use stiction with
relative humidity as high as 90% .
1. Upper: schematic illustrate of a cantilever; lower: Interference
microscope image of cantilever beam arrays with beam length of 2
mm. The image (composite of three frames) is obtained after the
beams are released in an all-liquid etching-coating process.
 L. J. Hornbeck, US Patent 5602671 (1997).
 J. R. Martin and Y. Zhao, US Patent 694740 (1997).
 M. P. de Boer and T. M. Mayer, MRS Bulletin, 26, 302-304
 R. Maboudian, Surf. Sci. Rep. 30, 207-269 (1998).
 Y. S. Jun, X.-Y. Zhu, J. Adh. Sci. Technol. (special MEMS issue,
2003), in press.