My watch list  

Hyperspectral imaging

Hyperspectral imaging, sometimes referred to as spectral imaging, is an electron microscopy technique that involves microanalysis using either Energy dispersive X-ray spectroscopy (EDS), Electron energy loss spectroscopy (EELS), Infrared Spectroscopy(IR), Raman Spectroscopy, or cathodoluminescence (CL) spectroscopy, in which the entire spectrum measured at each point is recorded. EELS hyperspectral imaging is performed in a scanning transmission electron microscope (STEM); EDS and CL mapping can be performed in STEM as well, or in a scanning electron microscope or electron probe microanalyzer (EPMA). Often, multiple techniques (EDS, EELS, CL) are used simultaneously.

In a "normal" mapping experiment, an image of the sample will be made that is simply the intensity of a particular emission mapped in an XY raster. For example, an EDS map could be made of a steel sample, in which iron x-ray intensity is used for the intensity grayscale of the image. Dark areas in the image would indicate not-iron-bearing impurities. This could potentially give misleading results; if the steel contained tungsten inclusions, for example, the high atomic number of tungsten could result in bremsstrahlung radiation that made the iron-free areas appear to be rich in iron.

By hyperspectral mapping, instead, the entire spectrum at each mapping point is acquired, and a quantitative analysis can be performed by computer post-processing of the data, and a quantitative map of iron content produced. This would show which areas contained no iron, despite the anomalous x-ray counts caused by bremsstrahlung. Because EELS core-loss edges are small signals on top of a large background, hyperspectral imaging allows large improvements to the quality of EELS chemical maps.

Similarly, in CL mapping, small shifts in the peak emission energy could be mapped, which would give information regarding slight chemical composition changes or changes in the stress state of a sample.

The primary advantages to hyperspectral mapping are that, because an entire spectrum is acquired at each point, the operator needs no a priori knowledge of the sample, and post-processing allows all available information from the dataset to be mined.

The primary disadvantage is that fast computers, sensitive detectors, and large data storage capacities are needed. This is because hyperspectral images are multi-dimensional datasets where the third and higher dimensions represent entire X-ray, EELS, or cathodoluminescence spectra at each specimen point. A 256x200 pixel EDS map could conceivably be ~100megabytes, as an entire x-ray spectrum must be stored for each pixel point.

As a relatively new analytical technique, the full potential of hyperspectral imaging has not yet been discovered

See also

This article is licensed under the GNU Free Documentation License. It uses material from the Wikipedia article "Hyperspectral_imaging". A list of authors is available in Wikipedia.
Your browser is not current. Microsoft Internet Explorer 6.0 does not support some functions on Chemie.DE