This Entry
PermalinkURI
URI Label
Revisions
Add to List
Edit Entry
Export:
XML
This Notebook
New Entry Notebook Settings
Timeline View
Export Notebook
Feed (+Comments)
Archives
- May 2014 (10)
- April 2014 (5)
- March 2014 (9)
- February 2014 (15)
Authors
Sections
- BLL - Background (3)
- BLL - ideas (1)
- Tambjamine (2)
- Gaussian (3)
- Notebook (1)
- Planning (2)
- Programs and Software (3)
- QSAR (10)
- QSAR - Descriptors (8)
- QSAR - general (2)
- QSAR - Introduction (4)
Project
- BLL (5)
- QSAR (27)
- Tambjamine (1)
Tools
Show/Hide QR CodeShow/Hide Keys
Following on from the post Testing Geometry Optimisation the 4 test files were as follows:
| Test No. | Structure Type | Program | Theory level | Basis Set | Successful run? | Time taken |
| 1 | 2D | ChemDraw | B3LYP | 6-311+G** | No - error | n/a |
| 2 | 3D | MarvinSketch | B3LYP | 6-311+G** | Yes | 4:49:43 |
| 3 | 3D | OpenBabel (from 2D) | B3LYP | 6-311+G** | No - timed out | >5:00:00 |
| 4 | 3D | OpenBabel (from 3D) | B3LYP | 6-311+G** | No - timed out | >5:00:00 |
So this shows that only 1 of the files completed successfully. However the time taken to complete seems excessively large as one of the previous script tests completed in just under 3 hours.
This is a comparison of the input molecular structure and output molecular structure.
Geom_test_2 input vs output
This shows a diference in input and output structures although not a huge difference.
Due to the long timescales encountered another functional/basis set was set up (from another similar paper) This is the HF/6-31+G* (HF is a Hartree-Fock method)
| Test No. | Structure Type | Program | Theory Level | Basis Set | Successful Run? | Time taken |
| HF_1 | 2D | ChemDraw | HF | 6-31+G* | No - error | N/A |
| HF_2 | 3D | MarvinSketch | HF | 6-31+G* | Yes | 0:18:16 |
| HF_3 | 3D | OpenBabel (from 2D) | HF | 6-31+G* | Yes | 0:30:47 |
| HF_4 | 3D | OpenBabel (from 3D) | HF | 6-31+G* | Yes | 0:13:49 |
As can be seen these tests take a significantly shorter time, with HF_4 taking the shortest time.
Comparison of test_2 between B3LYP and HF theory level optimisations:
HF vs B3LYP - test_2
Comparison of outputs from test 2, 3 and 4:
HF test output 2,3,4
This shows that the outputs from the different tests vary fairly significant;y. This may be due to the fact that geometry optimisation only finds the local minimum and not the global minimum.
The 3 input structures were optimised in different programs prior to their optimisation in Gaussian, the different programs appear to use different methods for optimisation, leading to the difference in 3D structures.
The actual change between the input and output structures of any of the tests is not very large.
Attached Files
Geom_test_2 input vs output
HF vs B3LYP - test_2
HF test output 2,3,4
Add comment to Entry