Benchmarks
Embedded Ionization Potential (IP) Benchmark
Benchmark Summary
Aspect |
Details |
|---|---|
Simulated Property Benchmarked |
Onset of IP distribution |
Size of Benchmark Set |
16 organic materials made of OLED molecules |
Computational Method |
Nanoscope software |
Reference Data Sources |
|
Accuracy Achieved |
Standard error of ~60 meV after offset correction |
Key Outcome |
Accuracy close to chemical accuracy |
Published In |
Reference [d] |
Overview
Here we study Nanoscope accuracy in computing the onset of ionization potential (IP) distributions for organic molecules embedded within organic materials by comparing our computational results with experimental data from the literature [a], [b], [c] for 16 OLED materials.
The IP distribution of the OLED molecules embedded within the respecting organic materials was computed and compared against Photoelectron Yield Spectroscopy (PYS) data from literature. For sake of consistency we use experimental data derived with similar experimental setup, i.e. the AC3 setup provided by RIKEN.
Initial Comparison
The following figure shows onsets of the IP distributions computed with the Nanoscope vs. PYS values measured with AC3, as reported in literature [a], [b], [c].
Within each dataset (indicated by different colors corresponding to references [a], [b], and [c]), there is agreement within 100 meV, except for the molecule tmbt. However, systematic offsets were observed between datasets from different studies.
Offset Correction
To address these systematic offsets, we applied corrections based on the mean differences between datasets. After this adjustment, a strong global correlation emerged:
The combined data has a standard error of approximately 60 meV, which is below the experimental error of measurement of 75 meV.
For a comprehensive discussion of this benchmark, please refer to our publication [d].
References
Full description of this benchmark: SID Symposium Digest of Technical Papers, 55: 607-610; DOI: 10.1002/sdtp.17597