[DoctorDomo]

What does it mean if you get a crappy signal to noise ratio? I'm an undergrad (4th year chem major doing my thesis), I don't get to run NMRs myself, I have to submit them to the postgrads in the NMR room, I dissolved my product (its a benzene ring which should have at least some fluorine substituents) in CDCl3 and submitted it for a HNMR and an FNMR. The proton NMR came out fine, but the fluorine NMR is a big mess, the FID is basically just a straight line. They tried twice but both attempts produced that shitty spectrum.

What does this mean? I considered that all F atoms mighta been lost in my reaction, but that would just produce a straight line wouldn't it? The spectrum is a big thick (wasted a lot of ink printing it) mess with a thousand peaks, and only one thing that could be considered possibly a real peak. Why would the proton NMR work fine, but the fluorine NMR not work? It looks kinda like this:

minus the shark peak. Also its not perfectly linear, its more wavy. I'll upload it when I have the time so you can see for yourself.

[Corribus]

You are showing two spectra. What is what? (Although they look somewhat identical).

[DoctorDomo]

Thats not my spectrum, I just posted it to show you an example of what my spectrum looks like. Theres so much noise, you can't determine anything. It rises at one point, but I took a look at the FID and it was just a straight line so something went wrong. The ran a proton NMR on the sample and it came out fine, I hear that fluorine NMRs are even more sensitive. Maybe I knocked all the F atoms off my product. Is there an open source database where people can upload their spectra for specific compounds? That would be pretty handy,

[Yggdrasil]

If you have access to a mass spec, that could be a good way to test whether you were successful in incorporating fluorine into your sample.

[DoctorDomo]

Yggdrasil: Heres my starting compound:

the F atoms are already there.

Heres the FNMR:

and heres a proton NMR run on the same sample:

my products clearly dirty, there are way too many protons in there.

Heres the FID for the FNMR:


as expected, I got a proper FID for the HNMR:


Any idea what would cause this to happen?

[Babcock_Hall]

In general one can obtain poor S/N when the probe is not tuned properly, but I don't have any reason to believe that this is the case here.  However, I do see one complication.  Are you familiar with the concept of magnetic equivalence?  Two nuclei can have the same chemical shift but not be magnetically equivalent.  Whether or not this issue explains your spectra, I could not say, but it might be worth trying a simulation of the spectrum.

[DoctorDomo]

However, I do see one complication.  Are you familiar with the concept of magnetic equivalence?  Two nuclei can have the same chemical shift but not be magnetically equivalent.  Whether or not this issue explains your spectra, I could not say, but it might be worth thinking about.

I'm not familiar with that concept but I'll learn it now that you mentioned it. The FID in this case though looks like something went completely wrong, I'm just not sure what. It looks like none of the F atoms got excited, otherwise there'd be something at the start of the FID, at first I thought maybe I knocked off all the F atoms on the starting material, but if that was the case (if there are no F atoms present), why didn't I get a straight line on my chemical shift spectrum?

Either way though, my theory is that there were no F atoms present in whatever I ran the NMR on. That would be an interesting result, but I have to be sure. Replacing those 4 electron withdrawing F atoms with electron donators will greatly alter the aromaticity of the benzene ring, but I'm not sure how to analyse that kinda thing. UV spectroscopy maybe? The colour that the product fluoresces at is the same as the starting material, I'm not entirely sure what that means, but I bet they absorb UV light at different frequencies (well I'm pretty sure of that, would make no sense if they didn't).

EDIT: Anyone know what wavenumber an aromatic C-F stretch or bend appears at on an IR spectrum? I haven't purified anything enough for C13 NMRs but I suppose an IR spectrum will tell me a thing or two about whats going on.

[Corribus]

Either way though, my theory is that there were no F atoms present in whatever I ran the NMR on. That would be an interesting result, but I have to be sure. Replacing those 4 electron withdrawing F atoms with electron donators will greatly alter the aromaticity of the benzene ring, but I'm not sure how to analyse that kinda thing. UV spectroscopy maybe? The colour that the product fluoresces at is the same as the starting material, I'm not entirely sure what that means, but I bet they absorb UV light at different frequencies (well I'm pretty sure of that, would make no sense if they didn't).

You should observe a shift in the electronic absorption spectrum but it may be hard to see unless you do it at low temperature. It depends on the magnitude of the shift. The unchanging fluorescence color may only mean you have starting material (whatever it is) contaminating your sample. Although a proper fluorescence experiment is much better than going by eye. Fluorescence spectroscopy can be extremely sensitive to impurities, so if you're not convinced your sample is clean (given you're proton NMR above, I'd say no), the fluorescence tells you absolutely nothing.

Do note that fluorine can couple to hydrogen in the NMR spectra, since they both have the same nuclear spin. This can make proton NMR spectra appear more complicated than you might otherwise assume.

EDIT: Anyone know what wavenumber an aromatic C-F stretch or bend appears at on an IR spectrum? I haven't purified anything enough for C13 NMRs but I suppose an IR spectrum will tell me a thing or two about whats going on.

Since C-O stretches occur usually around 1000 to 1200 wavenumbers, and F is slightly less massive than O, I'd assume that C-F stretch ranges from around 1100 to 1400 wavenumbers. I would also expect peaks should be strong because of the large dipole moment change.

[Corribus]

One other thing that might be worth doing is taking a ¹⁹F NMR spectrum of a model compound for which the ¹⁹F NMR spectrum is well known, like fluoroacetone. This will help to convince you that the experiment and equipment is working correctly.

EDIT: Also: what is your scan range? (Your thumbnails are too small for me to see.) Is it possible your fluorine resonances are outside your scan range?

[DoctorDomo]

Thanks for the info. I don't think the other purple spots are starting material because they don't move up the TLC plate like the starting material does. For my latest reaction, I separated a couple of compounds from the product, and one of them gives light blue fluorescence. So I need to purify the product before running an F19 NMR? I thought purities not so important with FNMR since only flourinated compounds are visible in that particular range. I'm guessing the F atoms leave as HF, not sure if thats soluble in DMF or not. Ah, so F-H coupling affects HNMR spectra too? That might explain some of the splitting in my spectrum.

EDIT: A weird observation is that an intense DCM peak appears in my HNMR spectrum. Weird since I didn't use DCM. Contamination from a shitty rotavap maybe? I ended up with sand in my flask after using this rotavap the other day so god knows what else my products have been tainted with.

[Irlanur]

To be honest, that all looks and sounds pretty unorganised and dirty... You wont find a solution in a forum.

[Babcock_Hall]

In reexamining your structure, I now realize that I must have misinterpreted it the first time I looked at it.  I thought that the two nitro groups were ortho with respect to each other, but I see that they are in a para relationship.  That means my point about magnetic nonequivalence is probably less important.

[DoctorDomo]

I think the problem all along is my sample wasn't concentrated enough in the NMR tube. Thats why my latest proton spectrum shows such a massive TMS peak, as well as a huge CHCl3 peak (the solvent was CDCl3). Some shifts for ethanol even showed up, I'm pretty sure there couldn't have been much ethanol in there since I wasn't using it for anything, could only be a trace impurity (from a pipette pump or contaminated CDCl3 bottle maybe). Its a good thing I decided to run HNMR spectra along with my FNMR spectra, or it woulda been harder to figure this out.