Validation

Approach to validation

The principles established for the validation of chemistry sets are:

1. Experimental benchmarking from open sources (where available) and also directly provided by Quantemol’s industrial partners (collaborating on the Powerbase project) and database contributors.
2. Calculations need be performed for a range of models thereby reflecting the underlying quality of input data (example models used for validation include HPEM, Global_Kin, ChemKin).
3. Attempts will be made to validate the models used to produce the data; performed on a case by case basis.
4. Numerical uncertainties will be quantified and a threshold set for validation where possible.

This methodology was specifically applied to atomic and molecular calculations using the principles established in the publication "Uncertainty Estimates of Theoretical Atomic and Molecular Data"[1], which was produced for the International Atomic Energy Agency and focused “on data that are most important for high temperature plasma modeling” with the “ultimate goal to develop guidelines for self-validation of computational theory for A+M processes”.

It is recognised that while the validation of chemistry sets directly may still be uncertain, the validation of data produced by models using this data will often be more easily obtained.

QDB users are invited to validate chemistry sets either directly or by validating the results of models which use these chemistry sets as inputs. Validation of the chemistry sets provided in the database will be based on the foundations of Uncertainty Quantification for calculations of complex systems [2,3].

Plasma researchers are encouraged to contribute to the validation process and receive 6 months free access to the subscription section of the database per contribution.

The chemistry set will be used for the modeling of different reactors. Comparison of the model output with measurement is the principal method for validation.

For chemistry simulation, the scaling law based on the parameter study is a common methodology for this validation [4]. For higher dimensional simulation, the behaviour of the species and the surface will be used for comparison [5].  

Part of this chemistry will be validated by comparison of different measurements which are and will be available to us in the context of the Powerbase project and different academic research groups from the reviewer’s panel.

We hope that the community will take part in the validation and subsequent scientific discussions, which will facilitate research in this field.

Comprehensive and complete referencing will be provided with users downloading chemistry sets, ensuring that relevant citations to chemistry set and validating experiments are included and can be used for publications.

Validation of Individual Chemistry Reactions:

Rate coefficients of each reaction are included in the validated chemistry set for a similar range of temperature and pressure.  

The main validation method for individual reactions  will be compared with alternative theoretical calculations/estimations and experimental measurements.  

For unknown reactions we are using different calculation methods which include:

• Quantemol-N (R-matrix method) calculations for electron molecule scattering reactions
• Scaling law, mathematical methods of estimation and expert opinion to estimate necessary data
• Quantum and transition state theory for unknown heavy particle reactions

User data for validation is welcomed with 6 months free subscription; references to the contributors’ original publications are included in publications using the data.

Validation of Surface reactions:

Surface reactions will be validated as part of the chemistry set. The main parameters used for comparison are etching rate and deposition rate, where available.

Unknown surface reactions will be calculated by quantum chemistry methods or estimated by scaling laws based on similar reactions.

Ideas on validation methods are very welcome.

Please contact us on support@quantemol.com.

References

1. http://www.siam.org/activity/uq
2. H. K. Chung et al “Uncertainty Estimates for Theoretical Atomic and Molecular Data”, (2016) J. Phys. D: Appl. Phys. (submitted)
3. National Research Council Committee on Mathematical Foundations of Verification, Validation and Uncertainty Quantification, Assessing the reliability of complex models: mathematical and statistical foundations of verification, validation and uncertainty quantification (National Academies Press, 2012)
4. Principles of Plasma Discharges and Materials Processing, Micheal A.Lieberman, Alan J. Lichtenberg, 1994,(John Wiley & Sons, 2005), ISBN 0-471-72001-1
5. Zhang, Da, and Mark J. Kushner. "Investigations of surface reactions during C2F6 plasma etching of SiO2 with equipment and feature scale models." Journal of Vacuum Science and Technology-Section A-Vacuum Surfaces and Films 19.2 (2001): 524-538.