MEMS metallic mechanical response performance testing evaluates the elastic, plastic, fatigue, and creep behavior of metallic thin films and microstructures used in micro-electro-mechanical systems. Because metal properties at the microscale differ significantly from bulk behavior due to size effects, grain structure, and process-induced residual stress, dedicated micro-scale characterization is essential. This article covers the sample types, testing items, methods, standards, and application scenarios for MEMS metal mechanical testing services.
Technical Overview
Micro-Electro-Mechanical Systems (MEMS) devices rely heavily on metallic materials such as gold, aluminum, copper, nickel, platinum, titanium, and tungsten for structural beams, membranes, springs, contacts, and interconnects. These metals are typically deposited as thin films with thicknesses ranging from tens of nanometers to several hundred micrometers through sputtering, evaporation, electroplating, or LIGA processes. At these dimensions, mechanical behavior is governed by size effects, columnar grain structures, surface oxide layers, and residual stresses introduced during deposition and patterning. Consequently, bulk material data sheets cannot be reliably extrapolated to MEMS design.
MEMS metallic mechanical response performance testing provides quantitative characterization of elastic modulus, strength, hardness, residual stress, fatigue life, creep resistance, and fracture behavior using micro-scale specimen geometries and specialized load frames. The resulting data supports design validation, process optimization, reliability assessment, and failure analysis for devices such as RF switches, resonators, micro-mirrors, accelerometers, and micro-relays. Testing is typically performed in accordance with ISO 14577 for instrumented indentation and adapted micro-tensile procedures, with environmental options including elevated temperature, humidity, and cyclic loading.
Testing Samples
Metallic Thin Films on Substrates
Sputtered or evaporated films of aluminum, gold, copper, nickel, titanium, tungsten, and alloys deposited on silicon, glass, or quartz wafers. Film thicknesses typically range from 100 nm to 10 µm. These samples are suitable for nanoindentation hardness and modulus measurement as well as wafer-curvature residual stress determination.
Freestanding Micro-Tensile Specimens
Dog-bone shaped specimens fabricated by surface micromachining and released by sacrificial layer etching, with gauge sections of 10–1000 µm in length and 1–50 µm in thickness. Electroplated nickel and nickel-iron specimens from LIGA or UV-LIGA processes fall into this category. These specimens enable direct uniaxial stress-strain measurement of the true film properties without substrate influence.
Patterned MEMS Microstructures
Cantilever beams, fixed-fixed bridges, clamped membranes, comb-drive resonators, folded flexures, and spiral springs made of metallic materials. These on-chip test structures are used for residual stress extraction, fatigue testing, and creep evaluation under actuated or external loading.
Bonding Layers, Interconnects, and 3D Micro-Parts
Metallic bonding layers such as Au-Au or Cu-Cu thermocompression bonds, through-silicon via (TSV) copper liners, sintered silver die-attach layers, and three-dimensional electroformed micro-components. These samples require specialized miniaturized test configurations to evaluate interfacial and structural integrity.
Testing Items
- Elastic modulus (Young's modulus) and effective stiffness of metal films and microstructures
- Poisson's ratio determination
- Yield strength and 0.2% offset proof stress
- Ultimate tensile strength and fracture strength
- Elongation to failure and uniform elongation
- Hardness (nanoindentation and micro-hardness)
- Residual stress (wafer curvature and structure deflection methods)
- Strain rate sensitivity and loading-rate dependent response
- High-cycle and low-cycle fatigue life, S-N curve generation
- Creep behavior, stress relaxation, and time-dependent deformation
- Fracture toughness and crack growth resistance
- Adhesion strength and interfacial delamination resistance
- Temperature-dependent mechanical response (room temperature to 300°C or higher)
- Cyclic contact resistance stability for MEMS switches and relays
Testing Methods
Nanoindentation
Instrumented indentation with a diamond Berkovich tip measures hardness and elastic modulus continuously as a function of penetration depth. The Oliver-Pharr method per ISO 14577 is applied, with substrate-effect corrections when indentation depth exceeds roughly 10% of film thickness. Continuous stiffness measurement (CSM) enables depth-profiling of graded films and dynamic modulus extraction.
Micro-Tensile Testing
A dedicated micro-force testing system with piezoelectric or electromagnetic actuation, load resolution at the micronewton level, and laser or image-based displacement measurement performs uniaxial tension on freestanding dog-bone specimens. Full stress-strain curves deliver modulus, yield strength, ultimate strength, and elongation. Strain rate and temperature can be varied to evaluate rate sensitivity and thermal effects.
Bulge and Membrane Deflection Testing
Pressure is applied to a clamped circular or rectangular metal membrane while central deflection is monitored by interferometry or laser displacement sensing. Fitting the pressure-deflection response yields plane-strain modulus and residual stress of the film, which is particularly valuable for films that are difficult to release as freestanding tensile bars.
Beam Bending and Residual Stress Extraction
Cantilever deflection and buckling of fixed-fixed beams are measured optically to extract gradient and average residual stress using established analytical models. Rotating-blade, buckling-array, and ring-and-beam test structures may also be employed for rapid on-wafer stress mapping.
Resonant Fatigue and Dynamic Testing
Electrostatic or electromagnetic excitation drives metal microstructures at resonance for millions to billions of cycles, with amplitude monitoring to detect stiffness degradation and crack initiation. This approach reproduces the high-frequency vibration environment of RF switches and resonators and supports S-N curve construction.
Creep and Stress Relaxation Testing
Constant-load or constant-displacement hold tests on micro-specimens and spring structures, performed at room and elevated temperatures, quantify creep strain rate, stress exponent, and relaxation behavior critical for long-term dimensional stability of metal MEMS components.
Wafer Curvature and X-Ray Diffraction Stress Analysis
The wafer curvature method measures film stress before and after thermal cycling, while X-ray diffraction using the sin-squared-psi technique provides phase-specific stress and texture information. These methods complement micro-structural test approaches for process monitoring.
Applications
- RF MEMS switches and micro-relays: fatigue life and contact reliability of gold and alloy contact metals
- MEMS mirrors and optical scanners: fatigue and creep resistance of aluminum and titanium hinge structures
- Resonators and gyroscopes: elastic modulus consistency and long-term frequency stability of metal electrodes
- Pressure sensors and microphones: residual stress control and burst strength of metal diaphragms
- Inkjet printheads: thermomechanical response of heated metal layers under cyclic thermal loading
- Advanced packaging: mechanical integrity of TSV copper, bonding layers, and sintered silver die attach
- Wearable and biomedical MEMS: corrosion-fatigue and biocompatibility-related mechanical evaluation
- Process development: comparing sputtering, evaporation, and electroplating recipes through mechanical benchmarking
- Failure analysis: verifying whether field failures originate from fatigue cracks, creep deformation, stress migration, or delamination
FAQ
Why cannot bulk metal data be used for MEMS design? Metallic thin films exhibit strong size effects: grain sizes comparable to film thickness, high dislocation densities, surface oxide influence, and deposition-induced residual stresses all cause yield strength and fatigue behavior to deviate substantially from bulk values. Micro-scale testing on the actual film or structure is therefore required.
What specimen dimensions can be tested? Freestanding tensile specimens are typically tested at gauge lengths from 50 µm to 2 mm and thicknesses from 0.5 µm to 200 µm. Nanoindentation can be performed on films as thin as approximately 50 nm with appropriate substrate correction.
Which standards apply to MEMS metal mechanical testing? ISO 14577 governs instrumented indentation, while micro-tensile testing generally follows adapted ASTM E8/E8M procedures with miniaturized geometry. SEMI and IEEE guidelines provide supplementary direction for MEMS material characterization and reliability qualification.
How is residual stress in metal films measured? Common approaches include wafer curvature measurement before and after deposition, deflection analysis of released cantilevers and buckled beams, bulge testing, and X-ray diffraction sin-squared-psi analysis. Combining methods separates average stress from through-thickness gradients.
What environmental conditions can be simulated? Testing can be conducted at temperatures from room temperature up to 300°C or higher, under controlled humidity, and under vacuum, enabling evaluation of creep, thermal cycling effects, and humidity-assisted fatigue relevant to real operating environments.
How many samples are required for statistical confidence? Because micro-scale specimens show greater scatter than bulk material, a minimum of five to ten valid specimens per condition is recommended for tensile data, and larger populations or longer runout populations are advised for fatigue S-N characterization.
What is the typical turnaround time? Standard nanoindentation and residual stress evaluations are usually completed within five to ten working days. Full micro-tensile or fatigue campaigns requiring specimen release, handling, and extended cyclic testing may require two to four weeks depending on sample count and conditions.