203 Wendy C. Crone The expanding and developing fields of micro-electromechanical systems (MEMS) and nano-electromechanical (NEMS) are highly interdisciplinary and rely heavily on experimental mechanics for materials selection, process validation, design development, and device characterization. These devices range from mechanical sensors and actuators, to microanalysis and chemical sensors, to micro-optical systems and bioMEMS for microscopic surgery. Their applications span the automotive industry, communications, defense systems, national security, health care, information technology, avionics, and environmental monitoring. This chapter gives a general introduction to the fabrication processes and materials commonly used in MEMS/NEMS, as well as a discussion of the application of experimental mechanics techniques to these devices. Mechanics issues that arise in selected example devices are also presented. 9.1 Background ......................................... 203 9.2 MEMS/NEMS Fabrication ......................... 206 9.3 Common MEMS/NEMS Materials and Their Properties ............................. 206 9.3.1 Silicon-Based Materials ................ 207 9.3.2 Other Hard Materials .................... 208 9.3.3 Metals ........................................ 9.3.4 Polymeric Materials ...................... 9.3.5 Active Materials ........................... 9.3.6 Nanomaterials ............................. 9.3.7 Micromachining ........................... 9.3.8 Hard Fabrication Techniques ......... 9.3.9 Deposition .................................. 9.3.10 Lithography ................................. 9.3.11 Etching ....................................... 208 208 209 209 210 211 211 211 212 9.4 Bulk Micromachining versus Surface Micromachining .............. 213 9.5 Wafer Bonding ..................................... 214 9.6 Soft Fabrication Techniques................... 215 9.6.1 Other NEMS Fabrication Strategies .. 215 9.6.2 Packaging ................................... 216 9.7 Experimental Mechanics Applied to MEMS/NEMS .......................... 217 9.8 The Influence of Scale ........................... 9.8.1 Basic Device Characterization Techniques .................................. 9.8.2 Residual Stresses in Films .............. 9.8.3 Wafer Bond Integrity .................... 9.8.4 Adhesion and Friction................... 217 218 219 220 220 9.9 Mechanics Issues in MEMS/NEMS ............. 221 9.9.1 Devices ....................................... 221 9.10 Conclusion ........................................... 224 References .................................................. 225 9.1 Background The acronym MEMS stands for micro-electromechanical system, but MEMS generally refers to microscale devices or miniature embedded systems involving one or more micromachined component that enables higher-level functionality. Similarly NEMS, nanoelectromechanical system, refers to such nanoscale devices or nanodevices. MEMS and NEMS are fab- ricated microscale and nanoscale devices that are often made in batch processes, usually convert between some physical parameter and a signal, and may be incorporated with integrated circuit technology. The field of MEMS/NEMS encompasses devices created with micromachining technologies originally developed to produce integrated circuits, as well as Part A 9 A Brief Introd 9. A Brief Introduction to MEMS and NEMS 204 Part A Solid Mechanics Topics Part A 9.1 Table 9.1 Sample applications of MEMS/NEMS Sensors Actuators Passive structures Accelerometers, biochemical analyzers, environmental assay devices, gyroscopes, medical diagnostic sensors, pressure sensors Data storage, drug-delivery devices, drug synthesis, fluid regulators, ink-jet printing devices, micro fuel cells, micromirror devices, microphones, optoelectric devices, radiofrequency devices, surgical devices Atomizers, fluid spray systems, fuel injection, medical inhalers non-silicon-based devices created by the same micromachining or other techniques. They can be classified as sensors, actuators, and passive structures (Table 9.1). Sensors and actuators are transducers that convert one physical quantity to another, such as electromagnetic, mechanical, chemical, biological, optical or thermal phenomena. MEMS sensors commonly measure pressure, force, linear acceleration, rate of angular motion, torque, and flow. For instance, to sense pressure an intermediate conversion step, such as mechanical stress, can be used to produce a signal in the form of electrical energy. The sensing or actuation conversion can use a variety of methods. MEMS/NEMS sensing can employ change in electrical resistance, piezoresistive, piezoelectric, change in capacitance, and magnetoresistive methods (Table 9.2). MEMS/NEMS actuators provide the ability to manipulate physical parameters at the micro/nanoscale, and can employ electrostatic, thermal, magnetic, piezoelectric, piezoresistive, and shape-memory transformation methods. Passive MEMS structures such as micronozzles are used in atomizers, medical inhalers, fluid spray systems, fuel injection, and ink-jet printers. MEMS have a characteristic length scale between 1 mm and 1 μm, whereas NEMS devices have a characteristic length scale below 1 μm (most strictly, 1–100 nm). For instance a digital micromirror device has a characteristic length scale of 14 μm, a quantum dot transistor has components measuring 300 nm, and molecular gears fall into the 10–100 nm range [9.2]. Additionally, although an entire device may be mesoscale, if the functional components fall in the microscale or nanoscale regime it may be referred to as a MEMS or NEMS device, respectively. MEMS/NEMS inherently have a reduced size and weight for the function they carry out, but they can also carry advantages such as low power consumption, improved speed, increased function in one package, and higher precision. There is no distinct MEMS/NEMS market, instead there is a collection of niche markets where MEMS/NEMS become attractive by enabling a new function, bringing the advantage of reduced size, or lowering cost [9.1]. Despite this characteristic, the MEMS industry is already valued in tens of billions of dollars and growing rapidly. The Small Times Tech Business DirectoryTM Table 9.2 Physical quantities used in MEMS/NEMS sensors and actuators (after Maluf [9.1]) Method Description Physical and material parameters Order of energy density (J/cm3 ) Electrostatic Attractive force between two components carrying opposite charge Certain materials that change shape under an electric field Thermal expansion or difference in coefficient of thermal expansion Electric current in a component surrounded by a magnetic field gives rise to an electromagnetic force Certain materials that undergo a solid–solid phase transformation producing a large shape change Electric field, dielectric permittivity ≈ 0.1 Electric field, Young’s modulus, piezoelectric constant Coefficient of expansion, temperature change, Young’s modulus Magnetic field, magnetic permeability ≈ 0.2 Transformation temperature ≈ 10 Piezoelectric Thermal Magnetic Shape memory ≈5 ≈4 A Brief Introduction to MEMS and NEMS ples include biofluidic chips for biochemical analyses, biosensors for medical diagnostics, environmental assays for toxin identification, implantable pharmaceutical drug delivery, DNA and genetic code analysis, imaging, and surgery. NEMS is often associated with biotechnology because this size scale allows for interaction with biological systems in a fundamental way. BioNEMS may be used for drug delivery, drug synthesis, genome synthesis, nanosurgery, and artificial organs comprised of nanomaterials. The sensitivity of such bioNEMS devices can be exquisite, selectively binding and detecting a single biomolecule. More complete background information on microfluidic devices can be found in Beeby [9.7], Koch [9.9] and Kovacs [9.10]. Semiconductor NEMS devices can offer microwave resonance frequencies, exceptionally high mechanical quality factors, and extraordinarily small heat capacities [9.11, 12]. Examples of NEMS devices also include transducers, radiating energy devices, nanoscale integrated circuits, and optoelectronic devices [9.13, 14]. NEMS manufacturing is being further enabled by the drive towards nanometer feature sizes in the microelectronics industry. Terascale computational ability will require nanotransistors, nanodiodes, nanoswitces, and nanologic gates [9.15]. NEMS also opens the door for fundamental science at the nanometer scale investigating phonon-mediated mechanical processes [9.16] and quantum behavior of mesoscopic mechanical systems [9.17]. Although there is some discussion as to whether the NEMS definition requires a characteristic length scale below 1000 nm or 100 nm, there is no argument that the field of NEMS is in its infancy. Existing commercial devices are limited at this point, but research on NEMS is extremely active and highly promising. Many challenges remain, including assembly of nanoscale devices and mass production capabilities. In the long term, a number of issues must be addressed in analysis, design, development, and fabrication for high-performance MEMS/NEMS to become ubiquitous. Of most relevance to the focus of this handbook, advanced materials must be well characterized and MEMS/NEMS testing must be further developed. Additionally for commercialization, MEMS/NEMS design must consider issues of market (need for product, size of market), impact (enabling new systems, paradigm shift for the field), competition (other macro and micro/nanoproducts existing), technology (available capability and tools), and manufacturing (manufacturability in volume at low cost) [9.18]. 205 Part A 9.1 lists more than 700 manufacturers/fabricators of microsystems and nanotechnologies [9.3]. High-volume production with lucrative sales have been achieved by several companies making devices such as accelerometers for automobiles (Analog Devices, Motorola, Bosch), micromirrors for digital projection displays (Texas Instruments), and pressure sensors for the automotive and medical industries (NovaSensor). Currently, the MEMS markets with the largest commercial value are ink-jet printer heads, optical MEMS (which includes the Digital Micromirror DeviceTM discussed below), and pressure sensors, followed by microfluidics, gyroscopes, and accelerometers [9.4]. The MEMS market was reported to be US $5.1 billion in 2005 and projected to reach US $9.7 billion by 2010 [9.4]. The NEMS industry, while still young, has been growing in value. The market research firm Report Buyer recently released a market report on Nanorobotics and NEMS indicating that the global market for NEMS increased from US $29.5 million to US $34.2 million between 2004 and 2005 and projecting that the market will reach US $830.4 million by 2011 [9.5]. Microfluidic and nanofluidic devices also fall under the umbrella of MEMS/NEMS and are often classified as bioMEMS/bioNEMS devices when involving biological materials. These devices incorporate channels with at least one microscale or nanoscale dimension in which fluid flows. The small scale of these devices allow for smaller sample size, faster reactions, and higher sensitivity. Microfluidic devices commonly use both hard and soft fabrication techniques to produce channels and other fluidic structures [9.6]. The common feature of these devices is that they allow for flow of gas and/or liquid and use components such as pumps, valves, nozzles, and mixers. Commercial and defense applications include automotive controls, pneumatics, environmental testing, and medical devices. The advantages of the microscale in these applications include high spatial resolution, fast time response, small fluid volumes required for analysis, low leakage, low power consumption, low cost, appropriate compatibility of surfaces, and the potential for integrate signal processing [9.7]. At the microscale, pressure drop over a narrow channel is high and fluid flow generated by electric fields can be substantial. The micrometer and nanometer length scales are particularly relevant to biological materials because they are comparable to the size of cells, molecules, diffusions length for molecules, and electrostatic screening lengths of ionic conducting fluids [9.8]. Device exam- 9.1 Background 206 Part A Solid Mechanics Topics Part A 9.3 The field of MEMS/NEMS is highly multidisciplinary, often involving expertise from engineering, materials science, physics, chemistry, biology, and medicine. Because of the breadth of the field and the range of activities that fall under the scope of MEMS/NEMS, a comprehensive review is not possible in this chapter. After providing general background, the focus will be on mechanics and specifically experimental mechanics as it is applied to MEMS/NEMS. Mechanics is critical to the design, fabrication, and performance of MEMS/NEMS. A broad range of experimental tools has been applied to MEMS/NEMS. This chapter will provide an overview of such work. Additional information on the application of mechanics to MEMS/NEMS can be found in the proceedings of the annual symposium held by the MEMS and Nanotechnology Technical Division of the Society for Experimental Mechanics (see, for example, [9.19]). 9.2 MEMS/NEMS Fabrication Traditionally MEMS/NEMS are thought of in the context of microelectronics fabrication techniques which utilize silicon. This approach to MEMS/NEMS brings with it the momentum of the integrated circuits industry and has the advantage of ease of integration with semiconductor devices, but fabrication is expensive in both the infrastructure and equipment required and the time investment needed to create a working prototype. An alternative approach that has seen significant success, especially in its application to microfluidic devices, is the use of soft materials such as polydimethylsiloxane (PDMS). Soft MEMS/NEMS fabrication can often be conducted with bench-top techniques with no need for the clean-room facilities used in microelectronics fabrication. Additionally, polymers offer a range of properties not available in siliconbased materials such as mechanical shock tolerance, biocompatibility, and biodegradability. However, polymers can carry disadvantages for certain applications because of their viscoelastic behavior and low thermal stability. Ultimately a combination of function and economics decides the medium of choice for device construction. Whether we talk about hard or soft MEMS/NEMS, the basic approach to device construction is similar. Material is deposited onto a substrate, a lithographic step is used to produce a pattern, and material removal is conducted to create a shape. For traditional microelectronics fabrication, the substrate is often silicon, material deposition is achieved by vapor deposition or sputtering, lithography involves patterning of a chemically resistant polymer, and material is removed by a chemical etch. Alternatively, for soft MEMS/NEMS materials, fabrication often utilizes a glass or plastic substrate, material in the form of a monomer is flowed into a region, a lithographic mask allows exposure of a pattern to ultraviolet (UV) radiation triggering polymerization, and the unpolymerized monomer is removed with a flushing solution. For both hard and soft MEMS/NEMS fabrication there are a number of variations on these basic steps which allow for a wide array of structures and devices to be constructed. 9.3 Common MEMS/NEMS Materials and Their Properties Materials used in MEMS/NEMS must simultaneously satisfy a range requirements for chemical, structural, mechanical, and electrical properties. For biomedical and bioassay devices, material biocompatibility and bioresistance must also be considered. Most MEMS/NEMS devices are created on a substrate. Common substrate materials include singlecrystal silicon, single-crystal quartz, fused quartz, gallium arsenide, glass, and plastics. Devices are made with a range of methods by machining into the substrate and/or depositing additional material on top of the substrate. The additional materials may be structural, sacrificial, or active. Although traditionally MEMS in particular have relied on silicon, the materials used in MEMS/NEMS are becoming more heterogeneous. Selected properties are given in Table 9.3 for comparative purposes, but an extensive list of properties for the wide range of materials used in MEMS/NEMS cannot be included here. It should be noted, however, that the constitutive behavior of materials used in MEMS/NEMS applications can be sensitive to fabrication method, processing parame- A Brief Introduction to MEMS and NEMS 9.3 Common MEMS/NEMS Materials and Their Properties Property Si SiO2 Si3 N4 Quartz SiC Si(111) Stainless steel Al Young’s modulus (GPa) Yield strength (GPa) Poisson’s ratio Density (g/cm3 ) Coefficient of thermal expansion (10−6 /◦ C) Thermal conductivity at 300 K (W/cm · K) Melting temperature (◦ C) 160 7 0.22 2.4 2.6 73 8.4 0.17 2.3 0.55 323 14 0.25 3.1 2.8 107 9 0.16 2.65 0.55 450 21 0.14 3.2 4.2 190 7 0.22 2.3 2.3 200 3 0.3 8 16 70 0.17 0.33 2.7 24 1.57 0.014 0.19 0.0138 5 1.48 0.2 2.37 1415 1700 1800 1610 2830 1414 1500 660 ters, and thermal history due to the relative similarity between characteristic length scales and device dimensions. A good resource compiling characterization data from a number of sources is the material database at http://www.memsnet.org/material/ [9.20]. The following books, used as references for the discussion here, are valuable resources for more extensive information: Senturia [9.18], Maluf [9.1], and Beeby [9.7]. 9.3.1 Silicon-Based Materials Silicon, Polysilicon, and Amorphous Silicon Silicon-based materials are the most common materials currently used in MEMS/NEMS commercial production. MEMS/NEMS devices often exploit the mechanical properties of silicon rather than its electrical properties. Silicon can be used in a number of different forms: oriented single-crystal silicon, amorphous silicon, or polycrystal silicon (polysilicon). Single-crystal silicon, which has cubic crystal structure, exhibits anisotropic behavior which is evident in its mechanical properties such as Young’s modulus. A high-purity ingot of single-crystal silicon is grown, sawn to the desired thickness, and polished to create a wafer. Single-crystal silicon used for MEMS/NEMS are usually the standard 100 mm (4 inch diameter, 525 μm thickness) or 150 mm (6 inch diameter, 650 μm thickness) wafers. Although larger 8-inch and 12-inch wafers are available, they are not used as prevalently for MEMS fabrication. The properties of the wafer depend on both the orientation of crystal growth and the dopants added to the silicon (Fig. 9.1). Impurity doping has a significant impact on electrical properties but does not generally impact the mechanical properties if the concentration is approximately < 1020 cm−3 . Silicon is a group IV semiconductor. To create a p-type material, dopants from group III (such as boron) create mobile charge carriers that behave like positively charged species. To create an n-type material, dopants from group V (such as phosphorous, arsenic, and antimony) are used to create mobile charge carriers that behave like negatively charged electrons. Doping of the entire wafer can be accomplished during crystal growth. Counter-doping can be accomplished by adding dopants of the other type to an already doped substrate using deposition followed by ion implantation and annealing (to promote diffusion and relieve residual stresses). For instance, p-type into n-type creates a pn-junction. Amorphous and polysilicon films are usually deposited with thicknesses of < 5 μm, although it is also possible to create thick polysilicon [9.21]. The residual stress in deposited polysilicon and amorphous silicon thin films can be large, but annealing can be used to provide some relief. Polysilicon has the disadvantage of a somewhat lower strength and lower piezoresistivity than single-crystal silicon. Additionally, Young’s (100) n-type Secondary flat (111) n-type Primary flat (100) p-type Primary flat Secondary flat Primary flat (111) p-type Primary flat Secondary flat Fig. 9.1 Flats on standard commercial silicon wafers used to identify crystallographic orientation and doping (after Senturia [9.18]) Part A 9.3 Table 9.3 Properties of selected materials (after Maluf [9.1] and Beeby [9.7]) 207 208 Part A Solid Mechanics Topics Part A 9.3 modulus may vary significantly because the diameter of a single grain may comprise a large fraction of a component’s width [9.22]. Silicon, polysilicon, and amorphous silicon are also piezoresistive, meaning that the resistivity of the material changes with applied stress. The fractional change in resistivity, Δρ/ρ, is linearly dependent on the stress components parallel and perpendicular to the direction of the resistor. The proportionality constants are the piezoresistive coefficients, which are dependent on the crystallographic orientation, and the dopant type/concentration in single-crystal silicon. This property can be used to create a strain gage. Silicon Dioxide The success of silicon is heavily based on its ability to form a stable oxide which can be predictably grown at elevated temperature. Dry oxidation produces a higherquality oxide layer, but wet oxidation (in the presence of water) enhances the diffusion rate and is often used when making thicker oxides. Amorphous silicon dioxide can be used as a mask against etchants. It should be noted that these films can have large residual stresses. Silicon Nitride Silicon nitride can be deposited by chemical vapor deposition (CVD) as an amorphous film which can be used as a mask against etchants. It should be noted that these films can have large residual stresses. Silicon Carbide Silicon carbide is an attractive material because of its high hardness, good thermal properties, and resistance to harsh environments. Additionally, silicon carbide is piezoresistive. Although it can be produced as a bulk polycrystalline material it is generally grown or deposited on a silicon substrate by epitaxial growth (single crystal) or by chemical vapor deposition (polycrystal). Quartz Single-crystal quartz, which has a hexagonal crystal structure, can be used in natural or synthesized form. Like silicon, it can be etched selectively but the results are less ideal than in silicon because of unwanted facets and poor edge definition. Single-crystal quartz can be used as substrate material in a range of cuts which have different temperature sensitivities for piezoelectric or mechanical properties. Detailed information about quartz cuts can be found in Ikeda [9.25]. Quartz is also piezoelectric, meaning that there is a relationship between strain and voltage in the material. Fused quartz (silica) is a glassy noncrystalline material that is also occasionally used in MEMS/NEMS devices. Glass Glasses such as phosphosilicate and borosilicate (Pyrex) can be used as a substrate or in conjunction with silicon and other materials using wafer bonding (discussed below). Diamond Diamond is also attractive because of its high hardness, high fracture strength, low thermal expansion, low heat capacity, and resistance to harsh environments. Diamond is also piezoresistive and can be doped to produce semiconducting and metal-like behavior [9.26]. Because of its hardness, diamond is particularly attractive for parts exposed to wear. The most promising synthetic forms are amorphous diamond-like carbon, nanocrystalline diamond, and ultra-nanocrystalline diamond films created by pulsed laser deposition or chemical vapor deposition [9.27–31]. 9.3.3 Metals Silicon on Insulator (SOI) Silicon on insulator (SOI) wafers are also used for MEMS sensors and actuators [9.23]. Different SOI materials are distinguished by their properties. Buried oxide layers can be produced either through ion implantation or wafer bonding processes; these techniques are discussed further below [9.24]. Metals are usually deposited as a thin film by sputtering, evaporation or chemical vapor deposition (CVD). Gold, nickel and iron can also be electroplated. Aluminum is the most common metal used in MEMS/NEMS, and is often used for light reflection and electrical conduction. Gold is used for electrochemistry, infrared (IR) light reflection, and electrical conduction. Chromium is often used as an adhesion layer. Alloys of Ni, such as NiTi and PermalloyTM , can be used for actuation and are discussed in more detail below. 9.3.2 Other Hard Materials 9.3.4 Polymeric Materials Gallium Arsenide Gallium arsenide (GaAs) is a III–V compound semiconductor which is often used to create lasers, optical devices, and high-frequency components. Photoresists Polymeric photoresist materials are generally used as a spin-cast film as part of a photolithographic process. The film is modified by exposure to radiation A Brief Introduction to MEMS and NEMS Polydimethylsiloxane Polydimethylsiloxane (PDMS) is an elastomer used as both a structural component in MEMS devices and a stamping material for creating micro- and nanoscale features on surfaces. PDMS is a common silicone rubber and is used extensively because of its processibility, low curing temperature, stability, tunable modulus, optical transparency, biocompatibility, and adaptability by a range functional groups that can be attached [9.32,33]. 9.3.5 Active Materials There are several types of active materials that successfully perform sensing and actuation functions at the microscale. Several examples of active materials are given below. NiTi Near-equiatomic nickel titanium alloy can be deposited as a thin film and used an as active material. This material is of particular interest to MEMS because the actuation work density of NiTi is more than an order of magnitude higher than the work densities of other actuation schemes. These shape-memory alloys (SMAs) undergo a reversible phase transformation that allows the material to display dramatic and recoverable stress- and temperature-induced transformations. The behavior of NiTi SMA is governed by a phase transformation between austenite and martensite crystal structures. Transformation between the austenite (B2) and martensite (B19) phases in NiTi can be produced by temperature cycling between the hightemperature austenite phase and the low-temperature martensite phase (shape-memory effect), or loading and unloading the material to favor either the highstrain martensite phase or the low-strain austenite phase (superelasticity). Thus both stress and temperature produce the transformation between the austenite and martensite phases of the alloy. The transfor- mation occurs in a temperature window, which can be adjusted from −100 ◦ C to +160 ◦ C by changing the alloy composition and heat treatment processing [9.34]. PermalloyTM PermalloyTM , Nix Fe y , displays magnetoresistance properties and is used for magnetic transducing. Multilayered nanostructures of this alloy give rise to a giant-magnetoresistance (GMR) phenomenon which can be used to detect magnetic fields. It has been widely applied to read the state of magnetic bits in data storage media. Lead Zirconate Titanate (PZT) Lead zirconate titanate (PZT) is a ceramic solid solution of lead zirconate (PbZrO3 ) and lead titanate (PbTiO3 ). PZT is a piezoelectric material that can be deposited in thin film form by sputtering or using a sol–gel process. In addition to natural piezoelectric materials such as quartz, other common synthetic piezoelectric materials include polyvinylidene fluoride (PVDF), zinc oxide, and aluminum nitride. Actuation performed by piezoelectrics has the advantage of being capable of achieving reasonable displacements with fast response, but the material processing is complex. Hydrogels Hydrogels, such as poly(2-hydroxyethyl methacrylate (HEMA)) gel, with volumetric shape-memory capability are now being employed as actuators, fluid pumps, and valves in microfluidic devices. In an aqueous environment, hydrogels will undergo a reversible phase transformation that results in dramatic volumetric swelling and shrinking upon exposure and removal of a stimulus. Hydrogels have been produced that actuate when exposed to such stimuli as pH, salinity, electrical current, temperature, and antigens. Since the rate of swelling and shrinking in a hydrogel is diffusion limited, the temporal response of hydrogel structures can be reduced to minutes or even seconds in microscale devices. 9.3.6 Nanomaterials Nanostructuring of materials can produce unique mechanical, electrical, magnetic, optical, and chemical properties. The materials themselves range from polymers to metals to ceramics, it is their nanostructured nature that gives them exciting new behaviors. Increased hardness with decreasing grain size allows for 209 Part A 9.3 such as visible light, ultraviolet light, x-rays or electrons. Exposure is usually conducted through a mask so that a pattern is created in the photoresist layer and subsequently on the substrate through an etching or deposition process. Resists are either positive or negative depending on whether the radiation exposure weakens or strengthens the polymer. In the developer step, chemicals are used to remove the weaker material, leaving a patterned photoresist layer behind. Important photoresist properties include resolution and sensitivity, particularly as feature sizes decrease. 9.3 Common MEMS/NEMS Materials and Their Properties 210 Part A Solid Mechanics Topics Part A 9.3 Table 9.4 Micromachining processes and their applications (after Kovacs [9.10]) Process Example applications Lithography Thin-film deposition Photolithography, screen printing, electron-beam lithography, x-ray lithography Chemical vapor deposition (CVD), plasma-enhanced chemical vapor deposition (PECVD), physical vapor deposition (PVD) such as sputtering and evaporation, spin casting, sol–gel deposition Blanket and template-delimited electroplating of metals Electroplating, LIGA, stereolithography, laser-driven CVD, screen printing, microcontact printing, dip-pen lithography Plasma etching, reactive-ion enhanced etching (RIE), deep reactive-ion etching (DRIE), wet chemical etching, electrochemical etching Drilling, milling, electrical discharge machining (EDM), focused ion beam (FIB) milling, diamond turning, sawing Direct silicon-fusion bonding, fusion bonding, anodic bonding, adhesives Wet chemical modification, plasma modification, self assembled monolayer (SAM) deposition, grinding, chemomechanical polishing Thermal annealing, laser annealing Electroplating Directed deposition Etching Machining Bonding Surface modification Annealing hard coatings and protective layers, lower percolation threshold impacts conductivity, and narrower bandgap with decreasing grain size enable unique optoelectronics [9.35]. Hundreds of different synthesis routes have been created for manufacturing nanostructured materials. (See, for example, the proceedings of the International Conferences on Nanostructured Materials [9.36].) A few examples of such materials are given below. Carbon Nanotubes and Fullerenes Carbon nanotubes (CNTs) and fullerenes (buckyballs, e.g., C60 ) are self-assembled carbon nanostructures. CNTs are cylindrical graphene structures of single- or multiwall form which are extremely strong and flexible. They possess metallic or semiconducting electronic behavior depending on the details of the structure (chirality). They can be created in an arc plasma furnace, laser ablation, or grown by chemical vapor deposition (CVD) on a substrate using catalyst particles [9.37]. Quantum Dots, Quantum Wires, and Quantum Films Quantum behavior occurs in semiconductor materials (such as GaAs) when electrons are confined to nanoscale dimensions. The confined space forces electrons to have energy states that are clustered around specific peaks, producing fundamentally different electrical and optical properties than would be found in the same material in bulk form. The number of directions free of confinement is used to classify structures, thus two-dimensional (2-D) confinement leads to a quan- tum film, one-dimensional (1-D) confinement leads to a quantum wire, and zero-dimensional (0-D) confinement leads to a quantum dot. The dimension of the confined direction(s) is so small that the energy states are quantized in that direction [9.37]. Nanowires A variety of methods have been developed for making nanowires of a wide range of metals, ceramics, and polymers. Examples include gold nanowires made by a solution method [9.38], palladium nanowires created by electroplating on a stepped surface [9.39], and zinc oxide nanowires created by a vapor/liquid/solid method [9.40]. In one popular technique, electroplating is conducted inside a nanoporous template of alumina or polycarbonate to direct the growth of nanowires [9.41, 42]. The template can be chemically removed, leaving the nanowires behind. In another application, lithographically patterned metal is used as a catalyst for silicon nanowire growth, creating predefined regions of nanowires on a surface [9.43, 44]. Using various combinations of metal catalysts and gases, a wide range of nanowire compositions can be created from chemical vapor deposition methods. 9.3.7 Micromachining Micromachining is a set of material removal and forming techniques for creating microscale movable features and complex structures, often from silicon. The micromachining processes listed in Table 9.4 can be applied to other materials such as glasses, ce- A Brief Introduction to MEMS and NEMS 9.3.8 Hard Fabrication Techniques Hard MEMS utilizes enabling technologies for fabrication and design from the microelectronics industry. The MEMS industry has modified advanced techniques, leveraging well beyond the capability to fabricate integrated circuits. Micromachining involves three fundamental processes: deposition, lithography, and etching. Deposition may employ oxidation, chemical vapor deposition, physical vapor deposition, electroplating, diffusion, or ion implantation. Lithography methods include optical and electron-beam techniques. Etching methods include wet and dry chemical etches, which can be either isotropic (uniform etching in all directions, resulting in rounded features) or anisotropic (etching in one preferential direction, resulting in well-defined features). 9.3.9 Deposition Physical Vapor Deposition (PVD) Physical vapor deposition (PVD) includes evaporation and sputtering. The evaporation method is used to deposit metals on a surface from vaporized atoms removed from a target by heating with an electron beam. This technique is performed under high vacuum and produces very directional deposition and can create shadows. Sputtering of a metallic or nonmetallic material is accomplished by knocking atoms off a target with a plasma of an inert gas such as argon. Sputtering is less directional and allows for higher deposition rates. Chemical Vapor Deposition (CVD) In chemical vapor deposition (CVD), precursor material is introduced into a heated furnace and a chemical reaction takes place on the surface of the wafer. The CVD process is generally performed under low-pressure conditions and is sometimes explicitly referred to as low-pressure CVD (LPCVD). A range of materials can be deposited by CVD, including films of silicon (formed by decomposition of silane (SiH4 )), silicon nitride formed by reacting dichlorosilane (SiH2 Cl2 ) with ammonia (NH3 )), and silicon oxide (formed by silane with an oxidizing species). LPCVD can produce amorphous inorganic dielectric films and polycrystalline polysilicon and metal films. Epitaxy is a CVD process where temperature and growth rate are controlled to achieve ordered crystalline growth in registration with the substrate. PECVD is a plasma-enhanced CVD process. Electroplating A variety of electroplating techniques are used to make micro- and nanoscale components. A mold is created into which metal is plated. Gold, copper, chromium, nickel, and iron are common plating metals. Spin Casting Spin casting is used to create films from a solution. The most common spin-cast material is polymeric photoresist. Sol–Gel Deposition A range of sol–gel processes can be used to make films and particles. The general technique involves a colloidal suspension of solid particles in a fluid that undergo a reaction to generate a gelatinous network. After deposition of the gel, the solvent can be removed to transform the network into a solid phase which is subsequently sintered. Piezoelectric materials such as PZT can be deposited with this method. 9.3.10 Lithography Most of the micromachining techniques discussed below utilize lithography, or pattern transfer, at some point in the manufacturing process. Depending on the resolution required to produce the desired feature sizes and the aspect ratio necessary, lithography is either per- 211 Part A 9.3 ramics, polymers, and metals, but silicon is favored because of its widespread use and the availability of design and processing techniques. Other advantages of silicon include the availability of relatively inexpensive pure single-crystal substrate wafers, its desirable electrical properties, its well-understood mechanical properties, and ease of integration into a circuit for control and signal processing [9.7]. Although often performed in batch processes, micromachining for MEMS application may make large-aspect-ratio features and incorporation of novel or active materials a higher priority than batch manufacturing. This opens the door for a wider range of fabrication techniques such as focused ion-beam milling, laser machining, and electron-beam writing [9.22, 45, 46]. A brief overview of micromachining is provided below. The following books, used as references for the discussion here, are valuable resources for more extensive information [9.1, 2, 10, 18, 22, 33]. Additional information can be found in Taniguchi [9.47] and Evans [9.48] on microfabrication technology, Bustillo [9.49] on surface micromachining, and Gentili [9.50] on nanolithography. 9.3 Common MEMS/NEMS Materials and Their Properties Solid Mechanics Topics Part A 9.3 formed with ultraviolet light, an ion beam, x-rays, or an electron beam. X-ray lithography can produce features down to 10 nm and electron beams can be focused down to less than 1 nm [9.50]. Optical lithography allows aspect ratios of up to three whereas x-ray lithography can produce aspect ratios > 100. This large depth of focus, lack of scattering effects, and insensitivity to organic dust make x-ray lithography very attractive for NEMS production. Electron-beam lithograph has the attractive feature that a pattern can be directly written onto a resist, as well as the fact that it produces lower defect densities with a large depth of focus, but the process must be performed in vacuum. In most cases a mask that carries either a positive or negative image of the features to be created must first be produced. Masks are commonly made with a chromium layer on fused silica. Photoresist covering the chromium is exposed with an optical pattern generated from a sequence of small rectangles used to draw out the pattern desired. Other mask production techniques include photographic emulsion on quartz, electron-beam lithography with electron-beam resist, and high-resolution ink-jet printing on acetate or mylar film. Photolithographic fabrication techniques have a long history of use with ceramics, plastics, and glasses. In the case of silicon fabrication, the wafer is coated with a polymeric photoresist layer sensitive to ultraviolet light. Exposure of the photoresist layer is conducted through a mask. Depending on whether a positive or negative photoresist is used, the light either weakens the polymer or strengthens the polymer. In the developer step, chemicals are used to remove the weaker material, leaving a patterned photoresist layer behind. The photoresist acts as a protective layer when etching is conducted. Contact lithography produces a 1:1 ratio of the mask size and feature size. Proximity lithography also gives a 1:1 ratio with slightly lower resolution because a gap is left between the mask and the substrate to minimize damage to the mask. A factor of 5–10 reduction is common for projection step-and-repeat lithography. Because this technique allows for the production of feature sizes smaller then the mask, only a small region is exposed at one time and the mask must be stepped across the substrate. 9.3.11 Etching A number of wet and dry etchants have been developed for silicon. Important properties include orientation dependence, selectivity, and the geometric details of the etched feature (Fig. 9.2). A common isotropic wet etchant for silicon is HNA (a combination of HF, HNO3 , and CH3 COOH), while anisotropic wet etchants include KOH, which etches {100} planes 100 times faster than {111} planes, tetramethylammonium hydroxide (called TMAH or (CH3 )4 NOH), which etches {100} planes 30–50 times faster than {111} planes but leaves silicon dioxide and silicon nitride unetched, and ethylenediamine pyrochatechol (EDP), which is very hazardous but does not etch most metals. Wet etchants such as HF for silicon dioxide, H3 PO4 for silicon nitride, KCl for gold, and acetone for organic layers, can be performed in batch processes with little cost [9.51]. An important feature of an etchant is its selectivity; for example, the etch rate of an oxide by HF is 100 nm/min compared to 0.04 nm/min for silicon nitride [9.51]. The etching reaction can be either reaction rate controlled or mass transfer limited. Because wet etchants act quickly, making it hard to control depth of the etch, electrochemical etching is sometimes employed using an electric potential to moderate the reaction along with a precision thickness epitaxial layer used for etch stop. The challenge comes with drying after the wet etching process is complete. Capillary forces can easily draw surfaces together, causing damage and stiction. Supercritical drying, where the liquid is converted to a gas, can be used to prevent this. Alternatively, application of a hydrophobic passivation layer such as a fluorocarbon polymer can be used to prevent stiction. Chemically reactive vapors and plasmas are highly effective dry etchants. Xenon difluoride (XeF2 ) is a commercially important highly selective vapor etchant for silicon. Dry etchants such as CHF3 + O2 for silicon dioxide, SF6 for silicon nitride, Cl2 + SiCl4 for Wet etch Plasma (dry) etch Isotropic Part A Anisotropic 212 {111} Fig. 9.2 Trench profiles produced by different etching processes (after Maluf [9.1]) A Brief Introduction to MEMS and NEMS all parts of the wafer. Ion milling refers to selective sputtering and can be done uniformly over a wafer or with focusing electrodes by focused ion-beam milling (FIB). FIB is also becoming a more important technique for test sample production and the application of gratings used for interferometry [9.52]. In addition to FIB, techniques such as scanning probe microscope (SPM) lithography and molecular-beam epitaxy can also be used to create micro- and nanoscale gratings [9.53]. Beyond the use of etching as part of the initial fabrication of a device, some small adjustments may be required after the device is fabricated due to small variations that occur in processing. Compensation can be performed by trimming resistors and altering mechanical dimensions via techniques such as laser ablation and FIB milling. Calibration can be performed electronically with correction coefficients. 9.4 Bulk Micromachining versus Surface Micromachining The processes for silicon micromachining fall into two general categories: bulk (subtraction of substrate material) and surface (addition of layers to the substrate). Other techniques used on a range of materials include surface micromachining, wafer bonding, thin film screen printing, electroplating, lithography galvanoforming molding (LIGA, from the German Lithografie-Galvanik-Abformung), injection molding, electric-discharge machining (EDM), and focused ion beam (FIB). Figure 9.3 provides a basic comparison of bulk micromachining, surface micromachining, and LIGA. Bulk Micromachining Removal of significant regions of substrate material in bulk micromachining is accomplished through a) b) c) Bulk micromachining Surface micromachining LIGA Resist structure Base plate Deposition of sacrificial layer Deposition of silica layers on Si Membrane <111> face Metal structure Electroforming Patterning with mask Gate plate Patterning with mask and etching of Si to produce cavity Silicon Silica Lithography Mold insert Mold fabrication Deposition of microstructure layer Molding mass Mold filling Etching of sacrificial layer to produce freestanding structure Silicon Polysilicon Sacrificial material Plastic structure Unmolding Fig. 9.3a–c Schematic diagrams depicting the processing steps required for (a) bulk micromachining (b) surface micromachining, and (c) LIGA. All views are shown from the side (after Bhushan [9.2] Chap. 50) 213 Part A 9.4 aluminum, and O2 for organic layers, are used as a plasma [9.51]. The process is conducted in a specially designed system that generates a chemically reactive plasma species of ion neutrals and accelerates them towards a substrate with an electric or magnetic field. Plasma etching is the spontaneous reaction of neutrals with the substrate materials, while reactive-ion etching involves a synergistic role between the ion bombardment and the chemical reaction. Deep reactive-ion etching (DRIE) allows for the creation of high-aspectratio features. DRIE involves periodic deposition of a protective layer to shield the sidewalls either through condensation of reactant gasses produced by cryogenic cooling of the substrate or interim deposition cycles to put down a thin polymer film. Ions can also be used to sputter away material. For example, argon plasma will remove material from 9.4 Bulk Micromachining versus Surface Micromachining 214 Part A Solid Mechanics Topics Part A 9.5 anisotropic etching of a silicon single-crystal wafer. The fabrication process includes deposition, lithography, and etching. Bulk micromachining is commonly used for high-volume production of accelerometers, pressure sensors, and flow sensors. are used to produce a free-standing structure. Surface micromachining is attractive for integrating MEMS sensors with electronic circuits, and is commonly used for micromirror arrays, motors, gears, and grippers. Surface Micromachining Alternating structural and sacrificial thin film layers are built up and patterned in sequence for surface micromachining. The process used by Sandia National Laboratory uses up to five structural polysilicon and five sacrificial silicon dioxide layers, whereas Texas Instrument’s digital micromirror device (discussed below) is made from a stack of structural metallic layers and sacrificial polymer layers [9.1, 54]. Deposition methods include oxidation, chemical vapor deposition (CVD), and sputtering. Annealing must sometimes be used to relax the mechanical stresses that build up in the films. Lithography and etching LIGA Lithography galvanoforming molding (LIGA, from the German Lithografie-Galvanik-Abformung) is a lithography and electroplating method used to create very high-aspect-ratio structures (aspect ratios of more than 100 are common). The use of x-rays in the lithography process takes advantage of the short wavelength to create a larger depth of focus compared to photolithography [9.14]. Devices can be up to 1 mm in height with another dimension being only a few microns and are commonly made of materials such as metals, ceramics, and polymers. See Guckel [9.55], Becker [9.56], and Bley [9.57] for additional details. 9.5 Wafer Bonding Although microelectronics fabrication processes allow stacking layers of films, structures are relatively two dimensional. Wafer bonding provides an opportunity for a more three-dimensional structure and is commonly used to make pressure sensors, accelerometers, and microfluidic devices (Fig. 9.4). Anodic and direct bonding are the most common techniques, but 2.5μm LPCVD polysilicon mask Sheat Sample inlet injector holes Fused silica substrate Electrochemical-discharge machined through-holes Two substrates thermally bonded together Fig. 9.4 Schematic diagram depicting a wafer bonding process used to create a microfluidic channel for flow cytometry (after Kovacs [9.10]) bonding can also be achieved by using intermediate layers such as polymers, solders, and thin-film metals. Anodic (electrostatic) bonding can be used to bond silicon to a sodium-containing glass substrate (with a matched coefficient of thermal expansion) using an applied electric field. This is accomplished with the application of a large voltage at elevated temperature to make positive Na+ ions mobile. The positively charged silicon is held to the negatively charged glass by electrostatic attraction. Direct (silicon-fusion) bonding requires two flat, clean surfaces in intimate contact. Direct bonding of a silicon/glass stack can be achieved by applying pressure. Direct wafer bonding allows joining of two silicon surfaces or silicon and silicon dioxide surfaces and is used extensively to create SOI wafers. After treatment of the surfaces to produce hydroxyl (OH) groups, intimate contact allows van der Waals forces to make the initial bond followed by an annealing step to create a chemical reaction at the interface. Grinding and polishing is sometimes needed to thin a bonded wafer. Annealing must be performed afterwards to remove defects incurred during grinding. Alternatively, chemomechanical polishing can be used to combine chemical etching with the mechanical action of polishing. A Brief Introduction to MEMS and NEMS 9.6 Soft Fabrication Techniques Self-Assembly Partly because of the high cost of nanolithography and the time-consuming nature of atom-by-atom placement using probe microscopy techniques, self-assembly is an important bottom-up approach to NEMS fabrication [9.59]. To offset the time it takes to build unit by unit to create a useful device, massive parallelism and autonomy is required. The advantage of self-assembly is that it occurs at thermodynamic minima, relying on naturally occurring phenomena that govern at the nanoscale and create highly perfect assemblies [9.58]. The atoms, molecules, collections of molecules, or nanoparticles self-organize into functioning entities using thermodynamic forces and kinetic control [9.60]. Such self-organization at the nanoscale is observed naturally in liquid crystals, colloids, micelles, and self-assembled monolayers [9.61]. Reviews of self-assembly can be found in [9.62–65]. At the nanoparticle level, a variety of methods have been used to promote self-assembly. Three basic requirements must be met: there must be some sort of bonding force present between particles or the particles and a substrate, the bonding must be selective, and the particles must be in random motion to facilitate chance interactions with a relatively high rate of occurrence. Additionally, for the technique to be practical, the particles must be easily synthesized. Selectivity can be facilitated by micromachining the substrate including patterns with geometric designs that allow for only certain orientations of the mating particle. Particularly powerful are self-assembly methods using complementary pairs and molecular building blocks (analogous to DNA replication). Complementary pairs can bind electrostatically or chemically (using functional groups with couple monomers). Molecular Table 9.5 Techniques for creating patterned SAMs [9.58] Method Scale of features Microcontact printing Micromachining Microwriting with pen Photolithography/lift-off Photochemical patterning Photo-oxidation Focused ion-beam writing Electron-beam writing Scanning tunneling microscope writing 100 nm – some cm 100 nm – some μm ≈ 10–100 μm > 1 μm > 1 μm > 1 μm ≈ some μm 25–100 nm 15–50 nm building blocks can use a number of different bonds and linkages (ionic bonds, hydrogen bonds, transition metal complex bonds, amide linkages, and ester linkages) to create building blocks for three-dimensional (3-D) nanostructures and nanocrystals such as quantum dots. Self-assembled monolayers (SAMs) can be produced in patterned form by several techniques that produce features in a range of micro- and nanoscale sizes (Table 9.5). Combined with lithography, defined areas of self-assembly on a surface can be created. Applications of SAMs include fundamental studies of wetting and electrochemistry, control of adhesion, surface passivation (to protect from corrosion, control oxidation, or use as resist), tribology, directed assembly, optical systems, colloid fabrication, and biologically active surfaces for biotechnology [9.58]. Soft Lithography The term soft lithography encompasses a number of techniques that can be used to fabricate microand nanoscale structures using replica molding and self-assembly. These techniques include microcontact printing, replica molding, microtransfer molding, micromolding in capillaries, and solvent-assisted micromolding [9.66]. As an example, microcontact printing uses a selfassembled monolayer as ink in a stamping operation that transfers the SAM to a surface (Fig. 9.5). The stamp is fabricated from of an elastomeric material such as PDMS by casting onto a master with surface features. The master can be produced with a range of photolithographic techniques. The polymeric replica mold is used as a stamp to enable physical pattern transfer. The advantages of microcontact printing are its simplicity, conformal contact with a surface, the reusability of the stamp, and the ability to produce multiple stamps from one master. Although defect density and registration of patterns over large scales can be issues, the flexibility of the stamp can be use to make small features (≈ 100 nm) using compression or pattern transfer onto curved surfaces [9.58]. The aspect ratio of features is a constraint with PDMS however. Ratios between 0.2 and 2 must be used to ensure defect-free stamps and molds [9.67]. 9.6.1 Other NEMS Fabrication Strategies Nanoscale structures can be created from both topdown and bottom-up approaches. Because of the push Part A 9.6 9.6 Soft Fabrication Techniques 215 216 Part A Solid Mechanics Topics Part A 9.6 to miniaturize commercial electronics, many top-down methods are refinements of micromachining techniques with the goal of achieving manufacturing accuracy on the nanometer scale. Bottom-up methods rely on additive atomic and molecular techniques, such as self-organization, self-assembly, and templating, using building blocks similar and size to those used in nature [9.68]. A brief review of some additional examples is provided below. Nanomachining Scanning probe microscopes (SPMs) are a valuable set of tools for NEMS characterization, but these tools Photoresist Si Photolithography is used to create a master Photoresist pattern (1–2 μm thickness) Si PDMS is poured over master and cured PDMS Photoresist pattern Si PDMS is peeled away from the master PDMS PDMS is exposed to a solution containing HS(CH2)15CH3 PDMS Alkanethiol Stamping onto gold substrate transfers thiol to form SAM SAMs (1–2 nm) Au (5–2000 nm) Si Ti (5–10 nm) Metal not protected by SAM is removed by exposure to selective chemical etchant Si Fig. 9.5 Schematic diagram depicting the processing steps required for microcontact printing. All views are shown from the side (after Wilbur [9.58]) can also be used for NEMS manufacturing. These microscopes share the common feature that they employ a nanometer-scale probe tip in the proximal vicinity of a surface. They are many times more powerful than scanning electron microscopes because their resolution is not determined by wavelength for the interaction with the surface under investigation. The scanning tunneling microscope (STM) can be used to create a strong electric field in the vicinity of the probe tip to manipulate individual atoms. Atoms can be induced to slide over a surface in order to move them into a desired arrangement by mechanosynthesis [9.69]. Resolution is effectively the size of a single atom but practically the process is exceptionally time consuming and the sample must be held at very low temperature to prevent movement of atoms out of place [9.70]. With slightly less resolution but still less than 100 nm, an STM can also be used to write on a chemically amplified negative electron-beam resist. Nanolithography Surface micromachining can be conducted at the nanoscale using electron-beam lithography to create free-standing or suspended mechanical objects. Although the general approach parallels standard lithography (see above), the small-scale ability of this technique is enabled by the fact that an electron beam with energy in the keV range is not limited by diffraction. The electron beam can be scanned to create a desired pattern in the resist [9.8]. Nanoscale resolution can also be obtained using alternative lithographic techniques such as dip-pen nanolithography (DPN) [9.71]. This technique employs an atomic force microscope (AFM) probe tip to deposit a layer of material onto a surface, much as a pen writes on paper. A pattern can be drawn on a surface using a wide range of inks such as thiols, silanes, metals, sol–gel precursors, and biological macromolecules. Although the DPN process is inherently slower that standard mask lithographic techniques, it can be used for intricate functions such as mask repair and the application of macromolecules in biosensor fabrication, or it can be parallelized to increase speed [9.72]. This and other nanofabrication techniques using AFM to modify and pattern surfaces are reviewed by Tang [9.73]. 9.6.2 Packaging Packaging of a MEMS/NEMS device provides a protective housing to prevent mechanical damage, minimize stresses and vibrations, guard against contamination, A Brief Introduction to MEMS and NEMS considerations include thickness of the device, wafer dicing (separation of the wafer into separate dice), sequence of final release, cooling of heat-dissipating devices, power dissipation, mechanical stress isolation, thermal expansion matching, minimization of creep, protective coatings to mitigate damaging environmental effects, and media isolation for extreme environments [9.1]. In the die-attach process, each individual die is mounted into a package, by bonding it to a metal, ceramic or plastic platform with a metal alloy solder or an adhesive. For silicon and glass, a thin metal layer must be placed over the surface prior to soldering to allow for wetting. Electrical interconnects can be produced with wire bonding (thermosonic gold bonding with ultrasonic energy and elevated temperature) and flip-chip bonding (using solder bumps between the die and package pads). Fluid interconnects are created by insertion of capillary tubes, mating of self-aligning fluid ports, and laminated layers of plastic [9.1]. 9.7 Experimental Mechanics Applied to MEMS/NEMS With a basic understanding of the materials and processes used to make MEMS/NEMS devices, the role of mechanics in materials selection, process validation, design development, and device characterization can now be discussed. The remainder of this chapter will focus on the forces and phenomena dominant at the micrometer and nanometer scales, basic device characterization techniques, and mechanics issues that arise in MEMS/NEMS devices. 9.8 The Influence of Scale To gain perspective on the micrometer and nanometer size scales, consider that the diameter of human hair is 40–80 μm and a DNA molecule is 2–3 nm wide. The weight of a MEMS structure can be about 1 nN and that of a NEMS components about 10−20 N. Compare this to the mass of a drop of water (10 μN) or an eyelash (100 nN) [9.2]. The minuscule size of forces that influence behavior at these small scales is hard to imagine. For instance, if you take a 10 cm length of your hair and hold it like a cantilever beam, the amount of force placed on the tip of the cantilever to deflect it by 1 cm is on the order of 1 pN. That piece of hair is 40–80 μm in diameter, which is large compared to most MEMS/NEMS components. In dealing with micro- and nanoscale devices, engineering intuition developed through experience with macroscale behavior is often misleading. It should be noted that many macroscale techniques can be applied at the micro- and nanoscales, but advantages come not from miniaturization but rather working at the relevant size scale using the uniqueness of the scale. The balance of forces at these scales differs dramatically from the macroscale (Table 9.6). Compared to a macroscale counterpart of the same aspect ratio, the structural stiffness of a microscale cantilever increases relative to inertially imposed loads. When the length scale changes by a factor of a thousand, the area decreases by a factor of a million and the volume by a factor of a billion. Surface forces, proportional to area, become a thousand times large than forces that are proportional to volume, thus inertial and electromagnet forces become negligible. At small scales, adhesion, friction, stiction 217 Part A 9.8 protect from harsh environmental conditions, dissipate heat, and shield from electromagnetic interference [9.15]. Packaging is critical because it enables the usefulness, safety, and reliability of the device. Hermetic packaging made of metal, ceramic, glass or silicon is used to prevent the infiltration of moisture, guard against corrosion, and eliminate contamination. The internal cavity is evacuated or filled with an inert gas. For MEMS/NEMS the packaging may also be required to provide access to the environment through electrical and/or fluid interconnects and optically transparent windows. In these cases, the devices are left more vulnerable in order for them to interact with the environment to perform their function. Although there are well-established techniques for packaging of common microelectronics devices, packaging of MEMS/NEMS presents particular challenges and may account for 75–95% of the overall cost of the device [9.1]. Packaging design must be conducted in parallel with design of the MEMS/NEMS component. Design 9.8 The Influence of Scale 218 Part A Solid Mechanics Topics Part A 9.8 Table 9.6 Scaling laws and the relative importance of phenomena as they depend on linear dimension, i (after Madou [9.22]) Importance at small scale Diminished Increased Phenomena Power of linear dimension Flow Gravity Inertial force Magnetic force Thermal emission Electrostatic force Friction Pressure Piezoelectricity Shape-memory effect Velocity Surface tension Diffusion van der Waal force l4 l3 l3 l 2 , l 3 or l 4 l 2 or l 4 l2 l2 l2 l2 l2 l l l 1/2 l 1/4 (static friction), surface tension, meniscus forces, and viscous drag often govern. Acceleration of a small object becomes rapid. At the nanoscale, phenomena such as quantum effects, crystalline perfection, statistical time variation of properties, surface interactions, and interface interactions govern behavior and materials properties [9.35]. Additionally, the highly coupled nature of thermal transport properties at the microscale can be either an advantage or disadvantage depending on the device. Enhanced mass transport due to large surface-to-volume ratio can be a significant advantage for applications such as capillary electrophoresis and gas chromatography. However, purging air bubbles in microfluidic systems can be extremely difficult due to capillary forces. The interfacial surface tension force will cause small bubble less than a few millimeters in diameter to adhere to channel surfaces because the mass of liquid in a capillary tube produces an insubstantial inertial force compared to the surface tension [9.10]. Some scaling effects favor particular micro- and nanoscale situations but others do not. For instance, large surface-to-volume ratio in MEMS devices can undermine device performance because of the retarding effects of adhesion and friction. However, electrostatic force is a good example of a phenomena that can have substantial engineering value at small scales. Transla- tional motion can be achieved in MEMS by electrostatic force because this scales as l 2 as compared to inertial force which scales as l 3 . Microactuation using electrostatic forces between parallel plates is used in comb drives, resonant microstructures, linear motors, rotary motors, and switches. In relation to MEMS testing, gripping of a tension sample can be achieved using an electrostatic force between a sample and the grip [9.74]. It is also important to note that, as the size scale decreases, breakdown in the predictions of continuumbased theories can occur at various length scales. In the case of electrostatics, electrical breakdown in the air gap between parallel plates separated by less than 5 μm does not occur at the predicted voltage [9.75]. In optical devices, nanometer-scale gratings can produce an effective refractive index different from the natural refractive index of the material because the grating features are smaller than the wavelength of light [9.76]. For resonant structures continuum mechanics predictions break down when the structure’s dimensions are on the order of tens of lattice constants in cross section [9.11]. Detailed discussions of issues related to size scale can be found in Madou [9.22] and Trimmer [9.77]. 9.8.1 Basic Device Characterization Techniques A range of mechanical properties are needed to facilitate design, predict allowable operating limits, and conduct quality control inspection for MEMS. As with any macroscale device or component, structural integrity is critical to MEMS/NEMS. Concerns include friction/stiction, wear, fracture, excessive deformation, and strength. Properties required for complete understanding of the mechanical performance of MEMS/NEMS materials include elastic modulus, strength, fracture toughness, fatigue strength, hardness, and surface topography. In MEMS devices the minimum feature size is on the order of 1 μm, which is also the natural length scale for microstructure (such as the grain size, dislocation length, or precipitate spacing) in most materials. Because of this, many of the mechanical properties of interest are size dependent, which precipitates the need for new testing methods given that knowledge of material properties is essential for predicting device reliability and performance. A detailed discussion about micro- and nanoscale testing can be found in Part B of this handbook as well as in references such as Sharpe [9.78], Srikar [9.79], Haque and Saif [9.80], Bhushan [9.2], and Yi [9.81]. The following sections A Brief Introduction to MEMS and NEMS 9.8.2 Residual Stresses in Films Many MEMS/NEMS devices involve thin films of materials. Properties of thin-film material often differ from their bulk counterparts due to the high surfaceto-volume ratio of thin films and the influence of surface properties. Additionally, these films must have good adhesion, low residual stress, low pinhole density, good mechanical strength, and good chemical resistance [9.22]. These properties often depend on deposition and processing details. The stress state of a thin film is a combination of external applied stress, thermal stress, and intrinsic residual stress that may arise due to factors such as doping (in silicon), grain boundaries, voids, gas entrapment, creep, and shrinkage with curing (in polymeric materials). Stresses that develop during deposition of thin-film material can be either tensile or compressive and may give rise to cracking, buckling, blistering, delaminating, and void formation, all of which degrade device performance. Residual stresses can arise because of coefficient of thermal expansion mismatch, lattice mismatch, growth processes, and nonuniform plastic deformation. Residual stresses that do not cause mechanical failure may still significantly affect device performance by causing warping of released structures, changes in resonant frequency of resonant structures, and diminished electrical characteristics. In some instances, however, residual stresses can be used productively, such as in shape setting of shape-memory alloy films or stress-modulated growth and arrangement of quantum dots. There are numerous techniques for measuring residual stresses in thin films. Fundamental techniques rely on the fact that stresses within a film will cause bending in its substrate (tension causing concavity, compression causing convexity). Simple displacement measurements can be conducted on a circular disk or a micromachined beam and stress calculated from the radius of curvature of the bent substrate or the deflection of a cantilever. Strain gages may also be made directly in the film and used to make local measurements. Freestanding portions of the thin film can be created by micromachining so that the films stresses can be explored by applied pressure, external probe, critical length for buckling, or resonant frequency measurements. For instance, the critical stress to cause buckling in a doubly supported beam can be estimated from: π 2t2 , K L2 where K is a constant determined by the boundary conditions (3 for a doubly supported beam), E is Young’s modulus, t is the beam thickness, and L is the shortest length of beam displaying buckling [9.10]. The stress or strain gradient over a region of a film can be found by measuring deflections in a simple cantilever. The upward or downward deflection along the length of the beam can be measured by optical methods and used to estimate the internal bending moment M from the expression: σCR = E (1 − ν2 ) δ (x) =K+ Mx , x 2E I where δ(x) is the vertical deflection at a distance x from the support, E is Young’s modulus, ν is Poisson’s ratio, I is the moment of the beam cross section about the axis of bending, and K is a constant determined by the boundary conditions at the support [9.83]. A number of techniques have been developed for determining residual stresses including an American Society for Testing and Materials (ASTM) standard involving optical interferometry [9.84]. The bulge test is a basic technique for measuring residual stress in a freestanding thin film [9.85]. The bulge test structure can be easily created by micromachining with well-defined boundary conditions. The M-test is an on-chip test that uses bending of an integrated free-standing prismatic beam [9.86]. The principle of an electrostatic actuator is used to conduct the test to find the onset of instability in the structure. The wafer curvature test is regularly used for residual stress measurement in nonintegrated film structures, and can be used even when the film thickness is much smaller than the substrate thickness [9.87]. Dynamic testing can be used to measure resonant frequency and extract information about residual stress and modulus. Resonant frequency increases with tension and decreases in compression [9.88, 89]. Air damping can significantly impact theses measurements, however, so they must be conducted in vacuum [9.18]. Other established techniques that can be employed to measure residual stresses in films include passive strain sensors, Raman spectroscopy, and nanoindentation [9.79]. More recently, nanoscale gratings created by focused ion-beam (FIB) milling have been used to 219 Part A 9.8 provide a review of some mechanics issues that arise at the device level. The following sources, used as references for the discussion below, should be consulted for additional background on mechanics, metrology, and MEMS: Trimmer [9.77], Madou [9.22], Bhushan [9.2], and Gorecki [9.82]. 9.8 The Influence of Scale 220 Part A Solid Mechanics Topics Part A 9.8 produce moiré interference between the grating on the specimen surface and raster scan lines of a scanning electron microscope (SEM) image [9.90]. This technique can be used to provide details of residual strains in microscale structures as they evolve with etching of the underlying sacrificial layer [9.52]. Digital image correlation (DIC) has also been applied to SEM and atomic force microscopy (AFM) images in combination with FIB. DIC is used to capture deformation fields while nearby FIB milling of the specimens releases residual stresses, allowing very local evaluation [9.91]. 9.8.3 Wafer Bond Integrity Wafer bonding is often an essential device fabrication step, particularly for microfluidic devices, microengines, and microscale heat exchangers. Although direct bonding of silicon can achieve strengths comparable to bulk silicon, the process is sensitive to bonding parameters such as temperature and pressure. The appearance of voids and bubbles at the interface is particularly undesirable for both strength and electrical conductivity [9.92]. An important nondestructive technique for assessing the bond quality of bonded silicon wafers is infrared transmission. At IR wavelengths of about 1.1 μm silicon is transparent [9.1]. Quantification of bond strength can be conducted with techniques such as the pressure burst test, tensile/shear test, knife-edge test, or four-point bend-delamination test [9.93]. Although a range of techniques and processes can be employed to bond both similar and dissimilar materials, the stresses and deformation of the wafers that develop are consistent. The residual stress stored in the bonded wafers is important because it may provide the elastic strain energy to drive fracture. Details of the wafer geometry can impact the final shape of the bonded pair and the integrity of the bond interface [9.94]. 9.8.4 Adhesion and Friction Adhesion is both essential and problematic for MEMS/NEMS. For multilayered devices, good adhesion between layers is critical for overall performance and reliability, where delamination under repetitive applied mechanical stresses must be avoided. Adhesion between material layers can be enhanced by improved substrate cleanliness, increased substrate roughness, increased nucleation sites during deposition, and addition of a thin adhesion-promoting layer. Standard tests for film adhesion include: the scotch-tape test, abrasion, scratching, deceleration using ultrasonic and ultracentrifuge techniques, bending, and pulling [9.95]. In situ testing of adhesion can also be conducted by pressurizing the underside of a film until initiation of delamination. This method also allows the determination of the average work of adhesion. Adhesion can be problematic if distinct components or a component and the nearby substrate come into contact, causing the device to fail. For example, although the mass in an accelerometer device is intended to be free standing at all points of operation, adhesion can occur in the fabrication process. Commonly with freestanding portions of MEMS structures, the capillary forces present during the drying of a device after etching to remove sacrificial material are large enough to cause collapse of the structure and failure due to adhesion [9.96]. To avoid this problem, supercritical drying is used. Contacting surfaces that must move relative to one another in MEMS/NEMS are minimized or eliminated altogether, the reason being that friction and adhesion at these scales can overwhelm the other forces at play. Because silicon readily oxidizes to form a hydrophilic surface, it is much more susceptible to adhesion and accumulation of static charge [9.97]. When contacting surfaces are involved, lubricant films and hydrophobic coatings with low surface energy can be applied to minimize wear and stiction (the large lateral force required to initiate relative motion between two surfaces). For instance, Analog Devices uses a nonpolar silicone coating in its accelerometers to resist charge buildup and stiction [9.98]. Processing plays a major role in surface properties such as friction and adhesion. Polishing will dramatically affect roughness, as in the case of polysilicon where roughness can be reduced by an order of magnitude from the as-deposited state [9.2]. The doping process can also lead to higher roughness. Organic monolayer films show promise for lubrication of MEMS to reduce friction and prevent wear. The atomic force microscope and the surface force apparatus used to quantify friction and MEMS test structures such as those developed at Sandia National Laboratory are aiding the development of detailed mechanics models addressing friction [9.99–102]. Flow Visualization Flow in the microscale domain occurs in a range of MEMS devices, particularly in bioMEMS, microchannel networks, ink-jet printer heads, and micropropulsion systems. The different balance of forces at micro- A Brief Introduction to MEMS and NEMS flow fields in microfluidic devices, where micron-scale spatial resolution is critical [9.103]. Microparticle image velocimetry (μPIV) has been used to characterize such things as microchannel flow [9.104] and microfabricated ink-jet printer head flow [9.105]. For the high-velocity, small-length-scale flows found in microfluidics, high-speed lasers and cameras are used in conjunction with a microscope to image the particles seeded in the flow. With μPIV techniques, the flow boundary topology can be measured to within tens of nanometers [9.106]. 9.9 Mechanics Issues in MEMS/NEMS 9.9.1 Devices A wide range of MEMS/NEMS devices is discussed in the literature, both as research and commercialized devices. These devices are commonly planar in nature and employ structures such as cantilever beams, fixed–fixed beams, and springs that are loaded in bending and torsion. A range of mechanics calculations are needed for device characterization, including the effective stiffness of composite beams, deflection analysis of beams, modal analysis of a resonant structures, buckling analysis of a compressively loaded beams, fracture and adhesion analysis of structures, and contact mechanics calculations for friction and wear of surfaces. A substantial literature is available on the application of mechanics to MEMS/NEMS devices. The selected MEMS/NEMS examples presented below were chosen for their illustrative nature. Digital Micromirror Device Optical MEMS devices range from bar-code readers to fiber-optic telecommunication, and use a range of wideband-gap materials, nonlinear electro-optic polymers, and ceramics [9.107]. (See Walker and Nagel [9.108] for more information on optical MEMS.) A wellestablished commercial example of an optical MEMS device is the Digital Micromirror DeviceTM (DMD) by Texas Instruments used for projection display (Fig. 9.6) [9.109]. These devices have superior resolution, brightness, contrast, and convergence performance compared to conventional cathode ray tube technology [9.2]. The DMD contains a surface micromachined array of half a million to two million independently controlled, reflective, hinged micromirrors that have a mechanical switching time of 15 μs [9.110]. This de- vice steers a reflected beam of light with each individual mirrored aluminum pixel. Pixel motion is driven by an electrostatic field between the yoke and an underlying electrode. The yoke rotates and comes to rest on mechanical stops and its position is restored upon release by torsional hinge springs [9.111]. Almost all commercial MEMS structures avoid any contact between structural members in the operation of the device, and sliding contact is avoided completely because of stiction, friction, and wear. The DMD is currently the only commercial device where structural components come in and out of contact, with contact occurring between the mirror spring tips and the underlying mechanical stops, which act as landing sites. To prevent adhesion problems in the DMD, a self healing perfluorodecanoic acid coating is used on the structural aluminum components [9.112]. Other challenges for the DMD include creep and fatigue behavior in the hinge, shock and vibration, and sensitivity to debris within the package [9.2]. The primary failure mechanisms are surface contamination and hinge memory due to creep in the metallic alloy resulting in a residual tilt angle [9.1]. Heat transfer, which contributes to the creep problem, is also an issue for micromirrors. When the reflection coefficient is less than 100% some of the optical power is absorbed as heat and can cause changes in the flatness of the mirror, damage to the reflective layer, and alterations in the dynamic behavior of the system [9.113]. Micromirrors for projection display involve rotating structures and members in torsion. Such torsional springs must be well characterized and their mechanics well modeled. For production devices extensive finite element models are developed to optimize performance [9.114]. For initial design calculations however, 221 Part A 9.9 scopic length scales can influence fluid flow to produce counterintuitive behavior in microscopic flows. Additionally, the breakdown in continuum laws for fluid flow begins to occur at the microscale. For instance, the no-slip condition no longer applies and the friction factor starts to decrease with channel reduction. Particle image velocimetry (PIV) is a technique commonly used at macroscopic length scales to measure velocity fields through the use of particles seeded in the fluid. The technique has been adapted to measure 9.9 Mechanics Issues in MEMS/NEMS 222 Part A Solid Mechanics Topics Part A 9.9 some closed-form solutions for mechanics analysis can be employed. For instance, an appropriate material can be chosen or the basic dimensional requirements can be found from calculation of the maximum shear stress τmax in a beam of elliptical cross section in torsion (with a and b the semi-axis lengths) using: 2Gαa2 b , for a > b , a 2 + b2 where G is the shear modulus and α in the angular twist [9.107]. Mechanical integrity of the DMD relies on low stresses in the hinge, thus the tilt angle is limited to ±10◦ [9.1]. τmax = atomic force microscopy [9.117, 118], and magnetic beads [9.119], but these techniques have the disadvantage of requiring external probes, labeling, and/or optical excitation. Alternatively, there are several methods using molecular recognition and the small-scale forces created by events such as DNA hybridization and receptor–ligand binding to produce bending in cantilevers to create sensors with high selectivity and resolution [9.115, 120]. Microcantilever sensors have been used for some time to detect changes in relative humidity, temperature, pressure, flow, viscosity, sound, natural gas, mercury vapor, and ultraviolet and infrared radiation. More re- Biomolecular Recognition Device Biological molecules can be probed by external methods using techniques such as optical tweezers [9.116], a) 100 μm Oligonucleotide x b) Mirror –10 deg Mirror +10 deg Hybridization Δx Hinge Yoke Landing tip CMOS substrate Fig. 9.6 (a) SEM image of yoke and hinges of one pixel with mirror removed (b) Schematic of two tilted pixels with mirrors (shown as transparent) (reprinted with permission, Hornbeck [9.111], Bhushan [9.2]) Fig. 9.7 SEM image of a portion of the cantilever sensor array and schematics illustrating functionalized cantilevers with selective sensing capability (reprinted with permission, Fritz [9.115]) A Brief Introduction to MEMS and NEMS Δm ≈ 2 Meff Δω ω0 where Meff is the effective vibratory mass of the resonator, and ω0 is the resonance frequency of the device [9.11]. The mass sensitivity of NEMS devices with micromachined cantilevers can be as small as a single small molecule (in the range of a single Dalton). In a device such as that shown in Fig. 9.7, a liquid medium, which contains molecules that dock to a layer of receptor molecules attached to one side of the cantilever, is injected into the device. Sensitizing an array of cantilevers with different receptor allows docking of different substances in the same solution [9.115]. Hybridization can be done with short strands of single-stranded DNA and proteins known a) to recognize antibodies. When docking occurs, the increase in the molecular packing density leads to surface stress, causing bending (10–20 nm of deflection). This deflection can be measured by a laser beam reflected off of the end of the cantilever [9.115]. Alternatively, simple geometric interference by interdigitated cantilevers that act as diffraction gratings can be used to provide output of a binding event [9.120]. Thermomechanical Data Storage Device Much of the drive to nanometer-scale devices originates in the desire for higher density and faster computational devices. Magnetic data storage has been pushed into the nanoscale regime, but limitations have prompted the development of alternative methods for data storage such as the NEMS device known as the Millipede, developed by IBM. The Millipede, or scanning probe array memory device, is an array of individually addressable scanning probe tips (similar to atomic force microscope probe tips) that makes precisely positioned indentations in a polymer thin film. The Millipede is scanned to address a large area for data storage. The indentations are bits of digital information. A polymer thin film (50 nm thick) of polymethyl methacrylate (PMMA) is used for write, read, erase, and rewrite operations. Each individual bit is a nanoscale feature, which allows the Millipede to extend storage density to the Tbit/in2 range with a bit size of 30–40 nm [9.123]. The device uses multiple cantilever probe tips equipped with integrated heaters which allow for data transfer rates of up to a few Mb/s [9.124]. Multiplex driver 2-D cantilever array chip b) Highly doped silicon cantilever leg Nickel bridge x Metal 1 (Gold) Metal 2 (Nickel) z3 y Polymer storage media on x/y/z scanner Low-doped Schottky diode area z1 z2 Stress-controlled nitride Silicon cantilever Heater platform Fig. 9.8 (a) Schematic illustration of the millipede device, (b) with a detail of one cantilever cell (reprinted with permission, Despont [9.122]) 223 Part A 9.9 cently micromachined cantilevers have been used to interact and probe material at the molecular level. Devices employing these micromachined cantilevers can be dynamic, which are sensitive to mass changes down to 10–21 g (the single molecule level), or static, which are sensitive to surface stress changes in the low mN/m range (changes in Gibbs free energy caused by binding site-analyte interactions) [9.121]. In this case adhesion is required between the device and the material to be detected. In a functionalized cantilever array device produced to measure biomechanical forces created by DNA hybridization or receptor–ligand binding, detection of the mass change is accomplished by measuring a shift in resonant frequency. The responsiveness of the device to a change in mass is given by the expression: 9.9 Mechanics Issues in MEMS/NEMS 224 Part A Solid Mechanics Topics Part A 9.10 Substrate Polymer Cantilever Write current Pit: 25nm deep 40nm wide (maximum) Erasure current Sensing current Inscribed pins Data stream 0 0 1 1 1 1 1 0 1 0 0 0 1 0 1 0 1 0 0 1 0 1 1 1 0 1 Output signal 0 The Millipede device is a massively parallel structure with a large array of thousands of probe tips (100 cantilevers/mm2 ), each of which is able to address a region of the substrate where it produces indentations for use as data storage bits (Fig. 9.8) [9.125]. As illustrated in Fig. 9.9, the probes writes a bit by heating and a mechanical force applied between a can- Fig. 9.9 Schematic of the writing, erasing, and reading op- erations in the Millipede device. Data are mechanically stored in pits on a surface (reprinted with permission, Vettiger [9.127]) tilever tip and a polymer film. Erasure of a bit is also conducted with heating by placing a small pit just adjacent to the bit to be erased or using the spring back of the polymer when a hot tip is inserted into a pit. Reading is also enabled by heat transfer since the sensing relies on a thermomechanical sensor that exploits temperature-dependent resistance [9.126]. The change in temperature of a continuously heated resistor is monitored while the tip is scanned over the film and relies on the change in resistance that occurs when a tip moves into a bit [9.123]. Scanning x, y manipulation is conducted magnetically with the entire array at once. The data storage substrate is suspended above the cantilever array with leaf springs which enables the nanometer-scale scanning tolerances required. The cantilevers are precisely curved using residual stress control of a silicon nitride layer in order to minimize the distance between the heating platform of the cantilever and the polymer film while maximizing the distance between the cantilever array and the film substrate to ensure that only the tips come into contact [9.123]. Fabrication details are given in Despont [9.125]. Thermal expansion is a major hurdle for this device since a shift of ≈ 30 nm can cause misalignment of the data storage substrate and the cantilever array. A 10 nm tip position accuracy of a 3 mm × 3 mm silicon area requires that temperature of the device be controlled to 1 ◦ C using several sensors and heater elements [9.123]. Tip wear due to contact between the tip and the underlying silicon substrate is an issue for device reliability. Additionally, the PMMA is prone to charring at the temperatures necessary for device operation (around 350 ◦ C) [9.128] so new polymeric formulations had to be developed to minimize this problem. The feasibility of using thin-film NiTi shape-memory alloy (SMA) for thermomechanical data storage as an alternative to the polymer thin film has also been shown [9.129]. 9.10 Conclusion The sensors, actuators, and passive structures developed as MEMS and NEMS devices require a highly interdisciplinary approach to their analysis, design, de- velopment, and fabrication. Experimental mechanics plays a critical role in design development, materials selection, prediction of allowable operating lim- A Brief Introduction to MEMS and NEMS MEMS/NEMS testing must be further developed. This chapter has provided a brief review of the fabrication processes and materials commonly used and experimental mechanics as it is applied to MEMS and NEMS. References 9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.8 9.9 9.10 9.11 9.12 9.13 9.14 9.15 9.16 N. Maluf: An Introduction to Microelectromechanical Systems Engineering (Artech House, Boston 2000) B. Bhushan (Ed.): Springer Handbook of Nanotechnology (Springer, Berlin, Heidelberg 2006) Small Tech Business DirectoryTM Guide, Small Times Media, http://www.smalltechdirectory.com/ directory/directory.asp. Accessed July 20, 2007. J.C. Eloy: The MEMS industry: Current and future trends, OnBoard Technol., 22–24 (2006), http://www.onboard-technology.com/ pdf_settembre2006/090604.pdf (Accessed July 20, 2007) NewswireToday: Nanorobotics and NEMS to Reach $830.4 million by 2011, 05/31/2007, http://www.newswiretoday.com/news/18867/ (Accessed July 20, 2007) B. Ziaie, A. Baldi, M. Lei, Y. Gu, R.A. Siegel: Hard and soft micromachining for BioMEMS: review of techniques and examples of applications in microfluidics and drug delivery, Adv. Drug Deliv. Rev. 56, 145–172 (2004) S. Beeby, G. Ensell, M. Kraft, N. White: MEMS Mechanical Sensors (Artech House, Boston 2004) H.G. Craighead: Nanoelectromechanical systems, Science 290, 1532–1536 (2000) M. Koch, A.G.R. Evans, A. Brunnschweiler: Microfluidic Technology and Applications (Research Studies, Baldock 2000) G.T.A. Kovacs: Micromachined Transducers – Sourcebook (McGraw-Hill, New York 1998) K.L. Ekinci: Electromechanical transducers at the nanoscale: Actuation and sensing of motion in nanoelectromechanical systems (NEMS), Small 1(8-9), 786–797 (2005) K.L. Ekinci, M.L. Roukes: Nanoelectromechanical systems, Rev. Sci. Instrum. 76(061101), 1–12 (2005) P. Gammel, G. Fischer, J. Bouchaud: RF MEMS and NEMS technology, devices, and applications, Bell Labs Tech. J. 10(3), 29–59 (2005) W.A. Goddard III, D.W. Brenner, S.E. Lyshevski, G.J. Iafrate (Eds.): Handbook of Nanoscience, Engineering and Technology (CRC, Boca Raton 2002) S.E. Lyshevski: Nano- and Microelectromechanical Systems: Fundamentals of Nano- and Microengineering (CRC, Boca Raton 2000) T.S. Tighe, J.M. Worlock, M.L. Roukes: Direct thermal conductance measurements on suspended 9.17 9.18 9.19 9.20 9.21 9.22 9.23 9.24 9.25 9.26 9.27 9.28 monocrystalline nanostructures, Appl. Phys. Lett. 70(20), 2687–2689 (1997) R.G. Knobel, A.N. Cleland: Nanometre-scale displacement sensing using a single electron transistor, Nature 424(6946), 291–293 (2003) S.D. Senturia: Microsystem Design (Kluwer Academic, Boston 2001) SEM: Society for Experimental Mechanics, Bethel, 2007 SEM Annual Conference Proceedings, Springfield (2007). MEMS and Nanotechnology Clearinghouse: Material Index, http://www.memsnet.org/material/ (Accessed July 20, 2007) P. Lange, M. Kirsten, W. Riethmuller, B. Wenk, G. Zwicker, J.R. Morante, F. Ericson, J.A. Schweitz: Thisck polycrystalline silicon for surface micromechanical applications: deposition, structuring and mechanical characterization, 8th International conference on solid-state sensors and actuators (Transducers ’95), Stockholm (1995) pp. 202–205 M. Madou: Fundamentals of Microfabrication (CRC, Boca Raton 1997) B. Diem, M.T. Delaye, F. Michel, S. Renard, G. Delapierre: SOI(SIMOX) as a substrate for surface micromachining of single crystalline silicon sensors and actuators, 7th International Conference on Solid-State Sensors and Actuators (Transducers ’93), Yokohama (1993) pp. 233–236 J.M. Noworolski, E. Klaassen, J. Logan, K. Petersen, N. Maluf: Fabricationof SOI wafers with buried cavities using silicon fusion bonding ans electrochemical etchback, 8-th International conference on solid-state sensors and actuators (Transducers ’95), Stockholm (1995) pp. 71–74 T. Ikeda: Fundamentals of Piezoelectricity (Oxford Univ., New York 1990) P. Gluche, A. Floter, S. Ertl, H.J. Fecht: Commercial applications of diamond-based nano- and microtechnolgy. In: The Nano-Micro Interface, ed. by H.-J. Fecht, M. Werner (Wiley-VCH, Weinheim 2004) pp. 247–262 S. Cho, I. Chasiotis, T.A. Friedmann, J.P. Sullivan: Young’s modulus, Poisson’s ratio and failure properties of tetrahedral amorphous diamond-like carbon for MEMS devices, J. Micromech. Microeng. 15, 728–735 (2005) A.R. Krauss, O. Auciello, D.M. Gruen, A. Jayatissa, A. Sumant, J. Tucek, D.C. Mancini, N. Moldovan, 225 Part A 9 its, device characterization, process validation, and quality control inspection. Commercial devices exist, and research in the area of MEMS/NEMS is extremely active, but many challenges remain. Advanced materials must be well characterized and References 226 Part A Solid Mechanics Topics Part A 9 9.29 9.30 9.31 9.32 9.33 9.34 9.35 9.36 9.37 9.38 9.39 9.40 9.41 9.42 9.43 9.44 9.45 9.46 A. Erdemir, D. Ersoy, M.N. Gardos, H.G. Busmann, E.M. Meyer, M.Q. Ding: Ultrananocrystalline diamond thin films for MEMS and moving mechanical assembly devices, Diamond Relat. Mater. 10(11), 1952–1961 (2001) D. Gruen: Nanocrystalline diamond films, Annu. Rev. Mater. Sci. 29, 211–259 (1999) D.M. Gruen, S. Liu, A.R. Krauss, J. Luo, X. Pan: Fullerenes as precursors for diamond film growth without hydrogen or oxygen additions, Appl. Phys. Lett. 64, 1502–1504 (1994) D.M. Gruen, S. Liu, A.R. Krauss, X. Pan: Buckyball microwave plasmas – fragmentation and diamondfilm growth, J. Appl. Phys. 75, 1758–1763 (1994) S.J. Clarson, J.A. Semlyen (Eds.): Siloxane Polymers (Prentice Hall, Englewood Cliffs 1993) C. Liu: Foundations of MEMS (Pearson Education, Upper Saddle River 2006) K. Otsuka, C.M. Wayman (Eds.): Shape Memory Materials (Cambridge Univ. Press, Cambridge 1998) H.-J. Fecht, M. Werner: The Nano-Micro Interface (Wiley-VCH, Weinheim 2004) M.J. Yacamaìn, T. Tsakalakos, B.H. Kear (Eds.): Proceedings of the First International Conference on Nanostructured Materials, Cancun, Nanostruct. Mater. 3(1/6), 22–26 (1993) V.E. Borisenko, S. Ossicini: What is What in the Nanoworld (Wiley-VCH, Weinheim 2004) B.D. Busbee, S.O. Obare, C.J. Murphy: An improved synthesis of high-aspect-ratio gold nanorods, Adv. Mater. 15(5), 414–416 (2003) F. Favier, E.C. Walter, M.P. Zach, T. Benter, R.M. Penner: Hydrogen sensors and switches from electrodeposited palladium mesowire arrays, Science 293(5538), 2227–2231 (2001) M.H. Huang, S. Mao, H. Feick, H.Q. Yan, Y.Y. Wu, H. Kind, E. Webber, R. Russo, P.D. Yang: Roomtemperature ultraviolet nanowire nanolasers, Science 292(5523), 1897–1899 (2001) R.M. Penner, C.R. Martin: Preparation and electrochemical characterization of ultramicroelectrode ensembles, Anal. Chem. 59(21), 2625–2630 (1987) J.C. Hulteen, C.R. Martin: A general template-based method for the preparation of nanomaterials, J. Mater. Chem. 7(7), 1075–1087 (1997) S. K. Chakarvarti: Science and art of synthesis and crafting of nano/microstructures and devices using ion-crafted templates: A review, Proc. SPIE 61702(61720G), 1–9 (2006) M.S. Dresselhaus, Y.-M. Lin, O. Rabin, M.R. Black, G. Dresselhaus: Nanowires. In: Springer Handbook of Nanotechnology, ed. by B. Bhushan (Springer, Berlin, Heidelberg 2007) W.R. DeVries: Analysis of Material Removal Processes (Springer, New York 1992) S. Kalpajian: Manufacturing Processes for Engineering Materials (Addison-Wesley, Reading 1984) 9.47 9.48 9.49 9.50 9.51 9.52 9.53 9.54 9.55 9.56 9.57 9.58 9.59 9.60 9.61 9.62 N. Taniguchi: Current status in, and future trends of, ultraprecision machining and ultrafine materials processing, Ann. CIRP 32, S573–S582 (1983) C. Evans: Precision Engineering (Cranfield, Cranfield 1989) J.M. Bustillo, R.T. Howe, R.S. Muller: Surface micromachining for microelectromechanical systems, Proc. IEEE 86(8), 1559–1561 (1998) M. Gentili, C. Giovannella, S. Selci (Eds.): Nanolithography (Kluwer Academic, Boston 1994) K.R. Williams, R.S. Muller: Etch rates for micromachining processing, J. Microelectromech. Syst. 5(4), 256–269 (1996) B. Li, X. Tang, H. Xie, X. Zhang: Strain analysis in MEMS/NEMS structures and devices by using focused ion beam system, Sens. Actuator A 111(1), 57–62 (2004) H. Xie, Z. Liu, H. Shang, Q. Xue, J. Jia, D. Fang: Micro/nano grating and its application to moire measurement, Proc. SPIE 5852, 740–744 (2005) Sandia National Laboratories, Micromachines: SAMPLESTM Program, http://mems.sandia.gov/ samples/index.html (Accessed July 20, 2007) H. Guckel: High-aspect ratio micromachining via deep x-ray lithography, Proc. IEEE 86(8), 1586–1593 (1998) E.W. Becker, W. Ehrfeld, P. Hagmann, A. Maner, D. Munchmeyer: Fabrication of microstructures with high aspect ratios and great structural heights by synchrotron radiation lithography, galvanoforming, and plastic moulding (LIGA process), Microelectron. Eng. 4(1), 35–56 (1986) P. Bley, W. Menz, W. Bacher, K. Feit, M. Harmening, H. Hein, J. Mohr, W.K. Schomburg, W. Stark: Application of the LIGA process in fabrication of three-dimensional mechanical microstructures, 4th International Symposium on MicroProcess Conference Kanazawa (1991) pp. 384–389 J.L. Wilbur, G.M. Whitesides: Self-assembly and self-assembled monolayers in micro- and nanofabrication. In: Nanotechnology, ed. by G. Timp (Springer, New York 1999) pp. 331–369 G.M. Whitesides, J.P. Mathias, C.T. Seto: Molecular self-assembly and nanochemistry: a chemical strategy for synthesis of nanostructures, Science 254, 1312–1318 (1991) G.M. Whitesides: Self-assembling materials, Sci. Am. 273, 146–149 (1995) H. Ringsdorf, B. Schlarb, J. Venzmer: Molecular architecture and function of polymeric oriented systems: Models for the study of organization, surface recognition, and dynamics, Angew. Chem. Int. Ed. Engl. 27, 113–158 (1988) G.M. Whitesides, P.E. Laibinis: Wet chemical approaches to the characterization of organic surfaces. Self-assembled monolayers, wetting, and the A Brief Introduction to MEMS and NEMS 9.64 9.65 9.66 9.67 9.68 9.69 9.70 9.71 9.72 9.73 9.74 9.75 9.76 9.77 9.78 9.79 9.80 9.81 9.82 9.83 9.84 9.85 9.86 9.87 9.88 9.89 9.90 9.91 9.92 9.93 9.94 9.95 9.96 9.97 9.98 C. Gorecki (Ed.): Microsystems Metrology and Inspection, Proc. EuroOpto Ser, Vol. 3825 (The Society of Photo-Optical Instrumentation Engineers, Washington 1999) W.C. Young, R.G. Budynas: Roark’s Formulas for Stress and Strain (McGraw-Hill, New York 2002) ASTM International: ASTM E2245-05, Standard Test Method for Residual Strain Measurements of Thin, Reflecting Films Using an Optical Interferometer (2005) M.G. Allen, M. Mehregany, R.T. Howe, S.D. Senturia: Microfabricated structures for the in-situ measurement of residual stress, Young’s modulus, and ultimate strain of thin films, Appl. Phys. Lett. 51(4), 241–243 (1987) P.M. Osterberg, S.D. Senturia: M-TESt: a test chip for MEMS material property measurements using electrostatically actuated test structures, J. Microelectromech. Syst. 6(2), 107–117 (1997) W.D. Nix: Mechanical properties of thin films, Metall. Trans. 20A, 2217–2245 (1989) K.E. Petersen: Dynamic micromechanics on silicon: Techniques and devices, IEEE Trans. Electron. Dev. 25, 1241–1250 (1978) L.M. Zhang, D. Uttamchandani, B. Culshaw: Measurement of the mechanical properties of silicon resonators, Sens. Actuator A 29, 79–84 (1991) B. Li, H. Xie, B. Xu, R. Geer, J. Castracane: Investigation of strain in microstructures by a novel moire method, J. Microelectromech. Syst. 11(6), 829–36 (2002) D. Vogel, D. Lieske, A. Gollhardt, J. Keller, N. Sabate, J.R. Morante, B. Michel: FIB based measurements for material characterization on MEMS structures, Proc. SPIE 5766, 60–69 (2005) A.A. Ayon, X. Zhang, K.T. Turner, D. Choi, B. Miller, S.F. Nagle, S.M. Spearing: Characterization of silicon wafer bonding for power MEMS applications, Sens. Actuator A 103, 1–8 (2003) P.G. Charalambides, H.C. Cao, J. Lund, A.G. Evans: Developments of a test method for measuring the mixed mode fracture resistance of bimaterial interfaces, Mech. Mater. 8(4), 269–283 (1990) K.T. Turner, M.D. Thouless, S.M. Spearing: Mechanics of wafer bonding: Effect of clamping, J. Appl. Phys. 95(1), 349–355 (2004) D.S. Campbell: Mechanical properties of thin films. In: Handbook of Thin Film Technology, ed. by L.I. Maissel, R. Glang (McGraw-Hill, New York 1970) C.H. Mastrangelo, C.H. Hsu: Mechanical stability and adhesion of microstructures under capillary forces. I. Basic theory, J. Microelectromech. Syst. 2(1), 33–43 (1993) M.P. de Boer, T.M. Mayer: Tribology of MEMS, MRS Bull. 26(4), 302–304 (2001) J.R. Martin, Y. Zhao: Micromachined device packaged to reduce stiction, US Patent 5694740 (1997) 227 Part A 9 9.63 physical-organic chemistry of the solid-liquid interface, Langmuir 6(1), 87–96 (1990) L.H. Dubois, R.G. Nuzzo: Synthesis, structure, and properties of model organic-surfaces, Annu. Rev. Phys. Chem. 43, 437–463 (1992) A. Ulman: Surface-absorption of monolayers, Mater. Res. Soc. Bull. 20(6), 46–51 (1995) A.R. Bishop, R.G. Nuzzo: Self-assembled monolayers: Recent developments and applications, Curr. Opin. Coll. Interface Sci. 1(1), 127–136 (1996) Y. Xia, G.M. Whitesides: Soft lithography, Annu. Rev. Mater. Sci 28, 153–184 (1998) E. Delamarche, H. Schmid, H.A. Biebuyck, B. Michel: Stability of molded polydimethylsiloxane microstructures, Adv. Mater. 9(9), 741–746 (1997) J.A. Preece, J.F. Stoddart: Concept transfer from biology to materials, Nanobiology 3, 149–166 (1994) J.A. Stroscio, D.M. Eigler: Atomic and molecular manipulation with the scanning tunneling microscope, Science 254, 1319–1326 (1991) D. Eigler: From the bottom up: building things with atoms. In: Nanotechnology, ed. by G. Timp (Springer, New York 1999) pp. 331–369 R.D. Piner, J. Zhu, F. Xu, S. Hong, C.A. Mirkin: “Dippen” nanolithography, Science 283, 661–663 (1999) D. Bullen, X. Wang, J. Zou, S.-W. Chung, C.A. Mirkin, C. Liu: Design, fabrication, and characterization of thermally actuated probe arrays for Dip Pen nanolithography, J. Microelectromech. Syst. 13(4), 594–602 (2004) Q. Tang, S.-Q. Shi, L. Zhou: Nanofabrication with atomic force microscopy source, J. Nanosci. Nanotechnol. 4(8), 948–963 (2004) W.N. Sharpe, K.T. Turner, R.L. Edwards: Tensile testing of polysilicon, Exp. Mech. 39(3), 162–70 (1999) S.F. Bart, T.A. Lober, R.T. Howe, M.F. Schlecht: Design considerations for micromachined electric actuators, Sens. Actuator 14, 269–292 (1988) D.W. Carr, L. Sekaric, H.G. Craighead: Measurement of nanomechanical resonant structures in singlecrystal silicon, J. Vac. Sci. Technol. B 16(6), 3821–3824 (1998) W.S. Trimmer (Ed.): Micromechanics and MEMS (IEEE, New York 1997) W.N. Sharpe: Mechanical properties of MEMS materials. In: MEMS handbook, ed. by M. Gad-el-Hak (CRC, Boca Raton 2002) V.T. Srikar, S.M. Spearing: A critical review of microscale mechanical testing methods used in the design of microelectromechanical systems, Exp. Mech. 43(3), 238–247 (2003) M.A. Haque, M.T.A. Saif: A review of MEMS-based microscale and nanoscale tensile and bending testing, Exp. Mech. 43(3), 248–255 (2003) T. Yi, C.J. Kim: Measurement of mechanical properties of MEMS materials, Meas. Sci. Tech. 10, 706–716 (1999) References 228 Part A Solid Mechanics Topics Part A 9 9.99 9.100 9.101 9.102 9.103 9.104 9.105 9.106 9.107 9.108 9.109 9.110 9.111 9.112 9.113 9.114 R.W. Carpick, M. Salmeron: Scratching the surface: Fundamanetal investigations of tribology with atomic force microscopy, Chem. Rev. 97, 1163–1194 (1997) B. Bhushan: Introduction to Tribology (Wiley, New York 2002) M.P. de Boer, J.A. Knapp, T.M. Mayer, T.A. Michalske: The role of interfacial properties on MEMS performance and reliability. In: Microsystems Metrology and Inspection, Proc. EuroOpto Ser., Vol. 3825, ed. by C. Gorecki (Society of Photo-Optical, Washington 1999) pp. 2–15 M.P. de Boer, N.F. Smith, N.D. Masters, M.B. Sinclair, E.J. Pryputniewicz: Integrated platform for testing MEMS mechanical properties at the wafer scale by the IMaP methodology, ASTM Spec. Tech. Pub. 1413, 85–95 (2001) J.G. Santiago, S.T. Wereley, C.D. Meinhart, D.J. Beebe, R.J. Adrian: A particle image velocimetry system for microfluidics, Exp. Fluids 25, 316–319 (1998) C.D. Meinhart, S.T. Wereley, J.G. Santiago: PIV measurements of a microchannel flow, Exp. Fluids 27, 414–419 (1999) C.D. Meinhart, H. Zhang: The flow structure inside a microfabricated inkjet printer head, J. Microelectromech. Syst. 9, 67–75 (2000) S. Wereley, C.D. Meinhart, S. Stone, V. Hohreiter, J. Chung: Experimental microfluidics toolbox for MEMS characterization, Proc. SPIE 4558, 124–132 (2001) E.T. Enikov: Structures and materials. In: Mechatronics Handbook, ed. by R.H. Bichop (CRC, Boca Raton 2002) S. J. Walker, D. J. Nagel: Optics and MEMS, Tech. Rep. NRL/MR/6336-99-7975 (Naval Research Lab, Washington 1999) L.J. Hornbeck: Frame addressed spatial light modulator, US Patent 4615595 (1986) P.F. Van Kessel, L.J. Hornbeck, R.E. Meier, M.R. Douglass: A MEMS-based projection display, Proc. IEEE 86(8), 1687–1704 (1998) L.J. Hornbeck: Digital light processing for highbrightness, high-resolution applications, Proc. SPIE 3013, 27–40 (1997) S.A. Henck: Lubrication of Digital Micromirror DevicesTM , Tribol. Lett. 3(3), 239–247 (1997) K. Hehr, S. Kurth, J. Mehner, W. Dotzel, T. Gessner: Investigation of heat transfer in micromirrors. In: Microsystems Metrology and Inspection, Proc. EuroOpto Ser., Vol. 3825, ed. by C. Gorecki (The Society of Photo-Optical Instrumentation Engineers, Washington 1999) pp. 24– 33 R. E. Meier: DMDTM pixel mechanics simulation, TI Tech. J., 64–74 (1998) Special issue on DLP—DMD Manufacturing and Design Challenges 9.115 J. Fritz, M.K. Baller, H.P. Lang, H.P. Rothuizen, P. Vettiger, E. Meyer, H.-J. Guntherodt, C. Gerber, J.K. Gimzewski: Translating biomolecular recognition into nanomechanics, Science 288(5464), 316–318 (2000) 9.116 S.C. Kuo, M.P. Sheetz: Force of single kinesin molecules measured with optical tweezers, Science 260(5105), 232–234 (1993) 9.117 E.L. Florin, V.T. Moy, H.E. Gaub: Adhesion forces between individual ligand-receptor pairs, Science 264(5157), 415–417 (1994) 9.118 G.U. Lee, L.A. Chrisey, R.J. Colton: Direct measurement of the forces between complementary strands of DNA, Science 266(5186), 771–773 (1994) 9.119 S.B. Smith, L. Finzi, C. Bustamante: Force of single kinesin molecules measured with optical tweezers, Science 258(5085), 1122–1126 (1992) 9.120 T. Thundat, G. Wu, M. Mao, A. Majumdar: Biochemoopto-mechanical (BioCOM) chip for chemical and biomolecular detection, 1st NASA and NCI Workshop on Sensors for Bio-Molecular Signatures (1999) 9.121 M. Sepaniak, P. Datskos, N. Lavrik, C. Tipple: Microcantilever transducers: a new approach in sensor technology, Anal. Chem. 74(21), 568A–575A (2002) 9.122 M. Despont, J. Brugger, U. Drechsler, U. Dürig, W. Häberle, M.I. Lutwyche, H.E. Rothuizen, R. Stutz, R. Widmer, G.K. Binnig, H. Rohrer, P. Vettiger: VLSI-NEMS chip for AFM data storage, Technical Digest, 12th IEEE International Micro Electro Mechanical Systems Conference (MEMS ’99), Orlando (1999) pp. 564–569 9.123 P. Vettiger, M. Despont, U. Drechsler, U. Dürig, W. Häberle, M.I. Lutwyche, H.E. Rothuizen, R. Stutz, R. Widmer, G.K. Binnig: The “Millipede” – more than one thousand tips for future AFM data storage, IBM J. Res. Dev. 44(3), 323–340 (2000) 9.124 P. Vettiger, B. Cross, M. Despone, U. Drechsler, U. Durig, B. Botsmann, W. Haberle, M.A. Lantz, H.E. Rothuizen, R. Stutz, G.K. Binnig: The “millipede” – nanotechnology entering data storage, IEEE Trans. Nanotechnol. 1(1), 39–55 (2002) 9.125 M. Despont, U. Drechsler, R. Yu, H.B. Pogge, P. Vettiger: Wafer-scale microdevice transfer/interconnect: its application in an AFM-based data-storage system, J. Microelectromech. Syst. 13(6), 895–901 (2004) 9.126 G.K. Binnig, J. Brugger, W. Häberle, P. Vettiger: Investigation and/or manipulation device, US Patent 6249747 (2001) 9.127 P. Vettiger, G. Binnig: The nanodrive project, Sci. Am. 288(1), 47–53 (2003) 9.128 B.J. Holland, J.N. Hay: The thermal degradation of poly(vinyl alcohol), Polymer 42(16), 6775–6783 (2001) 9.129 G.A. Shaw, J.S. Trethewey, A.D. Johnson, W.J. Drugan, W.C. Crone: Thermomechanical high-density data storage in a metallic material via the shapememory effect, Adv. Mater. 17, 1123–1127 (2005)
© Copyright 2025 Paperzz