The built-in strain in deposited polysilicon films depends on film morphology. The strain may be compressive or tensile. Its magnitude and uniformity may be measured by using surface micro machining with test structures which require a single photomask level and use Euler buckling. Anneal cycles may be used to convert compressive strain to tensile or zero strain. Deposition and anneal cycles may be optimized to produce fine grained polysilicon with isotropic mechanical behavior.

Strain field control can improve yield by reducing surface sticking. This mechanical stiffening is used in, for instance, pressure transducer construction. Built-in and applied axial strain are used to change the resonant frequencies of clamped-clamped beams in a vacuum envelope where quality factors up to 300,000 have been measured.

We present a MEMS microinstrument to interface submicron-scale structures and study their properties. Using this microinstrument integrated with a 150nm diameter silicon fiber that has high fracture strength and a high force micro-actuator, the non-linearity of n-type silicon's piezoresistive effect is measured up to the third order, using data from the high strain range of 0-1% for the first time.

The behavior of MEMS devices is limited by the strength of critical features such as thin ligaments, oxide cuts joining layers, pin joints and hinges. Devices fabricated at Sandia's Microelectronic Development Laboratory have been successfully tested to investigate these features. A series of measurements were performed on samples with gage lengths of 15 to 1000 microns, using conventional and tungsten coated samples as well as samples that include the critical features of standard components in the test section. Specimens have a freely moving pin joint on one end that anchors the sample to the silicon die to allow rotation to reduce effects of bending. Each sample is loaded in uniaxial tension by pulling laterally with a flat tipped diamond in a computer-controlled Nanoindenter. Load is calculated by resolving the measured lateral and normal forces into the applied tensile force and frictional losses. The specimen cross section and gage length dimensions were verified by measuring against a standard in the SEM. Multiple tests can be programmed at one time and performed without operator assistance allowing the collection of significant populations of data.

Determination of the mechanical properties of MEMS (microelectromechanical systems) materials is necessary for accurate device design and reliability prediction. This is most unambiguously performed using MEMS-fabricated test specimens and MEMS loading devices. We describe here a wafer-level technique for measuring the bend strength, fracture toughness, and tensile strength of MEMS materials. The bend strengths of surface-micromachined polysilicon, amorphous silicon, and polycrystalline 3C SiC are 5.1±1.0, 10.1±2.0, and 9.0±1.0 GPa, respectively. The fracture toughness of undoped and P-doped polysilicon is 1.2±0.2 MPa√m, and the tensile strength of polycrystalline 3C SiC is 3.2±1.2 GPa. These results include the first report of the mechanical strength of micromachined polycrystalline 3C SiC.

In an effort to develop thick, p-type polycrystalline silicon (polysilicon) films for microelectromechanical systems (MEMS) applications, in-situ boron-doped polysilicon films were deposited by a single-step APCVD process at susceptor temperatures ranging from 700°C to 955°C. The process produces boron-doped films at a deposition rate of 73 nm/min at 955°C. Spreading resistance measurements show that the boron doping level is constant at 2 × 1019 /cm3 throughout the thickness of the films. Doped films deposited at the low temperatures exhibit compressive stress as high as 666 Mpa; however films deposited at 955°C exhibited stress as low as 130 MPa. TEM and XRD show that the microstructure strongly depends on the deposition conditions. Surface micromachined, singly clamped cantilevers and strain gauges were successfully fabricated and used to characterize the residual stress of 5.0 µm-thick doped films deposited at a susceptor temperature of 955°C.

LPCVD polysilicon thin films deposited between ∼550 and ∼600 °C have an equiaxed microstructure (resulting from crystallization of an initially amorphous deposit) and contain ∼200 MPa residual tensile stresses after deposition, whereas polysilicon films deposited above ∼600 °C have a columnar microstructure and contain ∼300 MPa residual compressive stresses after deposition. Both types of films also contain stress gradients.

We have grown films containing multiple layers of polysilicon ("MultiPoly") by cycling the growth temperature between 570 and 615 °C. The multilayer films thus formed are comprised of alternating tensile and compressive layers, and by control of the thickness of the individual layers, the overall stress of the polysilicon can display any value between that of the tensile layer and that of the compressive layer. We have focussed on producing films with zero overall residual stress, as measured by wafer curvature, and have characterized the microstructures by transmission electron microscopy (TEM) and X-ray diffraction (XRD).

Because of the stress gradients present in both layers, devices made from films with zero overall residual stress may show distortions after release. We have measured the stress gradients in each type of layer and can design films with zero overall stress and zero overall stress gradients. It is also possible to design films with any level of overall (tensile or compressive) stress but with controlled displacements upon release.

Time dependent sub-critical cracking associated with environmental species such as moisture may have significant implications for the reliability of MEMS devices made of silicon. However, the existence of such stress corrosion phenomena in silicon remains controversial. Sub-critical crack-growth behavior in brittle materials is commonly characterized using crack velocity versus applied stress intensity curves (v-K curves). Crack velocity is inferred by curve-fitting crack length versus time data taken at low sample rates (<100 Hz) under the assumption that crack growth is a continuous process. However, we have observed discrete crack growth behavior in a micro-machined compression-loaded double cantilever beam. The samples are fabricated from (100) single crystal silicon wafers. A thin film resistor sputtered onto the sample surface using a lithographic technique is used to directly measure crack extension. The crack growth in all samples is characterized by periods of rapid crack growth interspersed with long periods of arrest in which no evidence of sub-critical cracking was observed. High speed data acquisition (up to 100 MHz) was performed and crack velocities as high as 1.7 km/s were accurately measured during these rapid growth periods.

Low residual stress silicon oxynitride thin films are investigated for use as a replacement for silicon dioxide (SiO2) as sacrificial layer in surface micromachined microelectrical-mechanical systems (MEMS). It is observed that the level of residual stress in oxynitrides is a function of the nitrogen content in the film. MEMS film stacks are prepared using both SiO2 and oxynitride sacrificial layers. Wafer bow measurements indicate that wafers processed with oxynitride release layers are significantly flatter. Polycrystalline Si (poly-Si) cantilevers fabricated under the same conditions are observed to be flatter when processed with oxynitride rather than SiO2 sacrificial layers. These results are attributed to the lower post-processing residual stress of oxynitride compared to SiO2.

The field of “non-silicon MEMS” is today rapidly expanding. The reasons are numerous: silicon has a limited functionality in many MEMS applications; the toolbox for micromaching of non-silicon materials is steadily growing in size and versatility; the price level of a variety of other advanced materials for MEMS has become more competitive etc. This development is here illustrated by a number of examples from a wide variety of materials classes, methods for synthesis and microprocessing, and fields of application. Among these examples are high-aspect ratio processing of different materials by ion track technology (MITE) and deep ion projection lithography (DIPL); laser microprocessing and free-space 3D laser writing; high-precision 3D diamond replicas from microstructured silicon masters; and microreplication in polymeric materials for microoptics and microsystems technology.

Silicon is currently the most commonly used material for the fabrication of microelectromechanical systems (MEMS). However, silicon-based MEMS will not be suitable for long-endurance devices involving components rotating at high speed, where friction and wear need to be minimized, components such as 2-D cantilevers that may be subjected to very large flexural displacements, where stiction is a problem, or components that will be exposed to corrosive environments. The mechanical, thermal, chemical, and tribological properties of diamond make it an ideal material for the fabrication of long-endurance MEMS components. Cost-effective fabrication of these components could in principle be achieved by coating Si with diamond films and using conventional lithographic patterning methods in conjunction with e. g. sacrificial Ti or SiO2 layers. However, diamond coatings grown by conventional chemical vapor deposition (CVD) methods exhibit a coarse-grained structure that prevents high-resolution patterning, or a fine-grained microstructure with a significant amount of intergranular non-diamond carbon. We demonstrate here the fabrication of 2-D and 3-D phase-pure ultrananocrystalline diamond (UNCD) MEMS components by coating Si with UNCD films, coupled with lithographic patterning methods involving sacrificial release layers. UNCD films are grown by microwave plasma CVD using C60-Ar or CH4-Ar gas mixtures, which result in films that have 3-5 nm grain size, are 10-20 times smoother than conventionally grown diamond films, are extremely resistant to corrosive environments, and are predicted to have a brittle fracture strength similar to that of single crystal diamond.

Ultra-hard film materials such as chemical vapor deposited (CVD) diamond are uniquely qualified for applications where superior tribological and electronic properties are required. Patterning of the film materials is essential to produce functional micro devices. Conventional lithography-based chemical etching is difficult or impossible on the ultra-hard diamond film materials, which have a high chemical resistance. Investigations on maskless patterning of the CVD diamond film are presented. Focused excimer laser pulses are used for dry etching on the film materials, and a micro computer numerical control (micro-CNC) stage is used for patterned translation of a target. The laser ablation of CVD diamond is observed to set up relationships among the processing parameters, such as the gas processing environments, the laser energy fluence and the number of laser pulses. The extent of the ablation-induced plasma is observed by time integrated image capturing. A cell-patterned structure, fabricated by the innovative maskless process, is presented for discussion.

Ultrashort laser pulses are utilized for fabrication of three-dimensional (3D) photonic crystals based on a multiphoton absorption process. The basic idea is, when a femtosecond laser pulse is tightly focused into some transparent media, a submicrometer hole will be generated due to microexplosion. By arraying these holes the same way as atoms in general solid crystals, 3D photonic lattices are achieved. Pronounced photonic bandgap effect shows that this technique is promising for tailoring arbitrary-lattice photonic crystals.

SU-8 has become a popular material for micromachining high aspect ratio structures. Typically, SU-8 is spun on a polished silicon wafer for processing. After patterning, the SU-8 is used for micromachined structures directly (such as fluidic channels) or as a mold for electroforming. Non-silicon substrates offer the possibility of cheaper processing, improved mold designs, and multi-material devices. Successful SU-8 processing depends strongly on surface properties of the substrate itself as well as environmental conditions during the processing. We explore the issues involved in transferring SU-8 technology to non-silicon substrates such as glass, plastics and metals. Issues such as wettability, adhesion, and surface tension are explored in this study. The findings indicate the merits of non-spinning approaches, such as dipping, spraying, and brushing and point to new SU-8 processes.

Metallization of plastic devices is of interest to newer non-silicon based MEMS which are being developed for biomedical and other applications. The electrical properties of the metallization will depend, in part, on the material on which they are patterned, and the method of deposition. Of interest are the electrical characterizations during usage which must be studied to allow plastic MEMS designs to proceed wisely. Flexible microdevices may develop open circuits or shorts after excessive bending; joule heating may adversely affect the plastic substrates, leading to failure. We present some experimental studies on electrical characterization of metallized plastics as it relates to various processing conditons and usage scenarios.

Stiction failures in microelectromechanical systems (MEMS) occur when suspended elastic members are unexpectedly pinned to their substrates. This type of device failure develops both in fabrication and during device operation, being a dominant source of yield loss in MEMS. Stiction failures require first a collapse force that brings the elastic member contact with the substrate followed by an intersolid adhesion sufficiently large to overcome the elastic restoring force. Stiction failure mechanisms have been studied extensively elsewhere [1]. This paper briefly summarizes these mechanisms in a the practical way. Over the last decade, stiction failure rates in MEMS have been minimized using a wide variety of processing, surface treatment, and physical schemes. An update of these methods is provided.

A test system has been assembled to examine factors controlling the reliability of microdevice switching. The system consists of a spherically tipped sputter-coated post driven by a piezoresistive actuator into a sputter-coated micromechanical membrane. Contact forces are calculated based upon the measured deflection of the micromechanical membrane, which is induced by post actuation. Membrane deflection is measured using white light interferometry. The known geometry of the membrane and probe tip permits the estimation of contact area and, hence, contact pressure. Preliminary testing of sputtered gold contacts resulted in localized melting, material transport to the membrane and failure of contact surfaces at relatively low contact pressures and current densities. Further testing will examine the factors governing the reliability of microdevice contacts, such as material properties, contact pressure and current density.

Microelectromechanical systems (MEMS) have been identified as a key technology for small-scale satellites, integrated sensors, and intelligent control systems. Using methods developed for highly integrated electronics, mechanical components are co-fabricated on planar wafers and subsequently etched free for mechanical movements in three dimensions. A major design limitation for these systems is their inability to withstand prolonged sliding surface contact. The fundamental problem is that the surface properties of silicon and poly-silicon, two of the most widely used materials for MEMS, are highly unsuitable for moving MEMS devices, resulting in high wear during operation. This work explores the feasibility and benefits of depositing thin, wear-resistant, low-friction coatings on silicon or poly-silicon. To achieve this goal, three-dimensional test silicon microstructures have been fabricated. Wear-resistant titanium carbide (TiC) coatings are deposited on these test structures using a novel non-line-of-sight pulsed laser deposition (PLD) process. In parallel, this paper addresses the integration of the TiC coating directly into the MEMS fabrication processes and its compatibility with standard silicon processing.

A MEMS test structure capable of measuring friction between polysilicon surfaces under a variety of test conditions has been refined from previous designs. The device is applied here to measuring friction coefficients of polysilicon surfaces under different environmental, loading, and surface conditions. Two methods for qualitatively comparing friction coefficients (µ) using the device are presented. Samples that have been coated with a self-assembled monolayer of the lubricating film perfluorinated-decyltrichlorosilane (PFTS) have a coefficient of friction that is approximately one-half that of samples dried using super-critical CO2 (SCCO2) drying. Qualitative results indicate that µ is independent of normal pressure. Wear is shown to increase µ for both supercritically dried samples and PFTS coated samples, though the mechanisms appear to be different. Super critically dried surfaces appear to degrade continuously with increased wear cycles, while PFTS coated samples reach a steady state friction value after about 105 cycles.

Two major problems associated with Si-based MEMS (MicroElectroMechanical Systems) devices are stiction and wear. Surface modifications are needed to reduce both adhesion and friction in micromechanical structures to solve these problems. In this paper, we will present a CVD (Chemical Vapor Deposition) process that selectively coats MEMS devices with tungsten and significantly enhances device durability. Tungsten CVD is used in the integrated-circuit industry, which makes this approach manufacturable. This selective deposition process results in a very conformal coating and can potentially address both stiction and wear problems confronting MEMS processing. The selective deposition of tungsten is accomplished through the silicon reduction of WF6. The self-limiting nature of this selective W deposition process ensures the consistency necessary for process control. The tungsten is deposited after the removal of the sacrificial oxides to minimize stress and process integration problems. Tungsten coating adheres well and is hard and conducting, requirements for device performance. Furthermore, since the deposited tungsten infiltrates under adhered silicon parts and the volume of W deposited is less than the amount of Si consumed, it appears to be possible to release stuck parts that are contacted over small areas such as dimples. The wear resistance of selectively coated W parts has been shown to be significantly improved on microengine test structures.

Fluorocarbon films are useful as antistiction films for suspended structure and also for electrical isolation purposes. Furthermore, due to their low dielectric constant and ease of deposition they are useful for VLSI manufacturing applications. Thus, in the interest of building a complete database for these plasma generated fluorocarbon films, we report a designed experiment with a 4-variable matrix to fully characterize deposited films using C4F8 as the feed gas in a high density inductively coupled plasma tool. We also demonstrate the in-situ microfabrication of electrostatic actuators that exhibit the corresponding passivation film for electrical isolation. The utilization of these films as a masking material for MEMS applications and in triple nested mask arrangements is also demonstrated.