A Tiny Compass in Every Nucleus
Earlier in this rung you met two detectives. Mass spectrometry weighs a molecule and watches it shatter, and infrared tells you which functional groups are present by how bonds vibrate. Both are powerful, but both are a bit indirect — they report on pieces, not on the whole map. Now meet the detective that draws the map itself: [[org-nmr-spectroscopy|nuclear magnetic resonance]], NMR, and in particular the proton (1H) version, which reports on the hydrogen atoms threaded all through an organic skeleton. More than any other single technique, this is how a chemist looks at an invisible molecule and says, with confidence, "that is what I made."
Where does the signal come from? A hydrogen nucleus is a single proton, and a proton behaves like a tiny spinning magnet — it has a property called spin, so it acts like a microscopic compass needle. Out in the open, those needles point every which way. But slide the sample into the huge magnet of an NMR spectrometer and each proton-compass must choose: align *with* the strong field (slightly lower energy) or *against* it (slightly higher energy). The two states are barely separated in energy, and that tiny gap is the whole game.
Now hit the sample with radio waves of exactly the right frequency, and protons absorb energy and flip from the lower state to the higher one — they *resonate*. "Magnetic" because of the magnet, "nuclear" because it is the nucleus flipping, "resonance" because absorption happens only at the matching frequency. The instrument records which frequency made which protons flip, and that record is the spectrum. The deep, beautiful fact is that not every hydrogen flips at quite the same frequency — and the small differences are dictated by the electrons surrounding each hydrogen. The molecule's own structure tunes each proton's note.
Clue One — Chemical Shift: Where the Peak Sits
Electrons are themselves moving charges, so the cloud around each hydrogen sets up its own tiny magnetic field. In most positions that little field opposes the big external one, so the proton feels a slightly weaker field than the magnet actually applies. We say the proton is shielded. A shielded proton needs a slightly different frequency to flip, and that horizontal position in the spectrum is its [[org-chemical-shift|chemical shift]], reported in parts per million (ppm) on a scale that runs from about 0 on the right to 12 on the left.
So the rule is one of [[shielding-deshielding|shielding and deshielding]]. Pull electron density away from a hydrogen — by hanging an electronegative atom like oxygen or chlorine nearby — and you thin out its protective cloud. With less shielding, the proton feels closer to the full external field, flips at higher frequency, and its peak slides to the *left* (higher ppm, called "downfield"). Hydrogens on a plain carbon chain, fat with electrons, sit shielded on the *right* near 0.9 ppm; a hydrogen on a carbon next to oxygen, like the CH3 of an ether, is dragged downfield to roughly 3.3 ppm. Each kind of environment lands in a fairly predictable neighbourhood, so the shift alone often names the functional surroundings of a hydrogen.
Clue Two — Integration, and Clue Three — Equivalence
The second clue is the simplest and one of the most useful. The *area* under each peak — not its height — is proportional to the number of hydrogens producing that signal. This is [[nmr-integration|integration]], and modern instruments print it as a number or a stepped curve over each peak. If one signal integrates to twice the area of another, it stands for twice as many hydrogens. Integration never gives you an exact count on its own; it gives a *ratio*, say 3 : 2 : 1. You combine that ratio with the molecular formula (often handed to you by mass spectrometry) to pin down the real numbers — a 3 : 2 : 1 ratio in a six-hydrogen molecule means 3, 2, and 1 hydrogens.
But integration only makes sense once you grasp the third clue: equivalence. NMR does not show one peak per hydrogen atom; it shows one signal per *kind* of hydrogen environment. Two hydrogens are equivalent if the molecule's symmetry makes their surroundings identical — and equivalent hydrogens resonate at exactly the same shift, merging into a single signal. The three hydrogens of a methyl group spin so fast around their bond that they are always equivalent and give one peak counting as three. The clearest test is the one you learned for stereochemistry: if swapping two hydrogens gives back an indistinguishable molecule, they are equivalent.
So the *number of distinct signals* counts how many different hydrogen environments the molecule has, and integration tells you how crowded each environment is. Take ethanol, CH3CH2OH. Its three carbons-and-oxygen hold three different kinds of hydrogen — the methyl CH3, the methylene CH2, and the lone OH — so you see exactly three signals, integrating 3 : 2 : 1. Before you ever read a single shift value, the count of peaks has already told you the molecule splits into three hydrogen worlds.
Clue Four — Splitting: Counting the Neighbours
Look closely at a real ethanol spectrum and the peaks are not simple single lines. The CH3 signal is split into three little peaks (a triplet) and the CH2 signal into four (a quartet). This is [[spin-spin-splitting|spin-spin splitting]], and it is the cleverest clue of the four, because it reports on a proton's *neighbours*. Each neighbouring hydrogen is itself a tiny magnet that can point with or against the field. A proton therefore feels not just the external field but the small extra nudge of every adjacent hydrogen-magnet, and the different combinations of those nudges split its single line into a cluster.
Counting the combinations gives the famous n+1 rule: a proton with *n* equivalent hydrogen neighbours on the *adjacent* carbon is split into *n*+1 peaks. In ethanol the CH3 has two neighbours on the CH2 carbon, so 2+1 = 3, a triplet; the CH2 has three neighbours on the CH3 carbon, so 3+1 = 4, a quartet. So splitting literally counts the hydrogens next door. One caution: equivalent hydrogens do not split *each other* — the three methyl hydrogens are identical and stay silent toward themselves; only neighbours on a *different* carbon split a signal.
Ethanol CH3 - CH2 - OH
CH3 : neighbours on next carbon = 2 (the CH2)
=> 2 + 1 = 3 lines -> TRIPLET
CH2 : neighbours on next carbon = 3 (the CH3)
=> 3 + 1 = 4 lines -> QUARTET
line heights follow Pascal's triangle
triplet 1 : 2 : 1 quartet 1 : 3 : 3 : 1There is even a fifth layer of detail. The *spacing* between the split lines, measured in hertz, is the [[org-coupling-constant|coupling constant]], written J. Two groups of hydrogens that are splitting each other must share the same J — that matched spacing is what lets you pair up partners across a spectrum and confirm they really are neighbours. Honest caveats: the plain n+1 rule assumes all the neighbours are equivalent to one another, which is not always true (neighbours on two different carbons can split a signal into more complex patterns), and OH or NH hydrogens often exchange so fast that they appear as a lone unsplit singlet. The rule is your reliable first pass, not the last word.
Putting the Four Clues Together
The real magic is that the four clues cross-check one another. Each peak tells you a position (what kind of neighbourhood), an area (how many hydrogens of that kind), and a splitting (how many hydrogens sit next door). Lay them side by side and the carbon skeleton starts to assemble itself like a jigsaw whose edges are forced to match. Here is the routine you would run on an unknown that mass spectrometry says is C3H6O2.
- Start outside NMR: from the formula C3H6O2 compute the degree of unsaturation. It comes to one, meaning one ring or one double bond — keep an eye out for a C=O.
- Count the signals: three peaks means three kinds of hydrogen environment. Read their integration ratio — here 3 : 2 : 1 over six total hydrogens, so 3, 2, and 1 hydrogens.
- Read the shifts: the 3H peak near 1.2 ppm is shielded (a CH3 on plain carbon); the 2H peak near 4.1 ppm is dragged downfield, the tell-tale of a CH2 wedged next to an oxygen; the 1H near 11 ppm, far downfield, screams a carboxylic-acid OH.
- Read the splitting: the CH3 is a triplet (2 neighbours on the CH2) and the CH2 is a quartet (3 neighbours on the CH3) — the classic ethyl fingerprint. The pieces -CH2CH3 next to oxygen, plus a C=O and an acidic OH, assemble into ethyl formate or, accounting for the acidic proton, the connectivity HCOO-CH2CH3. The four clues, cross-checked, hand you one structure.