Fabrication processes that produce chips for electronic-only functions use repetitive masking-deposition-removal processes to generate multiple layers of thin patterned materials functioning together as interconnected transistors, conductors and other circuit elements. The circuit layers, fabricated directly on silicon wafers, are separated into individual chips that are tested, combined with other circuit elements, and then assembled into packages. The drive for greater levels of integration in chip-based electronic circuits has pushed the development of lithographic fabrication processes to impressive levels, especially with regard to the achievement of finer resolution, higher speed, better efficiency, and tighter process control for improvement in yield.
Developments in the areas of packaging, assembly, and testing, while also impressive, have been less expansive. They have been focused on placing one or many chips in rigid, hermetically-sealed housings with focus primarily on economy and thermal control. Sealed package lead-throughs provide only electrical connection between the chip and the outside world for assembly into larger systems and after-packaging testing. Typical electronic packages are shown in Figure 6.1.
New systems, planned for the near future, will require subsystems that include not only electronic functions but others, such as mechanical, thermal, chemical, and optical. The new systems will include moving parts and require the passage of other-than-electric information through package walls. New systems will also include other requirements, such as nonplanar geometries, operation in more harsh environments, and other nonstandard characteristics.
Figure 6.1. A variety of typical closed, rigid electronic packages: dual in-line package (DIP), pin grid array (PGA), surface mounts, direct die mounting (from H.B. Bakoglu, Circuits, Interconnections and Packaging for VLSI, Reading, MA: Addison Wesley, 1990).
The micromechanics area has evolved, mainly over the past decade, with a focus on creating thicker, void-containing microassemblies which, in addition to electronics, include moving parts and utilize other physical processes, such as mechanical, optical, fluid, thermal, chemical, nuclear, and so on. Microassemblies targeted for development include subelements such as: (1) rigid shapes that form open or closed channels, coils to produce magnetic fields, orifices, large area capacitors; (2) structures that move by flexing, such as diaphragms, flexures, and springs; and (3) assemblies that permit relative motion (sliding) between elements such as links, valves, hinges, gears, and sliders. Figures in Chapters 2 and 5 of this report provide excellent examples of sacrificial layer etch-based techniques used to form complex geometry structures.
Microassemblies are interesting because they make possible the creation of microsubsystems which can be classified as sensors, actuators, and subsystems (SAS). SAS can be arranged to form components that function to enhance the performance, reliability, and economy of the larger systems in which they will function. Of particular importance will be the generation of SAS with high performance and reliability, small size, efficient intercommunication capability, local intelligence, and low cost. Present SAS development projects are investigating large numbers of materials and processes in which reciprocal energy-movement relationships exist (energy to movement or movement to energy). Approaches used include phenomena such as: magnetic fields, electrostatic fields, electrostrictive properties, magnetostrictive effects, piezoelectric bimorphs, thermal expansion elements, shape memory alloys, phase transition liquids and solids, impact systems, chemical reactions, and polymeric materials.
Other sections of this report show good examples of micromachines formed by various silicon micromachining processes. Other figures illustrate: (1) the formation of moving elements, such as comb drives or accelerometer masses, suspended by flexures; (2) revolute joints suspending micromotor rotor; (3) actuated microcantilever beams, and others.
MEMS is an emerging field focused on the integration of microelectronics and micromechanics to form systems of intercommunicating sensors, actuators, and subsystems that can be integrated to form components. The components can then function as the interconnected building blocks necessary for realization of systems based on new machine architectures that use larger numbers of sensors and actuators together with distributed subcontrollers.
The design and fabrication of MEMS-based systems is a complex undertaking. New mentalities, processes, and fabrication approaches are required, and even the act of requirements-definition is difficult. For example, just a few characteristics required for the specification of any real sensor include sensitivity, resolution, stability, dynamic range, bandwidth, drift, noise, reliability, ruggedness, addressability, cost, size, weight, absolute or incremental output, update rate, power consumption, local intelligence, output information architecture, life, stiction, backlash, required support systems, and so forth. Another example of just a few characteristics required for the specification of any real actuator include output force or torque, output speed, output impedance (stiffness, damping, inertia), input (voltage, pressure, current, flow, etc.) reversibility, efficiency, size, weight, stiction, backlash, thermal tolerance, reliability, ruggedness, emissions, cost, life, power dissipation, life, leakage, required support systems, and so forth.
The MEMS area is important because MEMS-based approaches can produce products with fundamental advantages in performance, reliability, and economy. MEMS-based systems can: (1) be used in many existing applications such that natural commercial motivations will push development, and (2) generate previously nonexistent systems that address important international priorities. Figure 6.2 shows a good example of a MEMS-based optical system (Texas Instruments), which includes numerous actuated micromirrors systematically driven to generate a color display. The optical display is packaged in a rigid-closed housing with a transparent optical pass-through for light passage to and from mirrors. Assembly at the microlevel is achieved by integrated, hands-off, techniques. Assembly of the chip into the package is achieved via conventional methods. Testing for individual pixel performance is achieved by examining dynamic image quality.
Figure 6.2. A MEMS-based array of actuated micromirrors (Texas Instruments).
Figure 6.3 shows a method of vacuum packaging by glass-silicon anodic bonding, which is being developed by Henmi, Shoji, Yoshimi, and Esashi at Tohoku University. Such methods will be used in the future to permit closed-rigid or closed-flexible packaging to occur at a more integrated process level.
Figure 6.3. A method of integrated vacuum packaging.
Complete systems can be formed by integrating elements from microelectronic, micromachine, and MEMS levels. In many cases the elements function as discriminating subsystems that permit a product to achieve properties not otherwise possible. It should not be forgotten that, in many cases, PAT consists of very important drivers of system reliability, ruggedness, maintainability, and cost. Good examples of complete systems are shown in Figures 6.4, 6.5, 6.6 and 6.7.
Figure 6.4 shows a cross section of a typical, Japanese-produced SLR camera. A typical system includes macromechanics, microelectronics, optics, sensors, very small actuators, seals, many input controls, internal intelligence, electric pass-throughs, and other elements. The modern camera, now typically taken for granted, represents a series of triumphs in both macro- and micro-PAT, in terms of reliability, ruggedness, and cost.
Figures 6.5 and 6.6 show two projects proposed by the Micromachine Center and supported by the Ministry of International Trade and Industry (MITI). The projects focus on "the development of micromachine technology, and to encourage a central role for these machines as they spread into various economic and social sectors." The projects are based on the idea that "micromachine technology must be established and steps taken to promote information exchange in this field." MITI plans that the projects will contribute both to development of domestic Japanese industry and to the international community. Figure 6.5 illustrates the concept for subsystems that can provide distributed, internal power plant maintenance. Figure 6.6 shows the concept for an intraluminal diagnostic and therapeutic system. Both projects are designed to promote developments in microelectronics, micromechanics, and MEMS. They will also require broad-based developments in PAT if any success is to be achieved. See Chapter 7 and the MITI site report for further information.
Figure 6.4. Cross section of a typical Japanese-produced SLR camera.
Figure 6.5. Concept for an advanced maintenance system for a power plant (MITI).
Figure 6.6. Concept for an intraluminal diagnostic and therapeutic system (MITI).
Figure 6.7 shows a duodenofiberscope currently marketed by Olympus. The system is a good example of current technology in systems for the execution of less invasive medical procedures. The scope shown is a commercial system (10 mm diameter with internal cannula pathways of 2.2 mm diameter). Emerging systems, not shown to the JTEC team, will achieve smaller sizes (1 to 2 mm diameters), which will intensify the need for MEMS-based systems for the tip. In general, scopes or catheters are introduced at an orifice and translated to a site in order to introduce or withdraw fluids, take biopsies, sense, view, treat, cut, expand, cauterize or otherwise interact with small, remote, anatomic features or lesions within a body. Although not themselves microsystems, endoscopes and catheters provide an excellent base for the application of microsystems on their tips. Here again, microsystems will provide discriminating features, but PAT-related issues, both macro and micro, will determine achievable levels of system reliability, ruggedness, and cost.
Figure 6.7. Duodenofiberscope currently marketed by Olympus.