Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO 25178-603: Design and Characteristics of Non-Contact (Phase Shifting Interferometry) Instruments
Geometrical Product Specifications (GPS) — Surface Texture: Areal — Part 603
1. Principles of Phase Shifting Interferometry
ISO 25178-603:2025 (second edition) specifies the design and metrological characteristics of phase shifting interferometry (PSI) instruments for areal surface topography measurement. PSI achieves sub-nanometer vertical resolution by analyzing interference patterns acquired over a sequence of controlled phase shifts, making it the method of choice for ultra-smooth surface characterization. The technique is most often employed for measurements of optically smooth surfaces as defined in ISO 25178-600, where height variations are much smaller than the wavelength of light, typically less than one-quarter of the equivalent wavelength.
The interference signal at the camera for an individual surface point follows the two-beam interference equation: I = IDC + IAC cos(phi – alpha + delta), where phi = 4pi z / lambda_eq is the surface-height-dependent phase, alpha is the phase shift imparted by the shifting mechanism, and delta is a phase offset related to reflection properties. The equivalent wavelength (lambda_eq) serves as the fundamental scaling factor for converting interference phase to surface height, typically approximately twice the mean source wavelength for low numerical aperture systems. The height at each point is calculated as z = (lambda_eq / 4pi) * phi, giving a direct path from phase measurement to surface topography.
PSI is best suited for optically smooth surfaces where height variations are much smaller than one-quarter of the equivalent wavelength. The equivalent wavelength (lambda_eq) converts interference phase to surface height, typically approximately twice the mean source wavelength. Phase unwrapping algorithms can extend this range by removing multiples of 2pi from the phase values of neighboring image points.
| Interference Objective | Magnification | Numerical Aperture | Optical Lateral Resolution (Sparrow, 570 nm) |
|---|---|---|---|
| Zygo Wide Field | 1.4x | Low | Large field of view |
| Michelson | 2.75x, 5.5x | 0.04-0.08 | 3.56-7.13 um |
| Mirau | 10x, 50x, 100x | 0.15-0.85 | 0.34-1.90 um |
| Linnik | Various | High NA | Sub-micrometer |
2. PSI Algorithms and Interference Objective Types
The standard describes both linear PSI (evenly spaced phase shifts) and sinusoidal PSI (sinusoidally-varying phase shifts). Linear PSI relies on sampling an interference signal over a sequence of evenly spaced phase shifts, typically generated by an axial scan motion of the interference objective or the sample surface along the z-axis. A simple 4-sample linear PSI algorithm calculates phase using tan(phi) = (I3 – I1)/(I1 – I0). Modern instruments use more sophisticated algorithms with up to 20 samples and floating-point coefficients for improved immunity to phase shift errors and environmental vibration. Sinusoidal PSI uses sinusoidally-varying phase shifts, which are less demanding mechanically but require more complex signal processing algorithms. The phase shifting mechanism is a critical component that imparts controlled phase shifts through a precisely calibrated mechanical scan.
PSI instruments use interference objectives in place of conventional microscope objectives. The standard describes four common types with different performance characteristics. Michelson objectives use a separate reference mirror and beam splitter, offering low magnification (2.75x-5.5x) with large working distances and field of view. Mirau objectives integrate the reference surface and beam splitter within the objective body, providing medium to high magnification (10x-100x) with excellent vibration stability. Linnik objectives use separate objective lenses for the reference and measurement paths, enabling the highest numerical apertures and best lateral resolution. Zygo Wide Field (ZWF) objectives offer very low magnification (1.4x) with extremely large field of view for measuring larger surface areas. Camera selection involves not only field size and pixel count but also acquisition speed, response linearity, quantum well depth, and digitization resolution.
PSI instruments are fundamentally sensitive to environmental vibration. Vibration frequencies below the camera framerate generate form distortions, while higher frequencies produce cyclic errors in the phase calculation. A vibration isolation table is essential for reliable PSI measurements. Some advanced PSI algorithms can partially compensate for vibrations but cannot eliminate their effects entirely.
3. Surface Films, PCOR Effects, and Practical Considerations
Surface films and dissimilar materials have a first-order effect on the phase offset in PSI measurements. The phase change on reflection (PCOR) varies with material type and can introduce apparent height offsets of tens of nanometers. For pure metallic surfaces, the PCOR is typically in the range of 0.1 to 0.5 radians, corresponding to height offsets of 5-25 nm for visible light. For multi-material surfaces or samples with thin films, PCOR variations across the field of view can produce significant measurement artifacts that are independent of actual surface topography. These effects must be modeled and compensated using known optical properties of the materials.
The standard also addresses slope-dependent effects where the simple phase-to-height model requires correction. When measuring steep slopes approaching the numerical aperture limit, the effective numerical aperture changes across the field, and the assumption of a constant equivalent wavelength breaks down. Phase unwrapping algorithms extend the measurement range beyond the fundamental 2pi ambiguity limit of lambda_eq/2 (approximately 300-400 nm for visible light). Fringe-order errors, which are integer multiples of lambda_eq/2 in height, can occur when phase unwrapping fails due to surface discontinuities or noise. The standard recommends using the smallest available equivalent wavelength (shortest source wavelength or largest effective NA) to minimize phase unwrapping errors while maintaining the required height resolution.
For best PSI practice: (1) always use vibration isolation, (2) calibrate equivalent wavelength using certified step-height standards traceable to the meter, (3) select objectives matched to the expected surface roughness and spatial frequencies, (4) account for PCOR when measuring dissimilar materials, and (5) use phase unwrapping to extend range beyond the 2pi ambiguity limit. Document all measurement conditions including temperature, vibration levels, and optical setup.
4. Frequently Asked Questions
Q: What is the maximum height range for PSI without phase unwrapping?
A: Limited to lambda_eq/2 (approximately 300-400 nm for visible light). Phase unwrapping extends this range by removing multiples of 2pi, enabling measurement of steps and features many times larger than the basic ambiguity limit.
A: Limited to lambda_eq/2 (approximately 300-400 nm for visible light). Phase unwrapping extends this range by removing multiples of 2pi, enabling measurement of steps and features many times larger than the basic ambiguity limit.
Q: What interference objectives are used in PSI microscopes?
A: Four main types: Michelson (low magnification, large working distance and FOV), Mirau (medium-to-high magnification, integrated design), Linnik (highest NA, best resolution), and Zygo Wide Field or ZWF (very low magnification, extremely large FOV).
A: Four main types: Michelson (low magnification, large working distance and FOV), Mirau (medium-to-high magnification, integrated design), Linnik (highest NA, best resolution), and Zygo Wide Field or ZWF (very low magnification, extremely large FOV).
Q: Why do surface films affect PSI measurements?
A: Films change the phase change on reflection (PCOR), introducing a material-dependent height offset of typically 5-25 nm that is unrelated to actual surface topography. For multi-material surfaces, PCOR variations can produce significant artifacts.
A: Films change the phase change on reflection (PCOR), introducing a material-dependent height offset of typically 5-25 nm that is unrelated to actual surface topography. For multi-material surfaces, PCOR variations can produce significant artifacts.
Q: What is the difference between linear PSI and sinusoidal PSI?
A: Linear PSI uses evenly spaced phase shifts typically generated by mechanical scanning, while sinusoidal PSI uses sinusoidally-varying shifts that are less demanding mechanically but require more complex algorithms for phase extraction from the modulated signal.
A: Linear PSI uses evenly spaced phase shifts typically generated by mechanical scanning, while sinusoidal PSI uses sinusoidally-varying shifts that are less demanding mechanically but require more complex algorithms for phase extraction from the modulated signal.
📥 Standard Documents Download
No download files available yet