Scratch drive actuator with mechanical links for self-assembly ofthree-dimensional MEMS

The self-assembling of three-dimensional (3-D) MEMS from polysilicon surface micromachined part is very attractive. To avoid risky external manipulation, the practical use of integrated actuator to perform the assembling task is required. To that goal, this paper presents detailed characteristics of the electrostatic surface micromachined scratch drive actuator (SDA). First, from numerous SDA tests, it is shown that this actuator is able to produce a threshold force of 30 μN, with a yield above 60%. With polysilicon devices consisting of SDA mechanically linked to buckling beam, a horizontal force of 63 mN has been demonstrated with ±112 V pulse, and up to 100 μN can be obtained with higher voltage. With buckling beams, displacements up to 150 μm have been obtained in the vertical direction. The generation of vertical force of 10 μN was confirmed with a 100 μm displacement producing 1 nJ work in the vertical direction. Finally, SDA overcomes the usual sticking of surface machined polysilicon by producing enough vertical force to completely release wide polysilicon plate (500 μm×50 μm) without external manipulation. The above characteristic, both in terms of structure releasing and vertical/horizontal forces and displacements provides the SDA with the capability of self-assembling complex 3-D polysilicon part, opening new integration capabilities and new application field of MEMS     

Simulation and fabrication of piezoresistive membrane type MEMS strain sensors

Two piezoresistive (n-polysilicon) strain sensors on a thin Si₃N₄/SiO₂ membrane with improved sensitivity were successfully fabricated by using MEMS technology. The primary difference between the two designs was the number of strips of the polysilicon patterns. For each design, a doped n-polysilicon sensing element was patterned over a thin 3 μm Si₃N₄/SiO₂ membrane. A 1000×1000 μm² window in the silicon wafer was etched to free the thin membrane from the silicon wafer. The intent of this design was to fabricate a flexible MEMS strain sensor similar in function to a commercial metal foil strain gage. A finite element model of this geometry indicates that strains in the membrane will be higher than strains in the surrounding silicon. The values of nominal resistance of the single strip sensor and the multi-strip sensor were 4.6 and 8.6 kΩ, respectively. To evaluate thermal stability and sensing characteristics, the temperature coefficient of resistance [TCR=(ΔR/R₀)/ΔT] and the gage factor [GF=(ΔR/R₀)/] for each design were evaluated. The sensors were heated on a hot plate to measure the TCR. The sensors were embedded in a vinyl ester epoxy plate to determine the sensor sensitivity. The TCR was 7.5×10⁻⁴ and 9.5×10⁻⁴/°C for the single strip and the multi-strip pattern sensors. The gage factor was as high as 15 (bending) and 13 (tension) for the single strip sensor, and 4 (bending) and 21 (tension) for the multi-strip sensor. The sensitivity of these MEMS sensors is much higher than the sensitivity of commercial metal foil strain gages and strain gage alloys.     

Etch rates for micromachining processing

The etch rates for 317 combinations of 16 materials (single-crystal silicon, doped, and undoped polysilicon, several types of silicon dioxide, stoichiometric and silicon-rich silicon nitride, aluminum, tungsten, titanium, Ti/W alloy, and two brands of positive photoresist) used in the fabrication of microelectromechanical systems and integrated circuits in 28 wet, plasma, and plasmaless-gas-phase etches (several HF solutions, H₃PO₄, HNO₃ +H₂O+NH₄F, KOH, Type A aluminum etchant, H ₂O+H₂O₂+HF, H₂O₂, piranha, acetone, HF vapor, XeF₂, and various combinations of SF₆, CF₄, CHF₃, Cl₂, O₂ , N₂, and He in plasmas) were measured and are tabulated. Etch preparation, use, and chemical reactions (from the technical literature) are given. Sample preparation and MEMS applications are described for the materials     

Thermal conductivity of doped polysilicon layers

The thermal conductivities of doped polysilicon layers depend on grain size and on the concentration and type of dopant atoms. Previous studies showed that layer processing conditions strongly influence the thermal conductivity, but the effects of grain size and dopant concentration were not investigated in detail. The current study provides thermal conductivity measurements for low-pressure chemical-vapor deposition (LPCVD) polysilicon layers of thickness near 1 μm doped with boron and phosphorus at concentrations between 2.0×10¹⁸ cm^-3 and 4.1×10¹⁹ cm^-3 for temperatures from 20 K to 320 K. The data show strongly reduced thermal conductivity values at all temperatures compared to similarly doped single-crystal silicon layers, which indicates that grain boundary scattering dominates the thermal resistance. A thermal conductivity model based on the Boltzmann transport equation reveals that phonon transmission through the grains is high, which accounts for the large phonon mean free paths at low temperatures. Algebraic expressions relating thermal conductivity to grain size and dopant concentration are provided for room temperature. The present results are important for the design of MEMS devices in which heat transfer in polysilicon is important     

Microsized Piston-Cylinder Pneumatic and Hydraulic Actuators Fabricated by Lithography

Future microrobotic applications require actuators that can generate a high actuation force in a limited volume. Up to now, little research has been performed on the development of pneumatic or hydraulic microactuators, although they offer great prospects in achieving high force densities. In addition, large actuation strokes and high actuation speeds can be achieved by these actuators. This paper describes a fabrication process for piston-cylinder pneumatic and hydraulic actuators based on etching techniques, UV-definable polymers, and low-temperature bonding. Prototype actuators with a piston area of 0.15 mm² have been fabricated in order to validate the production process. These actuators achieve actuation forces of more than 0.1 N and strokes of 750 mum using pressurized air or water as driving fluid.     

Asymmetric surface roughness measurements and meniscus modeling of polysilicon surface micromachined flaps

Polycrystalline silicon (polysilicon) films are primary structural materials for microelectromechanical systems (MEMS). Due to relatively high compliance, large surface-to-volume ratio, and small separation distances, micromachined polysilicon structures are susceptible to surface forces which can result in adhesive failures. Since these forces depend on surface properties especially surface roughness, three types of microhinged flaps were fabricated to characterize their roughness and adhesive meniscus properties. The flaps enabled access to both the top and bottom surfaces of the structural polysilicon layers. Roughness measurements using an atomic force microscope revealed that MEMS surfaces primarily exhibit non-Gaussian surface height distributions, and for the release procedures studied, the bottom surface of the structural layers was significantly smoother and prone to higher adhesion compared to the top surface. A non-symmetric surface roughness model using the Pearson system of frequency curves was coupled with a capillary meniscus adhesion model to analyze the effects of surface roughness parameters (root-mean-square, skewness, and kurtosis), relative humidity, and surface contact angle on the interfacial adhesion energy. Using the measured roughness properties of the flaps, four different surface pairs were simulated and compared to investigate their effects on capillary adhesion. It was found that since the base polysilicon layer (poly0) was rougher than the base silicon nitride and the structural layer on poly0 was also rougher than that on silicon nitride, depositing MEMS devices on poly0 layer rather than directly on silicon nitride will reduce the adhesion energy.     

Surface-micromachined capacitive differential pressure sensor withlithographically defined silicon diaphragm

A capacitive surface-micromachined sensor suitable for the measurement of liquid and gas pressures was fabricated. The structure consists of a polysilicon stationary electrode suspended 0.7 μm above a 20-μm-thick lightly doped silicon diaphragm formed by a patterned etch stop. The a priori patterning of the buried etch stop yields diaphragm widths independent of wafer thickness variations with excellent alignment. The design described here has a pressure range of 100 PSI, a nominal capacitance of 3.5 pF with a full scale span of 0.8 pF, and a temperature coefficient of 100 ppm°C^-1. Each device, including a matched reference capacitor, occupies 2.9 mm² , yielding approximately 2000 devices per 100-mm wafer     

Process-dependent thin-film thermal conductivities for thermal CMOSMEMS

The thermal conductivities κ of the dielectric and conducting thin films of three commercial CMOS processes were determined in the temperature range from 120 to 400 K. The measurements were performed using micromachined heatable test structures containing the layers to be characterized. The κ values of thermally grown silicon oxides are reduced from bulk fused silica by roughly 20%. The κ of phosphosilicate and borophosphosilicate glasses are 0.94±0.08 W m^-1 K^-1 and 1.18±0.06 W m^-1 K^-1, respectively, at 300 K. A plasma-enhanced chemical-vapor-deposition silicon-nitride layer has a thermal conductivity of 2.23±0.12 W m^-1 K^-1 at 300 K. This value is between published data for atmospheric-pressure CVD and low-pressure CVD nitrides. For the metal layers, we found thermal conductivities between 167 W m^-1 K^-1 and 206 W m ^-1 K^-1, respectively, at 300 K, to be compared with 238 W m^-1 K^-1 of bulk aluminum. The temperature-dependent product κρ of κ with the electrical resistivity ρ agrees better than 8.2% between 180-400 K with that of pure bulk aluminum. The κ values of the polysilicon layers are between 22.4 W m^-1 K^-1 and 37.3 W m^-1 K^-1 at 300 K. They are reduced from similarly doped bulk silicon by factors of between 2.0-1.3. The observed discrepancies between thin film and bulk data demonstrate the importance of determining the process-dependent thermal conductivities of CMOS thin films     

A high-Tc superconductor bolometer on a silicon nitridemembrane

In this paper, we describe the design, fabrication, and performance of a high-T_c GdBa₂Cu₃O_7-δ superconductor bolometer positioned on a 2× 2-mm² 1-μm-thick silicon nitride membrane. The bolometer structure has an effective area of 0.64 mm² and was grown on a specially developed silicon-on-nitride (SON) layer. This layer was made by direct bonding of silicon nitride to silicon after chemical mechanical polishing. The operation temperature of the bolometer is 85 K. A thermal conductance G=3.3·10^-5 W/K with a time constant of 27 ms has been achieved. The electrical noise equivalent power (NEP) at 5 Hz is 3.7·10^-2 WHz^-1/2, which is very close to the theoretical phonon noise limit of 3.6·10^-12 WHz ^-1/2, meaning that the excess noise of the superconducting film is very low. This bolometer is comparable to other bolometers with respect to high electrical performance. Our investigations are now aimed at decreasing the NEP for 84-μm radiation by further reduction of G and adding an absorption layer to the detector. This bolometer is intended to be used as a detector in a Fabry-Perot (FP)-based satellite instrument designed for remote sensing of atmospheric hydroxyl     

A new silicon gas-flow sensor based on lift force

This paper presents the first silicon-flow sensor based on lift force. The sensor is a bulk-micromachined airfoil structure that uses the lift force as a sensing principle. The lift force acts normal to the flow in contrast to drag-force sensor types, where the force acts in the flow direction. The sensor utilizes the special distribution of the lift force along the length of the sensor structure. Since the sensor, like an airfoil, is mounted at a small angle to the flow, it induces very little flow disturbance. The sensor consists of two plates connected to a center beam. Each plate is 5×5-mm square with a thickness of 30 μm. The flow-induced forces deflect the two plates in the same direction, but with different magnitude. The deflections are detected by polysilicon strain gauges. The differential mode bridge makes the sensor insensitive to common mode deflection, e.g., acceleration forces. The lift-force principle is characterized using fundamental airfoil theory. The sensor has been experimentally verified, and a flow sensitivity of 7.4 μV/V/(m/s)² has been measured in both flow directions     

A HARPSS polysilicon vibrating ring gyroscope

This paper presents the design, fabrication, and testing of an 80-μm-thick, 1.1 mm in diameter high aspect-ratio (20:1) polysilicon ring gyroscope (PRG). The vibrating ring gyroscope was fabricated through the high aspect-ratio combined poly and single-crystal silicon MEMS technology (HARPSS). This all-silicon single-wafer technology is capable of producing electrically isolated vertical electrodes as tall as the main body structure (50 to 100's (μm tall)) with various size air-gaps ranging from submicron to tens of microns. A detailed analysis has been performed to determine the overall sensitivity of the vibrating ring gyroscope and identify its scaling limits. An open-loop sensitivity of 200 μV/deg/s in a dynamic range of ±250 deg/s was measured under low vacuum level for a prototype device tested in hybrid format. The resolution for a PRG with a quality factor (Q) of 1200, drive amplitude of 0.15 μm, and sense node parasitic capacitances of 2 pF was measured to be less than 1 deg/s in 1 Hz bandwidth, limited by the noise from the circuitry. Elimination of the parasitic capacitances and improvement in the quality factor of the ring structure are expected to reduce the resolution to 0.01 deg/s/(Hz)^0.5     

Subcritical crack growth in silicon MEMS

New experimental techniques need to be developed to address fundamental materials issues in MEMS. Experimental protocols developed for macroscale testing are not necessarily applicable, and an understanding of the behavior of macroscale specimens cannot necessarily be relied upon to predict the behavior of microscale MEMS structures. An experimental protocol for studying slow crack growth in MEMS materials has been developed, and this protocol has been used to show that polycrystalline silicon (polysilicon) MEMS are susceptible to stress corrosion cracking. Using a model of the nonlinear dynamics of a specimen allowed an estimation of crack length and crack closure from the frequency response of the specimen. The procedure can resolve 1-nm crack extensions and crack growth rates below 10^-13 m/s. Crack closure, which has a pronounced effect on the dynamics of this nonlinear system, may be associated with the native oxide that grows on the faces of the crack. The data show that subcritical crack growth in polysilicon MEMS is driven by the synergistic effects of water and stress. In contrast to macroscale stress corrosion cracking behavior, a clear relationship between crack growth rate, stress intensity and humidity has not been found. Micrographs suggest that the crack path is transgranular     

Feasibility of Inductive Communication Between Millimeter-Sized Wireless Robots

This paper shows the feasibility of inductive data transmission between millimeter-sized (2 times 2 times 1 mm³) swarm robots. To fulfill the size restraints, the coils used for inductive transmission were fabricated by micromachining technologies. Since the induced signal at the receiving coil was very small due to the low coupling coefficient, a high amplification gain of the signal amplifier was necessary. Transmitter and receiver electronics for data transmission were constructed and the quality in data transmission between two 2 times 2 mm²-sized micromachined coils was evaluated. With a continuous power consumption of 270 muW in the transmitter, data transmission was successful without communication error up to a distance of 7 mm, which was considered sufficient to enable 2 times 2 mm² robots to behave as a swarm within the given operation area.     

A new organic modifier for anti-stiction

The chemical and mechanical characteristics of a new surface modifier, dialkyldichloromethylsilane (DDMS, CH₃)₂SiCl₂, for stiction-free polysilicon surfaces are reported. The main strategy is to replace the conventional monoalkyl-trichlorosilane (MTS, RSiCl₃) such as octadecyltrichlorosilane (ODTS) or 1H,1H,2H,2H-perfluorodecyltrichlorosilane (FDTS) with dichlorodisilane (DDS, R₂SiCl₂) with two short chains, especially DDMS. DDMS, with shorter chains in aprotic media, rapidly deposits on the chemically oxidized polysilicon surface at room temperature and successfully prevents long cantilevers 3 mm in length from in-use as well as release stiction. DDMS-modified polysilicon surfaces exhibit satisfactory hydrophobicity, long term stability and thermal stability, which are comparable to those of FDTS. DDMS as an alternative to FDTS and ODTS provides a few valuable advantages; ease in handling and storage of the solution, low temperature-dependence and low cost. In addition to the new modifier molecule, the simplified process of direct release right after washing the modified surface with isooctane was proposed to cut the processing time     

Development of Microfabricated Cylindrical Ion Trap Mass Spectrometer Arrays

A novel approach, in which microelectromechanical systems (MEMS) technology is used for constructing miniature cylindrical ion trap (CIT) mass spectrometer (MS) arrays in silicon (Si), is described. MEMS processes were used to fabricate precise CIT geometries in a stack of Si, SiO₂, and Si₃N₄. These geometries were then selectively coated with conductive (Cr/Au) layers to obtain a functional CIT array with individual CIT radii (r₀) of 360 mum, half-thickness (z₀) of 351 mum, and aperture size (r_H) of 162 mum. Each trap of a 5 times 5 CIT array was operated in the mass selective instability mode to analyze trichloroethylene and perfluorotributylamine at a pressure of 10^-5 torr. Mass spectra from individual CITs in the array were obtained using a rasterable electron beam for internal ionization. Investigation of the operation of individual CITs in the array is a critical step toward the understanding of the overall functioning of MS arrays.     

Bent-beam electrothermal actuators-Part II: Linear and rotarymicroengines

For Part I see L. Que, J.S. Park and Y.B. Gianchandani, ibid., vol.10, pp.247-54 (2001). This paper reports on the use of bent-beam electrothermal actuators for the purpose of generating rotary and long-throw rectilinear displacements. The rotary displacements are achieved by orthogonally arranged pairs of cascaded actuators that are used to rotate a gear. Devices were fabricated using electroplated Ni, p ⁺⁺ Si, and polysilicon as structural materials. Displacements of 20-30 μm with loading forces >150 μN at actuation voltages <12 V and power dissipation <300 mW could be achieved in the orthogonally arranged actuator pairs. A design that occupies <1 mm ² area is presented. Long-throw rectilinear displacements were achieved by inchworm mechanisms in which pairs of opposing actuators grip and shift a central shank that is cantilevered on a flexible suspension. A passive lock holds the displaced shank between pushes and when the power is off. This arrangement permits large output forces to be developed at large displacements, and requires zero standby power. Several designs were fabricated using electroplated Ni as the structural material. Forces >200 μN at displacements >100 μm were measured     

Lateral MEMS microcontact considerations

A lateral switching relay structure has been developed which provides a double gold contact with as low as 70-mΩ measured contact resistance, 0.45-A current-carrying ability at MEMS compatible force levels, TTL compatible actuation, and air gap isolation when open. The die area used for the relay mechanism itself (distinct from the actuation) is approximately 75 μm by 100 μm and was designed to allow fabrication of the relays in the MCNC MUMP's dual polysilicon foundry process with no assembly. Design analysis shows that substantial characterization is needed to design optimal microrelays. Temperature softening and failure modes have been characterized by current voltage techniques. Polysilicon vernier structures were used to develop force/current/conductance curves. Relays using thermal actuators have been built     

Computing LOGCFL certificates

The complexity class LOGCFL consists of all languages (or decision problems) which are logspace reducible to a context-free language. Since LOGCFL is included in AC¹, the problems in LOGCFL are highly parallelizable.

By results of Ruzzo (JCSS 21 (1980) 218), the complexity class LOGCFL can be characterized as the class of languages accepted by alternating Turing machines (ATMs) which use logarithmic space and have polynomially sized accepting computation trees. We show that for each such ATM M recognizing a language A in LOGCFL, it is possible to construct an L^LOGCFL transducer T_M such that T_M on input w A outputs an accepting tree for M on w. It follows that computing single LOGCFL certificates is feasible in functional AC¹ and is thus highly parallelizable.

Wanke (J. Algorithms 16 (1994) 470) has recently shown that for any fixed k, deciding whether the treewidth of a graph is at most k is in the complexity-class LOGCFL. As an application of our general result, we show that the task of computing a tree-decomposition for a graph of constant treewidth is in functional LOGCFL, and thus in AC¹.

We also show that the following tasks are all highly parallelizable: Computing a solution to an acyclic constraint satisfaction problem; computing an m-coloring for a graph of bounded treewidth; computing the chromatic number and minimal colorings for graphs of bounded tree- width.     

Fabrication of improved piezoresistive silicon cantilever probes for the atomic force microscope

This paper describes an improved design for a monolithic silicon atomic force microscope (AFM) probe using piezoresistive sensing. The probe is V shaped, with a sharp tip at the free end and two piezoresistors at the root, and is fabricated using silicon-on-insulator (SOI) starting material. The maximum sensitivity of the AFM probe is measured to be 4.0(± 0.1) × 10⁻⁷ Å⁻¹, which is larger than that of the previous parallel-arm piezoresistive AFM probe. The measured results are in reasonable agreement with the values predicted by theory. The minimum detectable force and minimum detectable deflection of the AFM probes are predicted to be 1.0 × 10⁻¹⁰ N and 0.29 Å_r.m.s., respectively, using a Wheatstone bridge arrangement biased at a voltage of ± 5 V and bandwidth of 10 Hz–1 kHz.     

A miniaturized high-voltage solar cell array as an electrostaticMEMS power supply

A hydrogenated amorphous silicon (a-Si:H) solar cell array that is designed as an on-board power source for electrostatic microelectromechanical systems (MEMS) is presented. A single cell consists of a triple layer of p-i-n/p-i-n/p-i-n a-Si:H and produces an open circuit voltage (V_OC) of 1.8~2.3 V, a short circuit current density (J_SC) of 2.8 mA/cm², and fill factor (FF) of 0.495. A series interconnected array of 100 single solar cells (total array area of 1 cm²) is fabricated in an integrated fashion and produces an array V_OC of 150 V, and array short circuit current (I_SC) of 2.8 μA under Air Mass (AM) 1.5 illumination. To demonstrate the usefulness of this solar cell array as an on-board power source for electrostatically driven micromachined devices, it has been packaged with a movable micromachined silicon (Si) mirror in a hybrid manner. The movable Si mirror is directly driven by the cell array electrical output, and the motion of the mirror plate has been observed reproducibly. Variation of light intensity and/or number of illuminated cells produces different values of array V_OC, thus enabling control of the deflection of the Si mirror by variation of incident light intensity