10. X-ray Fluorescence Analysis with Larch

X-ray Fluorescence Data can be manipulated and displayed with Larch. Some tools for XRF analysis, to convert XRF spectra into elemental concentrations, are also available and more are being added. A work in progress, this chapter discusses some of the existing tools.

10.1. MCA and ROI objects

Traditionally, XRF spectra have been measured with MCAs (Multi Channel Analyzers) which holds an array of intensities. Because of this history, the two concepts are often mixed, and one often talks about MCA traces as being the same as XRF spectra. To be sure, an XRF spectrum \(I(E)\) can be represented well with an MCA trace (intensities per bin) as long as one can convert bin number to energy. Fortunately, most measurement systems in use have a linear relation (calibration) between bin number and energy, and the use of an MCA trace as an XRF spectrum is straightforward. We will probably use the terms interchangeably here as well.

To be clear, an XRF spectrum collected by an MCA will have a one dimensional array of intensities (or counts, as detectors typically count X-rays the numbers are small enough to highlight the difference) with a single, well-defined energy for each bin. There will typically be between 1000 and 10,000 bins in each spectrum.

An ROI (Region of Interest) is a continuous portion of the XRF spectrum, generally representing a range of energies corresponding to a particular peak or X-ray emission line or family of lines. One often sums the counts in such an ROI to give a total number of counts for that emission line. An MCA spectrum may have many (typically 10s) of ROIs defined for particular emission lines.

In Larch, MCAs and ROIs are exposed as Groups, each with several attributes, and some built-in functions. For example, An MCA has arrays for energy and counts, as well as values for real_time and live_time, a deadtime correction factor dt_factor, and several ROIs. Each ROI has a left and right channel index, and a roi.get_counts() method.

10.2. Creating MCA objects

A simple way to create an MCA object is to read one from a disk file. For data collected at GSECARS, this can be done with the read_gsemca() function.


read a GSECARS MCA spectra file, returning a Group

Parameters:filename – name of GSECARS MCA file

The returned Group has the following components:

component name description
filename name of file
mcas list of MCA objects for each MCA saved in the file
rois list of ROIs
environ list of Environmental Variables
energy array of energy values
counts array of counts, deadtime corrected and summed over MCAs
raw array of counts, summed over MCAs, not corrected.
calib dictionary of calibration values
dt_factor deadtime correction factor
real_time real time for data acquisition
live_time live time for data acquisition
nchans number of energy points in spectra
get_roi_counts() function to get counts for a named ROI
save_mcafile() function to save MCA to file
xrf_plot(energy, counts, mca=None)

create an interactive window for displaying an X-ray Fluorescence spectra

  • energy – array of energy values
  • counts – array of counts
  • mca – MCA group (as read from read_gsemca()), containing ROI definitions, calibration values, and related data.

An example plot is shown below


Figure 10.1: Example XRF Display, showing X-ray Fluorescence spectra, defined ROIs (in red), and Periodic Table for showing predicted emission lines.