OpenSourceMalaria / Series4

Repository for Series 4 of the Open Source Malaria Consortium
17 stars 3 forks source link

2018 SSP Compounds #53

Open edwintse opened 6 years ago

edwintse commented 6 years ago

For the past 4 years, the first year Special Studies Program here at the University of Sydney have been synthesising new Series 4 compounds. This year, their target compounds were based around modification of the RHS aryl group with non-aromatic heterocycles (highlighted in green).

untitled wiley-10

The synthesis plan followed the typical route. The initial hydrazine SM was provided to them by us. This was then condensed with a commercially available CHO, cyclised and finally coupled with 2-phenylethanol. untitled wiley-5 Upon completion of the synthesis in the undergrad labs, I made attempts at purifying their crude products. This is where things became problematic... I ran the 4 compounds through 3 Isolera columns each (1 regular, 2 reverse phase), however none of the compounds ended up clean. Either the aliphatic region was still broad and messy or the aromatic region still had multiple impurity peaks, both of which I would have expected reverse phase to have cleaned up. I suspect that this may have been due to the a couple of reasons: Reactions could only be performed once a week within the timeframe of the 3 h lab. Reactions were left sitting until the next weeks lab session. Crude material was taken through for each step from the beginning.

Good news ––––––––––––––– I ended up resynthesising the 4 compounds with purifications of each intermediate before subsequent steps. The two protected amines were also deprotected to give a total of 6 compounds that will be sent for evaluation tomorrow. My ELN entries are here.

Not so good news ––––––––––––––– A comparison of their final products and the ones that I made are not consistent. In the spectra below, their compounds (after the 3 columns) are on top in blue and mine are below in red. It is clear that there are a number of differences between the spectra.

It is evident from this that they seem to have actually not made the desired compounds but something different. ssp comparison1 ssp comparison2 ssp comparison3 ssp comparison4 The only characterisation that the students performed was a 1H NMR of their crude condensation products. Comparison of these NMRs has identified yet another inconsistency. Again, their compounds are on top in blue and mine are below in red.

The only explanation for this is that they've somehow made the other isomer of the condensation product. The only difference between the reaction conditions is that they used E95 ethanol and I used E100 ethanol but I'm not sure how this would have made such a difference. No recrystallisations were done in either case. This may have also have influenced them not making the final products, however, I'm again not sure how this would have affected the subsequent cyclisation? ssp comparison 1 ssp comparison 2 ssp comparison 3 ssp comparison 4 I will look into the possibility that this could be due concentration dependence of the NMR samples to see if there's any shift in that signal. I will also get LRMS of their compounds to see if it matches.

MedChemProf commented 6 years ago

@edwintse I would tend to agree with the notion that all you are seeing is an artifact of concentration in the NMR in regards to the hydrazinyl intermediates. Unlikely, but do you happen to have any of their samples still available? If so, I would just mix the two and you would probably see a uniform spectra. A D2O washed sample might be worth running for comparison to future samples if your resynthesis does not go according to plan. In regards to the unknowns created by the students, you did not mention if they performed any LCMS runs on their products. From the complexity of the NMR, I would hazard a guess that the products might be heavy by the weight of a second pyrazine (or portion of the triazolopyrazine) due to some self-condensation while sitting unpurified throughout their synthesis.

edwintse commented 6 years ago

@MedChemProf I do still have some of their hydrazinyl intermediates so I'll definitely give that a go. Hopefully it's as simple as a concentration artifact. LCMS wasn't done on any of the students samples but I'll keep possible self-condensation products in mind when I've got the MS data.

edwintse commented 6 years ago

@MedChemProf I've just run an NMR of a mix of their product and mine and it indeed looks like just a concentration effect. There's just the one signal for the NH. Problem solved!! Now onto the mystery compounds...

mattodd commented 6 years ago

Nice work, @MedChemProf

MFernflower commented 6 years ago

Closing this issue - if some synthetic chemistry aspect has not been properly discussed feel free to reopen this issue or spin it off into a new one!

edwintse commented 6 years ago

Reopening until I get some more clarity on what the students actually made

edwintse commented 6 years ago

The mystery continues...

I have gotten the LRMS of each of the SSP samples and it indicates that the mass of the students samples matches that of the desired products, which seems odd considering the large differences in the NMRs above (pdfs attached): SSP 2018 4a LRMS SSP 2018 4b LRMS SSP 2018 4c LRMS SSP 2018 4d LRMS

Not sure how it would be, but I double checked to see if this was another concentration dependent thing, and as you can see from the spectra below (representative sample, mix of SSP 2018 4d and EGT 350-1), they are definitely different compounds. Top blue spectra is the mix, middle green spectra is my resynthesis, bottom red spectra is the SSP compound after purifications. ssp_egt_4d_mix

I'm stumped as to what they've made. I know that their condensation products are correct. No samples of their cyclisation products were kept to check. I can't see how it'd cyclise differently yet the aliphatic signals in the products are so different. There's a singlet for the pyrazine, so it's unlikely to be a 8-substituted product.

MedChemProf commented 6 years ago

@edwintse A possibility may that the presence of an sp3 hybridized carbon adjacent to the hydrazinyl intermediate allows for an alternative elimination route as compared to the normal sp2 aryl rings which are usually present. The alternative elimination product could then cyclize via a thermal electrocyclic reaction to give a spirocyclic product with the same molecular weight as the desired. Normal Oxidative Cyclization: image Alternative Oxidative Cyclization: image

MedChemProf commented 6 years ago

@edwintse If in fact the spirocycles have been made, I think this is actually a more interesting reaction as compared to the C5/C8 issue. Not because it is important to the project, but it produces a unique scaffold. I searched for the a generalized 4-, 5-, and 6-membered spirocycle (with and without heteroatoms) on a pyrazino-triazine core in scifinder and did not find any hits. I'd be interested in following up on this reaction if you are interested. Also, the C13 of these compounds should be easier to interpret as compared to the proton spectra.

MFernflower commented 6 years ago

If we can indeed confirm the spirocycle has been formed I think it would be wise to share this information with a building block supply house like enamine so as to have a commercially available novel core at a researchers disposal - who knows what it could lead to!

@medchemprof @edwintse

Do you think it would be worth making and then cyclizing this for study purposes? (condesation product of chlorohydrazinylpyrazine and formylcyclohexane) https://user-images.githubusercontent.com/3164942/42762576-b0c7a2c0-88de-11e8-805c-cf2d2cb70ed7.png

bendndi commented 6 years ago

Not spent too long looking at this, and the spirocyclic suggestion is theoretically a strong one, however this would not explain why the final compounds from students are missing the very clear CH2 triplets from the phenethyl ether. T The mass is clearly correct, suggesting that in the final step the phenethylalcohol has been added and the chloride displaced, but somehow the -OCH2CH2-Ph section is no longer symmetrical...

MFernflower commented 6 years ago

Is it possible that the chlorine was displaced by a free hydroxyl ion rather than the phenylethoxide ion we wanted? @bendndi @MedChemProf

bendndi commented 6 years ago

My understanding is that the mass spec of the final compounds is "as intended" for the final structure,

"I have gotten the LRMS of each of the SSP samples and it indicates that the mass of the students samples matches that of the desired products, which seems odd considering the large differences in the NMRs above (pdfs attached):"

This indicates that the chloride has somehow been replaced by the phenethyl (see post above). This will be very clear in the MS to see since not only the mass adds up but the characteristic Cl 35-37 fingerprint will have disappeared (@edwintse to confirm?). The lack of the clear -OCH2CH2- symmetry pattern in NMR suggests that it may now be in a chiral environment, resulting in a much more complex spectra for these 4 protons - difficult to tell much more without seeing the integrals...

@MFernflower

MFernflower commented 6 years ago

Would the spirocyclic system [assuming dr.Smith's idea is correct] on the right hand cause muddying of the phenylethanol fragment? @bendndi

bendndi commented 6 years ago

I've just looked at the mas specs @edwintse linked above - are these for the final compounds? If so, what is the mass ion you're seeing which suggests its the correct set of structures? Not being familiar with your MS system its difficult to know what we're looking at :-)

I'm seeing some pretty high MW ions (e.g. in MS for 4d) which suggest dimerisation type structures (some of which still contain the Cl 35/37 fingerprint...)

bendndi commented 6 years ago

@MFernflower In some this spirocyclic system would create a chiral environment (e.g. 4b and 4d) but in others it would still be a symmetric environment (e.g. 4a and 4c) - however I think it is not sufficient a perturbation to explain the complete lack of the O-CH2CH2Ph motif - such a chiral environment would only cause minimal impact in my opinion.

MFernflower commented 6 years ago

Would conversion of the halogen on the spirocyclic system to an alcohol be a possible explanation for the missing fragment - but weight should be different! 2222222222

@bendndi @edwintse @MedChemProf

david1597 commented 6 years ago

So good to wake up here in Aus and see a great discussion and suggestions while we slept. Looks like an interesting problem - @edwintse could you either post here or link in the ELN to the .fid files so we can play with the NMR spectra in detail?

The spirocyclic proposal is interesting, although as pointed out doesn't fully explain the differences seen in the phenethyl fragment - the aromatics are shifted as well as the lack of two clear triplets.

edwintse commented 6 years ago

Thanks for all this discussion everybody!

@MedChemProf Very interesting suggestion for the spirocyclic compound. Considering the novelty of this scaffold, it may be good to investigate further/try to replicate. I held off on getting other NMR because the protons weren't that clear, but I'll now get 13C and some 2D, at least for 4d (even if it'll be slightly messy).

@MFernflower If the phenylethanol fragment isn't there then the mass wouldn't make sense. The triplets seem missing/shifted but there are still the signals to the left of the CDCl3 that correspond to the phenyl group.

@bendndi Yes, the MS linked above are for the final compounds. The top MS is of the compound (+ve ESI), the second is zoomed in on the desired region, the third is a prediction of the desired mass+Na adduct, and the last two are -ve ion mode. In all cases, the predicted and experimental +ve MS match. The high MW ion in 4d would correspond to [2M+Na]+. From what I can see, there aren't any Cl isotope signals for those.

@david1597 I've attached a folder below that contains raw HNMR of the SSP compounds as well as my resynthesised compounds. I've also uploaded raw MS files in mzXML format for the SSP compounds (couldn't upload 4b, file size to big).

raw NMR files.zip SSP 2018 4a_LRMS.mzXML.zip SSP 2018 4c_LRMS.mzXML.zip SSP 2018 4d_LRMS.mzXML.zip

mattodd commented 5 years ago

Just reading through this fascinating issue again. @edwintse did you bring the samples over to London, and what are the chances of structural elucidation by crystallisation?

drc007 commented 5 years ago

How about using CryoEM?https://chemrxiv.org/articles/The_CryoEM_Method_MicroED_as_a_Powerful_Tool_for_Small_Molecule_Structure_Determination/7215332

mattodd commented 5 years ago

Well quite @drc007 I have spent a considerable part of this week looking into sourcing exactly this kit. Non trivial, and when I have clarity I might post on this further. If it works as advertised this could be powerful for us all in accelerating what we do. In the meantime, we could have a go the old fashioned way.

holeung commented 5 years ago

Every chemist in the world is probably talking about this article. I believe it requires a specialized micro-focus setup on an electron microscope. The experiment itself is simply electron crystallography.

MFernflower commented 5 years ago

Xray crystallography is probably still our best and most cost effective bet! @mattodd

MedChemProf commented 5 years ago

@edwintse I was just following up to see if you had any luck getting some additional data on the unknowns. Thanks.

edwintse commented 5 years ago

I've not been in the lab for a bit so nothing new just yet. The 2D NMR I got for one of the SSP compounds is still a bit too complicated to interpret, but I'm hoping to look into this a bit more early next year and will update again if I find anything new.

MedChemProf commented 5 years ago

@edwintse I have been having a few students continue to look at this sequence in order to help track down the anomalous products obtained by some of your students. Early in the synthetic sequence, after your students reacted the 2-chloro-6-hydrazinylpyrazine with their aldehydes to give the hydrazone products, how long was the time interval in between the synthesis of the hydrazones before they did the oxidative cyclization? Also, did they store the hydrazones as dry powers or were they in solution? Thanks!

edwintse commented 5 years ago

@MedChemProf The hydrazones were left at rt for a week before the cyclisations. All were stored as powders (either precipitated during the reaction and filtered or didn't precipitate so the solvent was removed). I've just started to look at the samples again too. I just ran them through LCMS last week. Haven't had a proper look through yet but the reports are attached below.

SSP 2018 4a.pdf SSP 2018 4b.pdf SSP 2018 4c.pdf SSP 2018 4d (f4-7).pdf SSP 2018 4d (f10-13).pdf

MedChemProf commented 5 years ago

@edwintse Thanks. The reason that I asked is that a student synthesized a new hydrazone (cyclohexyl) under two different conditions. The first was similar to the conditions in your notebook (ethanol solvent, hydrazine and aldehyde) and the second was similar to how we have done it in the past (acetonitrile, AcOH, hydrazine and aldehyde.) both conditions gave the same exact product that appeared to be a single isomer. Her notebook link is here. One of the NMR samples sat around for a few days and we just resubmitted it for NMR to get a COSY and we noticed that it now looks like the other isomer is slowly forming. Not conclusive yet, but I am going to let the sample stay in solution and see if the ratio changes. We will also check the dry powders again prior to running the oxidative cyclizations.

MFernflower commented 5 years ago

@maratsydney When you get a chance can you make a new issue about the 2019 SSP projects?