- SPO – high-energy optics
- Space instruments
- Airborne instruments
- HSI – hyperspectral imaging systems
- IMET – Industrial measurement systems
- PDS – Particle detection system
- MPS – Radiation detection system
- Multisensor camera systems
- CONDI – optical inspection systems
Particle Detection System
how clean is clean?
For the last several years cosine has been developing Silicon Pore Optics (SPO) for the next generation X-ray telescopes. The SPO must be assembled in a clean room to prevent dust particles that might deposit on the plates from affecting their eventual performance.
But how clean is clean? As it turns out while cleanliness is important, where the plates are clean is almost as important. In fact, in order to bond silicon plates, ribs 170 micron wide and spanning the entire plate length must be made on one side of the plate. Then the ribbed side of one plate is pushed on the non-ribbed side of the following plate. While this is enough to create a mechanically solid structure, any particle under those thin ribs can adversely affect the strength of the bond between plates.
In order to address this problem, cosine has developed a Particle Detection System (PDS) capable of locating sub-micron particles on optically smooth surfaces. Uniquely, our method allows us to deal with surfaces of variable shape, as the SPO must be bonded in a curved structure in order to be able to focus X-rays.
Optical plates are handled by a robotic arm which uses a customized gripper designed to prevent surface contamination. Using a multi-pass calibration scan, a plate is loaded into the inspection area and illuminated with a strong light. After an initial calibration scan, scan parameters are stored to allow batch scanning of subsequent plates.
Shown above are the key hardware components: (1) gripper in the inspection area, (2) robotic arm, (3) CCD line scanner,(4) light source, (5) linear stage.
The PDS employs a double-pass dark-field/bright-field surface scan to reliably detect the boundaries of the plate, derive the surface profile, and detect any particles present on the surface.
During the first pass, the precise shape of the plate is determined: this helps eliminating false positive identifications caused by excessive edge scattering. Additionally, the intensity profile of the plate is calculated. This enables intensity normalization during the second pass, when the system is configured to highlight surface defects the on the image.
Detected features are then isolated, analyzed, and categorized into particle classes including foreign objects, stains, image artifacts and other anomalies. The system is configured to look out for problem particles and to minimize the number of harmless or false positives reported. Potential problem particles are grouped by size and position making it possible to make high level decisions on whether to accept or reject the plate and/or low level decisions such as whether to actively try to remove problem particles on the fly. In particular, particles that would come to reside under one of the ribs, compromising the solidity of the bond.
The image below shows a small extruded three-dimensional region from sample scan data in which sub-micron particles are highlighted. As indicated, particles as small as 100 nm have been observed, although only particles 350 nm and larger have been verifiably detected and pin-pointed. The smooth surface to the right is the optical surface free of contaminants.
In the lab or on the factory floor, cosine can build effective customized metrology solutions.
For more information please contact Dipl.-Ing Max Collon.
- Tel+31 71 5284962
cosine measurement systems - Oosteinde 36, 2361 HE Warmond - The Netherlands