OctCarb (New)
Purpose and Objective of the Program
OctCarb is a script for Octave, which can be used to analyze and evaluate scattering data of non-graphitic carbons. The so-called non-graphitic carbons (NGCs) are an important class of sp2-based carbon materials that encompass a wide variety of natural and synthetic carbons in industry and research (e.g., charcoal, activated carbon, glassy carbon, soot, coal pitches). The bulk material can be used for various electrical and low-friction applications, while porous derivatives are used, for example, in gas storage/separation. Non-graphitic carbon can also be used to manufacture electrodes for batteries and supercapacitors. Carbons derived from phenol-formaldehyde resins (PF-R), the so-called glassy carbons, serve as containers in high-temperature applications due to their high chemical resistance.
Physical properties such as thermal and chemical resistance, as well as electrical characteristics, are directly related to the microstructure of NGCs (Fig. 1: Stack width and layer height (La and Lc) as well as their disorder (σ1 and σ3), average and minimum layer distance (a3 and a3 min), and the average C-C bond length (lcc)). However, the exact quantitative characterization of such substances is difficult: electron microscopy (TEM, HRTEM) and Raman spectroscopy only show a small part of the sample. Therefore, wide-angle X-ray/neutron scattering (WAXS/WANS) is often used to obtain quantitative structural parameters of NGCs. Due to the broad and overlapping reflections (Fig. 2), evaluating the scattering data using conventional methods is not possible. Thus, special software is needed to analyze such data. However, the currently available software for analyzing this data suffers from certain limitations.
Figure 1: Basic structure of non-graphitic carbons (NGCs):
NGCs consist of stacks of turbostratically arranged graphene layers (Taken from Ref. 1).
Figure 2: Example wide-angle scattering data of NGCs:
The total scattering is an overlay of layer scattering, stack scattering, and incoherent scattering (Taken from Ref. 1).
For all applications, quantitative determination of the microstructure based on experimentally accessible structural parameters is of crucial importance, both for tuning production processes and linking material properties with the microstructure. The latter includes macroscopic properties like hardness, chemical stability, or thermal and electrical conductivity, thus defining the final application. These parameters can be obtained by evaluating the WAXS data of NGCs using an approach published in 2002 by Ruland & Smarsly[2]. Based on the work of Pfaff, Badaczewski et al.[3, 4], a script (OctCarb) was developed that can be used with the freely available software Octave to automatically evaluate scattering data and validly quantify the microstructure of NGCs.
Advantages and Application
The main features are (not exhaustive):
- Completely free and open-source for Windows, Mac, and Linux
- Adjustment of various microstructure parameters, such as:
- La and σ1 (average layer extent and disorder)
- Lc (average stack height)
- a3,a3 min, and σ3 (average and minimum layer distance as well as standard deviation (disorder))
- lcc (C-C bond length)
- Consideration of multiple correction terms and support for different X-ray wavelengths and measurement geometries
- Thinning factor for faster calculation of the fit curve and faster automatic fitting, with the option to omit points at the beginning and end
- Various scales for the X-axis are possible (θ, 2θ, s, q)
- Automatic output and storage of all parameters and fitting graphs as text, table, and image
- Data fitting for both X-ray and neutron scattering possible
- The complex mathematical calculations are visible to the user but do not need to be changed or manually considered
- Fully implemented mathematically well-known fitting algorithm (Levenberg-Marquardt)
The surface of Octave (Fig. 3) essentially consists of a script interface, and the results are output (Fig. 4) as text, table, and image. All data is automatically saved on the hard drive.
Figure 3: Main graphical user interface (GUI) of Octave:
In the top bar, Octave files (*.m) can be opened and saved, and various windows can be activated and deactivated. On the left side, the current working directory can be changed, and the currently used variables and executed commands are displayed there. The lower section allows switching between the command window (interface for the Octave command line), the editor (for editing script files), and the variable editor and documentation (Adapted from Ref. 1).
Figure 4: Results of the WANS refinement with the Octave script (OctCarb, top left):
In addition to the microstructural parameters in text form, categorized by stack and layer parameters, Octave displays the measurement data, the fit, and the deviation, saving everything in a CSV file and as images (Adapted from Ref. 1)
Figure 5: Noise level test with simulated WANS data, which was affected by statistical noise using a Gaussian distribution (σ = 0.05) (black):
The resulting fit of the noisy data (red) is very close to the simulated curve (blue). The applied noise level corresponds to the typical noise of a real WANS experiment. The deviations between the original and refined structural parameters are very small, meaning that such a noise level has no significant impact on the resulting calculated microstructural data. The data roughly corresponds to a resin treated at intermediate heat treatment temperatures (1800 – 2500 °C) or a pitch treated at lower temperatures (1200 – 2000 °C) (Adapted from Ref. 1).
Further Information, Contact and Download
The presented script (OctCarb) has been thoroughly tested and worked flawlessly. Nonetheless, the following disclaimer applies. The software is provided as-is as open-source software. Bug reports and suggestions are welcome (smarsly@uni-giessen.de or carbon).
The resulting script (OctCarb), manuals, videos, and further information are provided at https://github.com/osswaldo/NGCs.
An article about this software tool was published in the journal C - Journal of Carbon Research:
Osswald, O. & Smarsly, B. M. (2022). C. 8(4), 78.
Disclaimer
This software is provided without warranty and any express or implied warranty, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose, is disclaimed. The authors or copyright holders shall in no event be liable for any direct, indirect, incidental, special, exemplary, or consequential damages (including, but not limited to, the procurement of substitute goods or services; loss of use, data, or profits; or business interruption) arising in any way from the use of this software, whether in an action of contract, strict liability, or tort (including negligence or otherwise), even if advised of the possibility of such damage.
References
[1] Osswald, O. & Smarsly, B. M. (2022). C. 8(4), 78
[2] Ruland, W. & Smarsly, B. (2002). J. Appl. Cryst. 35, 624–633
[3] Pfaff, T. et al. (2018). J. Appl. Cryst. 51, 219-229
[4] Pfaff, T., Badaczewski, F. M. et al. (2019). J. Phys. Chem. C 123 (33), 20532–20546