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

Infrared Spectroscopy

Every bond in a molecule is a tiny spring, and each kind of spring jiggles at its own frequency. Shine infrared light through a sample and the spectrum that comes out is a quick, honest readout of which functional groups are there — and which are not.

Bonds Are Springs That Jiggle

By now you can read a mass spectrum for a molecular weight and a few fragments. Infrared spectroscopy (IR) answers a different, complementary question: not how heavy the molecule is, but what kinds of bonds it contains. The trick is to stop picturing a bond as a rigid stick. A real chemical bond behaves like a tiny spring connecting two masses — the two atoms — that is constantly stretching longer and shorter, and bending side to side, even in a molecule sitting quietly at room temperature. These motions are the molecular vibrations, and they are the whole game.

Like any spring, each bond vibrates at a characteristic frequency, and Hooke's law tells you what sets it. Two things matter: how stiff the spring is (the bond strength) and how heavy the masses are (the atoms). Stiffer springs vibrate faster — so a strong triple bond jiggles faster than a double bond, which jiggles faster than a single bond. Lighter masses also vibrate faster — which is why any bond to a tiny hydrogen atom, like O-H or C-H, vibrates at a very high frequency. Hold onto those two rules; they explain the entire layout of an IR spectrum before you memorize a single number.

How Light Pumps a Vibration

Infrared light is just light with less energy than visible red — its frequency happens to match the frequency at which bonds vibrate. When you shine a whole spread of IR frequencies through a sample, a bond will absorb a photon only when that photon's frequency matches its own natural jiggle frequency, the way a child on a swing only goes higher if you push in rhythm. Absorb that photon and the bond vibrates harder. The instrument records which frequencies got swallowed, and a dip appears in the spectrum exactly there. So every absorption band is a bond announcing 'this is the frequency I vibrate at.'

The horizontal axis of an IR spectrum is labelled wavenumber, in units of cm-1, running by convention from about 4000 on the left down to 400 on the right. Wavenumber is just a stand-in for frequency, so high on the left means fast-vibrating bonds, low on the right means slow ones. The vertical axis is usually percent transmittance, which means the peaks point downward — a strong absorption is a deep valley, not a tall mountain. It looks upside-down at first, but the logic is simple: the deeper the dip, the more light that bond drank.

The Diagnostic Region: Tell-Tale Stretches

The high-frequency left half of the spectrum, roughly 4000 down to 1500 cm-1, is the diagnostic or functional-group region, and it is where you do almost all your detective work. Because light bonds and stiff bonds live here, this is exactly where the most informative functional groups sing out. Reading it well is mostly a matter of learning a short list of tell-tale stretches and what each one means — and, just as importantly, what their absence rules out.

Start with the bonds to hydrogen, up near 3000-3700 cm-1. An O-H stretch of an alcohol is unmistakable: a very broad, rounded band centered around 3300 cm-1, broadened because hydrogen bonding smears the spring stiffness over a range. A carboxylic acid O-H is even more dramatic — an enormous, draping band from about 2500 to 3300 cm-1 that swallows everything beneath it. An N-H of an amine sits in a similar place near 3300-3500 cm-1 but is narrower and weaker, often showing one notch for a secondary amine or two for a primary one. Just to their right, around 2850-3000 cm-1, sit the ordinary C-H stretches that nearly every organic molecule has — so this band confirms almost nothing by itself.

Now the headline peak of all of IR: the C=O stretch, a strong, sharp, hard-to-miss spike near 1700 cm-1. If you see one bold absorption around 1700 and nothing big nearby, a carbonyl is almost certainly present. Its exact position is a fine clue to which carbonyl you have — an aldehyde or ketone near 1715, an ester a bit higher near 1740, an amide lower near 1650 — but at this stage just learning 'sharp peak near 1700 means C=O' already lets you crack many problems. Two more stretches round out the headline list: a sharp, medium C#N stretch of a nitrile in the quiet window near 2250 cm-1, and a weak C=C of an alkene near 1650 cm-1, often faint because that bond is barely polar.

wavenumber (cm-1)   bond / group        look
3200-3550 (broad)   O-H  alcohol         wide, rounded
2500-3300 (huge)    O-H  carboxylic acid  enormous, drooping
3300-3500           N-H  amine            1-2 sharp notches
~2250               C#N  nitrile          sharp, lonely
~1700               C=O  carbonyl         strong, sharp <- headline
~1650 (weak)        C=C  alkene           faint
A starter cheat-sheet for the diagnostic region: position roughly tells you the bond, and shape (broad vs sharp, strong vs weak) confirms it.

The Fingerprint Region

Below about 1500 cm-1 lies the fingerprint region, a dense forest of peaks that comes from bending motions and from single-bond stretches of the carbon skeleton — C-C, C-O, C-N. These bonds are all heavy, similar, and coupled together, so their vibrations mix into a complicated pattern that no simple table can decode peak by peak. Trying to assign each notch here to one bond is usually a fool's errand, and it is honest to admit it.

But the region earns its name precisely because it is complicated. The exact tangle of fingerprint peaks is unique to each compound, like a human fingerprint — even two close isomers differ here. So although you do not read it band by band, it has a powerful use: if your unknown's full spectrum, fingerprint and all, lies perfectly on top of a known reference spectrum, you have proof they are the same compound. The diagnostic region tells you what groups are present; the fingerprint region confirms the exact identity by overall match.

Reading a Spectrum Like a Detective

In practice IR is a fast first pass, not a full solution. In a minute you can take a sample, run a spectrum, and ask a handful of yes/no questions that drastically narrow what you are holding. The real power is as much in the peaks that are missing as the ones present: no broad O-H means no alcohol or acid; no sharp peak near 1700 means no carbonyl at all, which instantly rules out aldehydes, ketones, acids, esters, and amides in one stroke. Here is the routine.

  1. Look near 1700 first. A strong sharp peak there means a C=O is present; its absence rules out every carbonyl-containing group at once.
  2. Scan 2500-3500 for broad bumps: a wide rounded O-H says alcohol; a huge drooping one says carboxylic acid; sharp narrow notches say N-H of an amine.
  3. Check the quiet window near 2250: a lone sharp peak there is the C#N of a nitrile (or, slightly different, a C#C of an alkyne).
  4. Combine the IR group list with the molecular formula from MS to count the degree of unsaturation, then hand the surviving candidates to NMR.