Chemical Imaging
Technical Contact: Dr. John Henry Scott

Chemical analysis on microscopic samples may use a variety of microscopes and particle-beam analytical probes. These instruments are frequently used to yield point analyses, but these same approaches become much more powerful when used to attain give chemical information as maps or images. The purpose of this project is to extract information in these images into understandable, visual, useful, and chemically meaningful form.

We explore new image collection, processing, segmentation, and analysis techniques, and multiple-image analysis methods such as principal components and factor analysis. Other approaches include modeling the interaction of the electron or ion beam with solids to better understand the chemical information displayed in the images.

Technical Approach

Chemical images are usually formed from image stacks where each image in the stack represents the intensity map of a different spectral region. These images are typically created in one of two generic ways ... spectral imaging or filtered imaging. In spectral imaging we collect a spectrum at each pixel of an image whereas in filtered imaging we collect a whole image at a selected spectral region. While these two modes are quite different for collection, there end use can be very similar. Some of the first chemical images were collected as simple x-ray "dot maps" where a small energy region or window of x-rays was collected as a function of position

We define spectral imaging (also frequently referred to as hyperspectral imaging) as the method where one or more full spectra or significant portions of spectra (x-ray, EELS, Auger, IR, and/or Raman ...) are collected at every pixel of a digital image.

Filtered imaging may be defined as collecting a whole image at one somewhat narrow band of a spectrum (x-ray, EELS, Auger, IR, and/or Raman ...). This is currently how many satellite images are taken ... a whole image in each of many different spectral band pass regions (from the IR through the visible to the UV) to collect a suite or stack of images that represent useful portions of the light spectrum. Our research interests are in developing methods to efficiently collect and quantitatively interpret these chemical images. New approaches to three dimensional chemical imaging are rapidly becoming required to characterize biosystems and the complex nanotechnology devices now being manufactured. Facilities include a variety of locally and commercially developed image processing, spectroscopic and statistical analysis systems; many electron microscopes with a suite of state-of-the-art detectors, cooperative microscopists, and lots of data on interesting samples.

Web Contact micro@nist.gov