Post-edge Correction

The postedge command removes the post-edge background from the absorption signal and determines the edge step (\(J_{0}\)).

Basic Usage 

postedge <range> [options]

Fits a polynomial to the absorption signal after the edge and removes it from the data to determine the absorption step.

Examples:

postedge 50eV .. --linear

postedge 8300eV 9000eV --degree 3

Command options

Option Description
<range> Energy range for fitting the post-edge polynomial (required). See range for details.

Polynomial degree options

These options specify the degree of the polynomial used to fit the post-edge background, and are mutually exclusive.

Option Description
--degree <n>
--deg <n>		 Polynomial degree (default: 2)
--constant
-C		 Use constant (degree 0)
--linear
-l		 Use linear fit (degree 1)
--quadratic
-q		 Use quadratic fit (degree 2, default)
--cubic
-c		 Use cubic fit (degree 3)

Fitting axis options

These options specify which column to use for fitting the post-edge background, and are mutually exclusive.

Option Description
--xaxis <col> Use specified column for fitting (default: E)
--E-axis
-E		 Use energy column for fitting (default)
--k-axis
-k		 Use k column for fitting
--e-axis
-e		 Use relative energy (E - E0) for fitting

Mode options

These options specify how the post-edge background is removed from the data, and are mutually exclusive.

Option Description
--mode <mode> Post-edge correction mode (default: division)
--subtraction
--subtract
--sub
-s		 Subtract fitted polynomial from data
--division
--divide
--div
-d		 Divide data by fitted polynomial (default)

Results

The command fits a polynomial to the specified post-edge region, and uses it to correct the absorption signal. The a column is updated with the corrected absorption, and a new post column is defined containing the fitted polynomial values. The edge step (\(J_{0}\)) is determined from the value of the fitted polynomial at the edge energy (\(E_0\)) and stored in the J0 variable.

The subtraction mode removes the fitted polynomial from the absorption signal, such that the corrected absorption is given by:

\[a_{\text{corrected}} = a_{\text{original}} - \text{post} + J_{0}\]

The division mode divides the absorption signal by the fitted polynomial, such that the corrected absorption is given by:

\[a_{\text{corrected}} = \frac{a_{\text{original}}}{\text{post}} \cdot J_{0}\]

The command does not perform normalization by the edge step, it only determines its value. Normalization is performed in the subsequent normalize step. This could seem counterintuitive, but it allows for a consistent workflow between the two modes, and gives users more control over the normalization process.

Tips and Best Practices

  1. Range selection:
    • Start 50-200 eV after the edge
    • End as far from the edge as your data allows
    • Avoid including any early EXAFS features
  2. Polynomial degree:
    • Quadratic (degree 2) works for most cases
    • Constant (degree 0) is suited for XANES analysis, where the post-edge background may be severely distorted by early EXAFS features
    • Use linear (degree 1) only if background is nearly linear
    • Use cubic (degree 3) for strongly curved backgrounds
    • Avoid higher degrees unless absolutely necessary

See also: