Week n+15¶

Sunday 10/10¶

I had a meeting with Katya which was mostly just a “you’re doing great, just keep pushing through” chat
Did a little bit of polishing to my presentation

It’s 2:30 and I’ve not done a huge amount. I was teaching before, and I watched some honours presentations this morning. Once my teaching hour is up (in case any students pop back in and need a hand), I’ll probably practice my talk a bit and maybe try and work on the “Relaxed \(\vec F\)” section of my thesis.
- Okay, working on my thesis never happened…

Gave my final talk
- It went really well and I was genuinely surprised by how many academics had questions to ask
I’ve been working on the “Relaxed \(\vec F\)” section, and I reckon that I need one more calculation which is to obtain rate constants for the \(\ce{R1/R2=NO2}\) derivative, so that I can provide an indication of the kind of increases that derivatisation would afford, so I’ve got that running

So I got distracted and have trying to compile Psi4 on ARM64… * The good news is that Psi4 actually compiles on my M1 mac now, which is awesome! the bad news is that I’ve definitely not done much writing today…
I’ve decided that what this thesis really needs is a final benchmarking section to compare the \(\ce{R1/R2=NO2}\) deriv in both \(\vec F = 0.1\) and \(0.2\:V\cdot\AA^{-1}\) as well as to compare the na1 relaxed geom in \(\vec F = 0.2\:V\cdot\AA^{-1}\), so I’ve got those jobs queued atm.

Back in the city today I’m starting to get some results out from the new benchmark jobs and they’re looking interesting. the \(\ce{R1/R2=NO2}\) species has a much higher(\(21.4\:\kjmol\)) barrier in \(\vec F = 0.2\:V\cdot\AA^{-1}\) than it does in \(0.1\:V\cdot\AA^{-1}\), but the product is more stable by \(\sim10\:\kjmol\), so perhaps it’s just not been optimised properly, or perhaps that’s just how it works. The barrier is still higher than the \(0.1\:V\cdot\AA^{-1}\) stabilised non-derivatised species, so perhaps the benefit of the \(\ce{NO2}\) groups is limited to stereoselection, which would be pretty cool, since it would mean that the \(\ce{NO2}\) groups themselves would ultimately support the stereoselection.

I’ve got all the PyMOL scripts ready to produce the figures for the final section, and have got my Jupyter scrips ready to go as well. I just need the final job to finish!
The results from the derivatised benchmarks are proving to be really weird and may invalidate some earlier results, but I’m still pushing them through and have started writing bits of my conclusion and abstract

Job	Reactant	Transition State	Product
na1 \(\vec F = 0.2\:V\cdot\AA^{-1}\)	Done	Done	Done
Deriv \(\vec F = 0.1\:V\cdot\AA^{-1}\)	Done	Done	Running (S)
Deriv \(\vec F = 0.2\:V\cdot\AA^{-1}\)	Running (R)	Running (C/R/S)	Done

Finish off the rest of the drafty first edit, including all of the front/tail matter

Review the paper by Wang et al. to find any other analysis that you should perform, or any other things that should be discussed.

Section	Status	Notes
Abbreviations	First draft done
Abstract	To do (up to results has been drafted)
Intro	Second draft done
Computational details	First draft done
Reaction benchmarking	First draft done
Static \(F_Y\)	First draft done
Static \(F_Z\)	First draft done
Static \(\varepsilon_r\)	First draft done	more discussion tying together all of the static scans?
Efield Scans	First draft done	I’m not sure what else to add here, but it feels weak.
EDD maps	First draft done	perhaps needs a discussion of the implication of these mechanisms?
Relaxed \(\vec F\)	First draft done
Derivatives	First draft done
Final Benchmark	To do
Conclusion	To do (future work has been drafted)
Acknowledgements	First draft done
Appendices	WIP	should I add in lists of partial charges?