JOVANA
Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
All guides

Carbon NMR & Solving Structures

Proton NMR mapped the hydrogens; now meet the spectrum that counts the carbons themselves. Then watch mass, IR, and NMR converge as we crack one real unknown together — the full detective kit in action.

The Spectrum That Counts Carbons

In the previous guide, proton NMR handed you a beautiful but indirect picture of the molecule: it showed you every kind of hydrogen, counted them by integration, and let neighbouring hydrogens whisper to each other through splitting. But hydrogen lives on the outside of a molecule, decorating a skeleton it never lets you see directly. The skeleton is carbon — and [[carbon-13-nmr|carbon-13 NMR]] lets you look straight at it. Instead of listening to the C-H protons, you now listen to the carbon nuclei themselves, one signal for each chemically distinct carbon in the molecule.

Why carbon-13 and not plain carbon? Because the common carbon isotope, carbon-12, has no nuclear magnetism at all — it is invisible to NMR. Only carbon-13, which makes up barely 1 in 100 carbon atoms in nature, has the spinning nucleus that a magnet can grab. That scarcity is the catch: with so few responsive nuclei, the signal is naturally weak, so a carbon spectrum takes far more sample or far more scans than a proton spectrum to pull a clean picture out of the noise. The reward for that patience is a clean count of the carbon framework.

The scarcity buys an unexpected gift, too. Because two carbon-13 nuclei sitting side by side is statistically rare (1 in 100, times 1 in 100), carbons essentially never split each other. And in the standard 'proton-decoupled' experiment, the machine deliberately silences the carbon-hydrogen splitting as well. So an ordinary carbon-13 spectrum is gloriously simple — a forest of single sharp lines, one per distinct carbon, with none of the multiplet thicket you learned to untangle in proton NMR. You count the lines, and you have counted the kinds of carbon.

Reading the Carbon Map

The same logic of position that you mastered for protons carries straight over. A carbon's [[org-chemical-shift|chemical shift]] — where its line lands on the scale — reports on the electron cloud and neighbours around that carbon, by exactly the shielding and deshielding you already understand. The numbers are simply spread across a wider range, roughly 0 to 220 ppm instead of the proton's 0 to about 12. A carbon buried in a plain alkyl chain sits low, near 10 to 40 ppm; a carbon bearing oxygen climbs to 50 to 90; aromatic and alkene carbons crowd 110 to 150; and a carbonyl carbon, the C=O itself, sails far downfield to 160 to 220, almost alone out there.

There is one more thing the simple spectrum quietly tells you, and it is a powerful cross-check with the degree of unsaturation you compute from the formula. A line out beyond 160 ppm almost always means a carbonyl carbon — a C=O — and so it doubles as a witness that one of your pi bonds is a carbonyl group. A cluster of lines in the 110 to 150 region announces double bonds or an aromatic ring. Read the carbon shifts and you are reading, peak by peak, how the molecule spends its degrees of unsaturation.

DEPT and a Glance at 2D

The plain carbon spectrum tells you how many kinds of carbon there are, but not how many hydrogens hang off each one. That is exactly the gap [[dept-nmr|DEPT]] fills. DEPT (Distortionless Enhancement by Polarization Transfer) is a cleverly timed pulse sequence that sorts the carbons by how many hydrogens they carry: it makes CH and CH3 carbons point one way, CH2 carbons point the opposite way, and quaternary carbons (those bearing no hydrogen at all) vanish entirely. Compare a DEPT trace against the full spectrum and any carbon that appears in one but is missing from the other must be a bare, quaternary carbon.

DEPT-135 readout (peaks up / down / absent):

  CH3  -> UP        CH2  -> DOWN
  CH   -> UP        C    -> ABSENT (no attached H)

appears in full 13C spectrum but ABSENT in DEPT  =  quaternary C
(e.g. a carbonyl C=O, or a ring carbon bearing no H)
DEPT sorts carbons by attached-hydrogen count; the carbons that go missing are the quaternary ones — often the carbonyl or substituted ring carbons that matter most.

Beyond DEPT lie the two-dimensional experiments, and you only need the idea of them for now, not the machinery. A 2D spectrum spreads the peaks across a square map so that a spot's position pins down two facts at once — and the most useful ones draw lines of connection between nuclei. HSQC links each carbon to the hydrogen actually bonded to it; COSY links hydrogens that sit on neighbouring carbons; HMBC reaches across two or three bonds to stitch fragments together over a quaternary carbon or a heteroatom that 1D methods cannot bridge. Think of 2D NMR as the detective finally laying the witness statements side by side on one board and drawing string between the ones that match.

One Unknown, Solved End to End

Now let us put the whole kit to work on a fresh unknown and watch the witnesses converge — the convergence that is the heart of structure elucidation. A colourless liquid comes off a reaction. We take it to every instrument in turn and let each one narrow what the last left open, exactly the disciplined order from the opening guide of this rung.

  1. Mass spectrometry: the molecular ion sits at mass 88, and the isotope and fragment pattern read out the formula C4H8O2 — four carbons, eight hydrogens, two oxygens. This is the witness on weight and atom count.
  2. Free cross-check: DoU = (2*4 + 2 - 8) / 2 = 1. Exactly one ring or one pi bond hides inside. Both oxygens are 'free' atoms in the chain, since oxygen never enters the unsaturation count.
  3. IR: a strong, sharp band in the carbonyl region says C=O is present, and — decisively — there is NO broad O-H band. So this is not a carboxylic acid. A C=O plus two oxygens but no acidic O-H points hard at an ester linkage, -C(=O)-O-, which spends both oxygens and the single degree of unsaturation in one stroke.
  4. Carbon-13 NMR: four distinct lines, one per carbon, confirming all four carbons are inequivalent. One sits past 170 ppm — the ester carbonyl carbon. Two land in the 20 to 60 ppm region (ordinary chain carbons), and one is pushed up near 60 to 70 ppm, the tell-tale of a carbon bonded to the ester oxygen (O-CH2).
  5. DEPT to assign the carbons: the carbonyl carbon is ABSENT (quaternary, no attached H, exactly as expected). The carbon near 60 ppm points DOWN — it is a CH2. The remaining two carbons point UP — they are CH3 groups. So the heavy atoms are: O=C, an O-CH2, and two CH3 groups to distribute.
  6. Proton NMR seals it: a 3H triplet and a 2H quartet (the classic CH3-CH2 ethyl pattern, coupled to each other) plus a lone 3H singlet with no neighbours. Thread it together: the isolated CH3 sits on the carbonyl (CH3-C=O), and the ethyl group is the O-CH2-CH3. The one structure that satisfies every witness is CH3-C(=O)-O-CH2CH3 — ethyl acetate.

Notice how no single witness solved it. Mass left a wide field of C4H8O2 isomers — it could have been butanoic acid, methyl propanoate, dioxane, or our ester. IR knocked out the carboxylic acids by the missing O-H. Carbon NMR and DEPT laid out the skeleton's pieces and flagged the quaternary carbonyl, and proton NMR's ethyl-and-singlet pattern chose, among the survivors, the one arrangement that fit. The answer is simply the only structure that contradicts none of them — and it balances atom for atom against the formula.

Two More Tools, and the Honest Limits

Two supporting witnesses round out the kit. The first is [[org-uv-vis-spectroscopy|UV-Vis spectroscopy]], which mostly stays quiet — until a molecule carries extended conjugation, a run of alternating double bonds or an aromatic system. Such a chromophore absorbs ultraviolet or visible light, and the longer the conjugated run, the longer the wavelength it soaks up. UV-Vis rarely pins a whole structure, but it is a sensitive yes-or-no on conjugation: a strong absorption means an extended pi system is present; near silence means it is not. It is the witness you call when you suspect your molecule has a conjugated backbone.

The second supporting tool answers a question all the others quietly assume: is the sample even one pure compound? Chromatography — thin-layer for a quick look, gas or high-performance liquid (GC, HPLC) for the careful version — separates a mixture by how strongly each component clings to a stationary surface versus a moving solvent, so different molecules travel at different speeds and arrive apart. It does not tell you what a compound IS, but it tells you how many things you have and whether your 'unknown' is clean. Every spectrum above silently assumes a single substance; if two compounds are present, every witness is testifying about a blend, and the case is corrupted from the start.

Be honest about where the kit runs out. Even a full set of spectra often cannot, on its own, fix stereochemistry — which enantiomer you hold, or the relative arrangement on a ring — and that usually needs specialised experiments (like NOE) or comparison against authentic known data. A weak or absent molecular ion can hide the formula. And these textbook unknowns are unusually obliging; a real natural product can take a team weeks, leaning hard on the 2D methods we only glanced at. The lesson of this whole rung is not that any one machine sees the molecule, but that patient, overlapping, cross-checked evidence — held to the unforgiving arithmetic of the formula — is how chemists earn the right to draw a structure and defend it.