IR Spectrum Peak Identifier – Read Functional Groups from Infrared Bands
An infrared spectrum records which bond vibrations a molecule absorbs. Because stretching frequencies depend on bond strength and the masses of the atoms involved, each functional group leaves bands in fairly predictable places. This IR spectrum peak identifier takes your peak list, or a spectrum file, matches every band against a correlation table and then weighs the evidence across all peaks to say which functional groups are present, likely, possible or ruled out.
How to interpret an IR spectrum step by step
Experienced chemists rarely read an IR spectrum peak by peak. They ask a few first-look questions, which the checklist under the results answers for you:
- Is there a strong band between 1850 and 1650 cm⁻¹? If so, the molecule almost certainly contains a carbonyl (C=O), and its exact position narrows the type.
- Is there a broad band between 3550 and 2500 cm⁻¹? A broad band centred near 3300–3400 suggests an alcohol; one sprawling across 3300–2500 and overlapping the C–H bands is the signature of a carboxylic acid.
- Are the C–H stretches above or below 3000 cm⁻¹? Bands just above 3000 come from sp² C–H (alkenes, aromatic rings); bands only below 3000 mean saturated sp³ C–H.
- Is there anything at 2300–2000 cm⁻¹? That window is nearly empty unless a nitrile, alkyne, isocyanate or azide is present.
| Region | Name | What to look for |
|---|---|---|
| 4000–2500 cm⁻¹ | X–H stretches | O–H (broad), N–H, ≡C–H, sp² and sp³ C–H, aldehyde C–H |
| 2500–2000 cm⁻¹ | Triple bonds | C≡N, C≡C, N=C=O, azides; CO₂ artifact near 2350 |
| 2000–1500 cm⁻¹ | Double bonds | C=O (the strongest clue), C=C, C=N, aromatic ring, N–H bends |
| 1500–400 cm⁻¹ | Fingerprint | C–O, C–N, C–X, out-of-plane bends; confirms rather than proves |
How the matching works
Each peak is compared with roughly a hundred entries from standard IR correlation tables. A peak inside a range scores 1; outside it, the score falls linearly to zero at your tolerance. If you give an intensity (s, m, w) or a shape (br, sh), mismatches lower the score: a strong peak where a weak band is expected is multiplied by 0.5, and a sharp peak where a broad O–H belongs is too. That is what separates the broad alcohol O–H from the sharp terminal alkyne ≡C–H, which both sit near 3300 cm⁻¹.
Functional groups are then judged on required and supporting bands. An ester needs both its C=O and its C–O stretch and is confirmed by a second C–O band; an aldehyde needs its C=O and the characteristic C–H near 2720 cm⁻¹. When a strong group already explains a peak, weaker readings of the same peak are demoted, so a 1688 cm⁻¹ band that belongs to a carboxylic acid is not also reported as a ketone.
3300-2500 br s, or add br. Without a shape the tool cannot tell a carboxylic acid O–H from an ordinary C–H stretch.Using the molecular formula and sample preparation
An optional molecular formula removes impossible groups: with no nitrogen there can be no amine, amide or nitrile, and the degree of unsaturation (DBE) decides whether a π bond, a triple bond or an aromatic ring (DBE ≥ 4) is possible at all. The sample preparation adds known artifacts, such as Nujol's C–H bands, water absorbed by KBr pellets and solvent bands from CCl₄ or chloroform, so they are not mistaken for your compound.
Importing a spectrum and picking peaks
You can drop a two-column CSV or a JCAMP-DX file, including compressed ASDF data. The trace is converted to absorbance with A = −log₁₀(T), lightly smoothed, and local maxima with enough prominence become peaks. Their relative size sets s, m or w, and their width at half prominence decides whether they are broad. The picked list lands in the text box, where you can correct it before the analysis runs.
Limits of IR interpretation
Correlation ranges overlap, and compilations disagree by 10–20 cm⁻¹ at their edges, which is why a tolerance setting exists. Hydrogen bonding, conjugation, ring strain and the physical state all shift bands. The fingerprint region below 1500 cm⁻¹ is crowded and rarely conclusive by itself. Treat the verdicts as a structured second opinion, and confirm a structure with NMR, mass spectrometry or a reference spectrum.