[gorillagrrl]

I am looking at an 1H NMR spectra where not all the integration values are whole numbers. (More specifically, one is a whole number and two are not, and neither of them look very "friendly.") I have never encountered a problem like this before because in all previous problems the integration numbers were already whole. Is there a method to converting these into whole numbers?

[thetada]

I haven't experienced that either, but I would round each figure to the nearest 0.5. Remember, the integration values give the proportion of hydrogen atoms rather than actual numbers. If you get 1:1:1.5, I'd go with 2:2:3 (or 4:4:6 etc)

[Borek]

Integration numbers are never guaranteed to be right. Usually they are proportional to the number of protons, but like every experimentally measured quantity they are sometimes in some way wrong. The most common problem is, the value depends on the integration interval, and sometimes choosing the lower and upper limit is quite tricky. But there can be other reasons why it doesn't work the way we expect. Sorry, I don't remember any further details.

[gorillagrrl]

I haven't experienced that either, but I would round each figure to the nearest 0.5. Remember, the integration values give the proportion of hydrogen atoms rather than actual numbers. If you get 1:1:1.5, I'd go with 2:2:3 (or 4:4:6 etc)

So if my numbers are 0.6330, 1.0000, and 2.7773, I should round them to 0.5, 1, and 3 respectively?

[mjc123]

That looks suspiciously close to 2:3:8 (though bear in mind Borek's comments). Do you know what your compound is supposed to be? How many protons are you expecting?

[tomek]

Is it an experimental spectrum we're talking about? If it is than you almost never get perfect integrals. This has to do with the relaxation time of nuclei mostly. For example protons on amines give broad peaks which usually integrate appreciatly below the number of actual protons. If it's a "student" spectrum then simply find the common denominator and bring those values to integers.

[gorillagrrl]

That looks suspiciously close to 2:3:8 (though bear in mind Borek's comments). Do you know what your compound is supposed to be? How many protons are you expecting?

I was not given what compound I was supposed to find, but I actually ended up getting 2:3:9 for the integration ratio and found the structure.

[thetada]

It's a bit late now, but another time I'd divide each figure by the smallest figure (ie divide all 3 by 0.633) like you would when you need to simplify the figures that come out of an empirical formula calculation. In this case the ugly 2.7773 figure comes out at 4.5 on the nail, suggesting that all figures should be doubled to give the ratio you got.