Inspection tool enables quality control of SiC epiwafers

Inspection tool enables quality control of SiC epiwafers

Technology News |
By eeNews Europe

Despite significant quality improvements in the past years, residual structural defects in the epiwafers are limiting to yield and performance of SiC bipolar devices. Currently, several methods are established to identify and assess such defects at the wafer level. However, all these methods have their drawbacks: Defective Selective Etching destroys the wafer, and Synchrotron X-Ray Topography is extremely cost-intensive. And both of them are very time-consuming – which rules out methods for use in an industrial series production.

But there are alternatives: A well known method allowing for the identification of structural defects is the photoluminescence (PL) method. It is non-destructive and can be used at ambient temperature. In PL images, structural defects can be identified easily as bright or dark artefacts on a grey SiC background. However, this method is not fast enough for an inline defect analysis on full waferscale in an industrial production environment.

In the SiC-WinS project, Fraunhofer IISB and Intego, both located in Erlangen, Germany, developed a defect luminescence scanner (DLS) fast enough for industrial production environments. Intego, a manufacturer of customer-specific camera systems for quality assurance applications, contributed its extensive expertise in the field of image processing while Fraunhofer IISB took the overall project responsibility and contributed its semiconductor know-how. The DLS enables short PL measurement cycles and high throughput of SiC epiwafers at a high lateral resolution of 5 µm.

The system is currently installed at Fraunhofer IISB and embraces an UV laser with a wavelength of 325 nm for PL excitation, a sample stage for wafer scans, and an electron multiplying charge-coupled device (EMCCD) camera for fast image recording at high signal quality. The high lateral resolution is achieved by a magnifying lens in front of the camera. Several bandpass filters enable the users to identify defect types by their specific spectral characteristics, or "fingerprints". The system can determine the defect types and their distribution on SiC 150 mm (6") epiwafers in less than 30 minutes. A routine for automated defect identification and counting is currently under development; this routine will allow for predict the device yield per wafer.

For more information visit

Related links and articles:

EU promotes metrology leading-edge research

Agilent, Cascade combine forces on wafer-level measurements

Power wafer-level testing and automated characterization

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News


Linked Articles