[organicmechanic]

Something that has been bothering me for a while, and I would like to know where others may weigh in on this because I cannot find anything written/posted about it: So let's say you're going to write up experimental data about a new compound you've synthesized, and you have a large (>10) amount of aliphatic protons. You know based on previous reactions and mass spec exactly how many protons "should" be there, but there's grease and water buried in that big signal. So what if you know there should be 16 protons in the big grease pile but it actually integrates for 25. In your peak list do you list this as 16 protons (what "should" be there) or 25 (what the big thing actually integrates for)? Both options seem viable to me, with potential problems for each. If you call the signal 25H, you're violating the molecular formula of the compound that you synthesized. If you call the signal 16H, you're not reporting exactly what is in the spectrum. Might seem like a minor thing but there are enough errors out there in the literature, I would like to avoid contributing any of my own.

[phth]

For overlapping peaks there isn't much that can be done to get an accurate integration because overlap causes the number to skew.  I would just go from the cosy to see how many of the target protons are under a peak.  Watch out for water because it will change the chemical shift (ternary mixtures).  Water can be removed using an azeotrope.  If the molecule is fragile, then cyclohexane (mixture boils ~60°C).  Benzene, toluene, MeCN etc. will work. To get the best results, I would recrystallize.  The final thing that can be done is to use an instrument with a larger magnet to get the peaks to separate more, and to use a different solvent than CDCl₃.  The larger the diamagnetic current of the solvent, the more peak separation will be seen.  CDCl₃<DMSO<acetone<nitromethane.  The samller π to π* gap the better. 

[Irlanur]

Since this is about reporting EXPERIMENTAL Data, in my opinion, the only viable approach is to report what is there, not what should be there. (I mean one could also just publish the spectra....)

[TheUnassuming]

Agreed with Irlanur.  If you are reporting experimental data, especially for publication, you MUST report what you get and not what you want or expect.  Doing otherwise can get you into a lot of trouble.

For publication, especially in a chemistry related journal, you need to clean up the spectra more before calling it good.  Water/grease in your compound will throw off the weights/yields you report, as well as not actually proving you made what you claim.  If there is water in your spectra that is the easiest to take care of.  Grease will require you to re-purify your compound and take extra care to keep you compound clean. 

Once you have the water and grease taken care of, I've found it helpful to use a proton that integrates to 2-4 for the zeroing to make the large peaks integrate correctly.

[Dan]

You know based on previous reactions and mass spec exactly how many protons "should" be there, but there's grease and water buried in that big signal. So what if you know there should be 16 protons in the big grease pile but it actually integrates for 25.

Well, 16 vs 25 is a big difference. I would clean up the sample and get a better spectrum.

Sometimes it's not possible - when the peak of interest overlaps with the NMR solvent residual peak, or the water peak in DMSO/acetone/methanol. In these cases I report what should be there (but an integrated spectrum, sometimes annotated, is always included as well).

A small discrepancy, say 10%, is within the limits of experimental error. If I have a Boc integrating at 9.9, I report it as 9H (but an integrated spectrum is also included in the SI). When the discrepancy is >50%, as in the example you gave, then it looks dodgy - 16H is not a reasonable interpretation of a 25 integral.

[wildfyr]

You can often get rid of grease by running the compound on a plug of silica gel with hexane, this will push off grease and unless you have something very nonpolar the product won't move. you can then switch to DCM or acetone and get the product off the plug. check this process by TLC first.

To remove water, just dissolve it in an organic like DCM and let it sit over MgSO4 or sieves, then rotovap.

Finally, if things are wonky and unpurifyable, you should state that, and perhaps include the spectrum along with some analysis. Usually with enough work you can purify something enough to make it look at least presentable.

[Irlanur]

You could also selectively pre-saturate the unwanted signal

[zuriel]

I often remove grease by dissolving my compound in MeCN and washing with hexane. You could try doing this with cold solvents to avoid losing some of your product. It might also be an idea to run your compound through a hexane column using distilled hexane (I used to use calcium chloride to dry hexane but found that just collecting it without using calcium chloride was equally as good).