Proton NMR Spectroscopy

[Stick]

I'm really bad at this...

I understand how the technique works and I know how to interpret the spectra that we are often provided with. I get the significance of chemical shift, number of peaks, peak heights/areas and peak splitting. I just can't figure out what the damn molecule looks like when I'm asked to draw its structure (which caused me to lose a few marks on my Chemistry trial exam at school). I know that it's really a matter of problem-solving, but I've practiced it heaps and it's not really getting any better. Something that particularly confuses me is the fact that so many of the chemical shift values are not quite exact with what's in the data book and vary at times. My teacher didn't really cover the problem-solving aspect of this topic, so I was wondering if there are any tips anyone could provide. Perhaps there's a methodical approach to solve these questions step-by-step? Any help would be great. Thanks.

[lzxnl]

Example? It really comes with experience when you notice certain patterns.

[Stick]

I don't have a specific example to reference, but I just get them wrong in general because I don't really know what to do.

[Stevensmay]

I found this helped alot.
http://www.youtube.com/watch?v=2ul2oSdjnvM
Second half of the video.

http://www.chem.ucla.edu/~webspectra/
If you need more questions to work on.

[P0ppinfr3sh]

Well, a small tip I can give is to look at the molecular formula you're given. For example:
If i've got something like C3 H6 O2, i know i've probably got a carboxylic acid or an ester . From this point you could look at the chemical shifts of the peaks in the 1H NMR spectrum and determine whether its an ester or carboxylic acid. From there, you could probably look at peak splitting, hydrogen environments and/or perhaps the relative peak heights to determine the complete structure of the molecule. e.g. If you determined it was an ester, and there is a quartet, then you know the molecule is ethyl methanoate.
If you find that its neither a carboxylic acid nor an ester, swear loudly. (But at least you've eliminated some options right?)
 

Hahaha me too.

Although I've kinda given up on chem, (raising one of my top four study scores by just 1 would be equivalent to me raising my chemistry study score by 10.. No difference between 30 and 40 on chem for me) this has really irked me.

I see the peak at around 3.3 and I think to myself:
  "Hmm... It could be a R-CH2-X... Or a R-CH2-OH... OR a R-C-NHCH2R double bond to O...  One in three gues- OH WAIT it could be R-O-CH3 OR R-O-CH2R... Whatever one in five guess, same as spesh multiple choice... HEY WTF R-O-H CAN BE ANYWHERE FROM 1-6!?!? SAME AS R-NH2!?!?
...Fuck it, next question."


I know how you feel Stick, haha.

I feel this way too..........................sometimes.

[lzxnl]

Wait...you guys actually bother with chemical shifts?

I don't even use them to start off with. The majority of the compound's structure can be inferred without the chemical shifts. For instance, you can work out how many double bonds there must be in the compound. Remove all of the oxygens and look at the remaining hydrocarbon (if it only has carbon, oxygen, hydrogen). For instance, C2H4O must have a double bond somewhere as C2H4 is ethene. Oxygens do not contribute to the double bond count.

Then you look at the splitting and the areas if they give you them. That should generally be enough info.

[jgoudie]

C₂H₄O could be a ethanal H₃C-CHO (2 carbon aldehyde).  Ketones and aldehydes, whilst you would not be expected to name them, their structure could be deduced. Chemical shifts are the last resort fair enough, but should not be exempt from things.

Wait...you guys actually bother with chemical shifts?

I don't even use them to start off with. The majority of the compound's structure can be inferred without the chemical shifts. For instance, you can work out how many double bonds there must be in the compound. Remove all of the oxygens and look at the remaining hydrocarbon (if it only has carbon, oxygen, hydrogen). For instance, C2H4O must have a double bond somewhere as C2H4 is ethene. Oxygens do not contribute to the double bond count.

Then you look at the splitting and the areas if they give you them. That should generally be enough info.

[lzxnl]

C₂H₄O could be a ethanal H₃C-CHO (2 carbon aldehyde).  Ketones and aldehydes, whilst you would not be expected to name them, their structure could be deduced. Chemical shifts are the last resort fair enough, but should not be exempt from things.

What could C2H4O be? It could be...ethenol or ethanal. I cannot think of a third possibility. There'd be a world of difference in the hydrogen spectrum as ethanal has two hydrogen environments (CH3CHO) and ethenol has three (CH2=CHOH). That doesn't require chemical shifts.

Chemical shifts would only separate REALLY similar compounds, and by then the problem is pretty much solved anyway.

[jgoudie]

True, chemical shifts are a last resort, but my point was that could can't just plainly say the oxygens play no part in determining double bonds.

What could C2H4O be? It could be...ethenol or ethanal. I cannot think of a third possibility. There'd be a world of difference in the hydrogen spectrum as ethanal has two hydrogen environments (CH3CHO) and ethenol has three (CH2=CHOH). That doesn't require chemical shifts.

Chemical shifts would only separate REALLY similar compounds, and by then the problem is pretty much solved anyway.

[lzxnl]

I think you misunderstood. You can remove the oxygens when considering how many double bonds the molecule has as a whole. I never said where these double bonds had to be. In both ethenol and ethanal, there is a double bond.
And before someone else comes along, I'm going to cover my bases by saying benzene counts as four double bonds and a ring counts as one.