If you want to know what’s really going on inside your NMR machine, read “Spin Dynamics” (2nd Edition) by M. H. Levitt. It weighs in at 700 pages, contains a 15 page symbol table (with at least 20 symbols/page). It will probably tell you more than you want to know. I don’t claim to have read whole thing, just the first 100 pages or so which describes the machine, the essential NMR experiment, and the physics. Be prepared for detail — including how the Fourier transform of the data produces the spectra we see. If I could inhabit another universe where time expands, I’d read the whole thing.
mnemonic loadd of bs
It’s far from perfect, so if you can think of something better, please post a comment.
p. 270 — “This 10 ppm scale is not the same as any part of the 13C NMR spectrum. It is at a different frequency altogether.” This is because even though they use the same reference compound (Me4Si), the split between the two energy levels (with and against the field) of 13C and 1H are quite a bit different — so the frequency of the transition (energy is proportional to frequency by Planck’s constant) will be quite different.
It’s worth while thinking (again) why Me4Si should be used. Essentially it’s because the electronegativity of Si is 1.9 and that of carbon 2.5 meaning that carbon in C – Si bonds has far more electrons close to the nucleus than carbon in C – C bonds, making it highly shielded. This means that the C – H bond has more electrons shielding the hydrogen.
p. 274 — “Rotation about single bonds is generally very fast” — well just how fast is it?
p. 278 — The pictures of a cyclophane and an annulene are not to be missed.
p. 292 — The color labeling HA in the chemical structure is incorrect, it should be that labelling HX. The colors are correct for the spectra.
p. 293 — dd should be in the table of abbreviations as well as in the text.
p. 293 — The example shows why coupling is NOT through space (or the coupling constant for cis hydrogens on a double bond would be larger than trans as they are physically closer). However, I don’t see why coupling of bonds aligned parallel should be higher.
p. 293 — Double triplet is all very nice, but only 5 peaks are shown in the spectrum not 6 (although the middle peak of the quintet looks a bit funny). Some note should be made of this.
p. 296 — W-coupling. Something is fishy as in the pictures shown the angle between the two coupled H bonds is the same as it is in cis vicinal hydrogens (e.g. 120 degrees), yet the bonds are said to “line up in a zig-zag fashion to maximize interaction”.
p. 303 — Rotating structures in space using Jmol is almost as good as having a model to play with. The structure itself is rigid, perhaps later we will be able to play with conformations of a given molecule (rather than rotating the molecule as a whole).
p. 306 — They do show rotation of a C – C bond beween two sp3 hybridized carbons. But you can’t change the conformation of the other two sp3 hybridized carbons, although you can rotate the whole molecule in space to see how the rotation looks from various angles — pretty neat !
p. 308 — The periodic table is found in the back of the book, not the front.
p. 320 — Watching a wire model of BINAP rotate in space is particularly impressive. Notice how, at times, during the rotation the drawing appears to be rather flat (it’s a drawing after all), while at other times the molecule appears to jump off the page into three dimensional space. This isn’t your brain playing tricks on you, but it’s your brain doing what your brain does best, inferring 3 dimensionality from the 2 dimensional projection of an image on your retina. It’s worthwhile stopping the rotation, and seeing how two dimensional the images actually are. Your brain is inferring 3 dimensionality by comparing images received over time. Impressive, isn’t it?
p. 323 — I don’t think “Why nature uses only one enantiomer of most important biochemicals is an easier question to answer” — I don’t think this was answered well. They should have said what the ‘enormous number of diastereomers’ would mean. What do you think? (answer below)
Chemiotics II 
Comments
I’m a major fan of Levitt’s text. I kind of wish I had first stumbled across it when I was first *really* learning NMR (above and beyond the brief exposure to it as an undergraduate), as I think it does the best job of getting someone to a point where they can tackle a more specialized text on biological or solid-state NMR, as well as start doing research.
TMS also serves as an acceptable reference for silicon-29 NMR. On a more practical note, TMS is something that you can add to a solution NMR sample in organic solvent – it’s rather inert and can be evaporated. For solution NMR of biomolecules in aqueous solution, DSS (sodium salt of 2,2-dimethyl-2-silapentane-5-sulphonic acid) is used as it’s water-soluble in contrast to TMS. For solids NMR, adamantane is a popular chemical shift reference – although with solids, there are other complications at work there. There are, of course, known offsets to reconcile any issues between the various schemes. Realistically, one uses whatever is more suitable and then just plugs in the offset to get you to the TMS standard.
Given that I am sometimes happy just to see a signal when doing NMR, I don’t spend much time thinking about how to rationalize proton chemical shifts. So take this with a grain of NaCl. Are you observing different shifts for ortho, meta, and para protons upon substitution, and not just a collective translation up- or downfield of the spectrum as a whole? Also, I presume that the chemical shift data for benzene versus these various aromatic derivatives were all done identically – one can get solvent-induced shifts, especially if using aromatic solvents. In general, since I don’t know specifically what molecules they’ve got listed in the book, I will just mention this – the more complicated and messy-looking the aromatic compound, the more difficult it’s going to be to intuit the chemical shift differences. Even the tabulated empirically-based charts that are out there for doing some fairly detailed estimates begin to fail if the molecule gets too messy.
I really do like the Jmol complement they’ve got going there. Although – I may just be missing it – they don’t have the space-fill option for all of the figures. I really think it can be a helpful thing to see along with the skeleton representation.