FYSC23 Powder XRD Analysis

This project provides a Python-based tool for analyzing powder X-ray diffraction (XRD) data. It determines the lattice constant of a sample and automatically decides whether the sample has a face-centered cubic (fcc) or body-centered cubic (bcc) structure based on linear regression quality. Additionally, the project includes options for cleaning raw data files and running crystallite size calculations using the powerxrd module.

Features

• Lattice Constant Determination:
  Calculates the lattice constant from XRD data using linear regression on sin²θ versus theoretical Q values.

• Crystal Structure Identification:
  Automatically selects between fcc and bcc structures based on regression quality.

• Data Cleaning:
  Provides functions to clean raw XRD data (e.g., Sample1.txt, Sample2.txt, etc.) using a rolling median filter to remove outliers, saving the cleaned files as filtered_sample<X>.xy.

• Crystallite Size Calculations:
  Includes size calculations routines for additional analysis steps such as back-subtraction, all peaks detection, and Scherrer peak analysis.

Requirements

• Python 3.x (tested on 3.12.6)
• NumPy
• Pandas
• Matplotlib
• SciPy
• powerxrd

You can install the required packages using pip:

pip install -r requirements.txt

Installation

git clone https://github.com/sebebbeb/FYSC23-PXRD-Lab.git
cd FYSC23-PXRD-Lab

Usage

The main script is main.py. Assuming the format of the data doesn't change in later iterations of the course. I'd recommend running the complete anlysis command-line option:

Complete Analysis

This cleans all raw data files found in the specified directory, saves them as filtered_sample<X>.xy and uses corresponding filtered file for the analysis. This analysis includes everything mentioned in the Features section above.

python main.py samples/Sample1.txt --clean --size

Basic Analysis

Here an already cleaned data is required

python main.py samples/filtered_sample1.xy

Clean Raw Data + Basic Analysis

If you have raw data files(e.g. Sample1.txt, Sample2.txt, etc) and want to clean them before analysis (recommended), use the --clean flag. This will:

• Clean all raw data files found in the specified directory.
• Save cleaned files as filtered_sample<X>.xy.
• Use the corresponding filtered file for further analysis.

python main.py samples/Sample1.txt --clean

Project Structure

• main.py:
  The main script that loads data, optionally cleans raw files, performs analysis (peak detection, regression, lattice constant calculation), and generates plots.

• data_cleaner.py:
  Contains functions for cleaning raw XRD data files. The cleaning functions process raw files (e.g., Sample1.txt) and save the output as filtered files (e.g., filtered_sample1.xy).

• powderxrd_patch.py:
  Contains additional functions for calculating the size of the crystallites and patching behaviors of the powerxrd library, including routines like test_allpeaks.

• samples/:
  A folder containing raw and filtered data files.

Troubleshooting

• File Paths:
  Ensure that the file paths provided to the script are correct. Relative paths are interpreted from the script's working directory.

• Dependencies:
  Verify that all required packages are installed and that you are using a compatible Python version.

• Data Format:
  The script expects data files to be in a two-column, tab-separated format (2θ and intensity). Ensure your data conforms to this format.

• Peaks Not Being Recognized:
  This is something that might occur. I'd recommend changing the height_frac and prominence in the detect_peaks function in main.py.
  If your problem lies in the Gaussian fits and Scherrer calculations take a look in src/powderxrd_patch.py and add the peaks manually in the test_allpeaks function. one peak around 2θ=35(deg) is already added manually.

• Dead Pixels/Spots Being Counted:
  If this is the case the data has somewhat changed since the code was written. The window size in the clean_data function of src/data_cleaner.py might need to be changed then. Furthermore, if this is the case, then you might need to change/delete the manually added peak in test_allpeaks as mentioned above.

Additional Notes

I've tried to make this program as modifiable as possible, I encourage you to modify all you see fit.