Logo

MonoCalc

/

IR Spectrum Peak Identifier

Chemistry
One peak per line: position, then optional s/m/w and br/sh. Ranges like 3300-2500 are read as broad bands.

Examples

Or drop a CSV/TSV/TXT (x, y) or JCAMP-DX file here. Up to 5 MB, read in your browser, never uploaded.

Filters out groups whose elements or DBE the formula rules out.

Only then can a missing band rule a group out.

A teaching and sanity-check aid, not a structure solver
IR correlation ranges overlap, and the fingerprint region (< 1500 cm⁻¹) is rarely conclusive on its own. Confirm with the formula, NMR, MS or a reference spectrum.

Enter IR peaks or load an example

Type one peak per line, e.g. 1715 s, 3340 br s or 3300-2500 broad, or import a spectrum file.

About This Tool

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:

  1. 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.
  2. 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.
  3. 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.
  4. Is there anything at 2300–2000 cm⁻¹? That window is nearly empty unless a nitrile, alkyne, isocyanate or azide is present.
RegionNameWhat to look for
4000–2500 cm⁻¹X–H stretchesO–H (broad), N–H, ≡C–H, sp² and sp³ C–H, aldehyde C–H
2500–2000 cm⁻¹Triple bondsC≡N, C≡C, N=C=O, azides; CO₂ artifact near 2350
2000–1500 cm⁻¹Double bondsC=O (the strongest clue), C=C, C=N, aromatic ring, N–H bends
1500–400 cm⁻¹FingerprintC–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.

Mark your broad bands
Enter a broad band as a range such as 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.

Frequently Asked Questions

Is the IR Spectrum Peak Identifier free?

Yes, IR Spectrum Peak Identifier is totally free :)

Can I use the IR Spectrum Peak Identifier offline?

Yes, you can install the webapp as PWA.

Is it safe to use IR Spectrum Peak Identifier?

Yes, any data related to IR Spectrum Peak Identifier only stored in your browser (if storage required). You can simply clear browser cache to clear all the stored data. We do not store any data on server.

How does this IR peak identifier work?

Each peak you enter is compared with a correlation table of about 100 bond vibrations. A peak scores on position (with your tolerance), intensity and shape, and each functional group then needs its required bands, gains confidence from supporting bands, and loses it when a stronger group already explains the same peak.

What does "ruled out" mean, and why does it need a complete peak list?

A group is ruled out only when one of its diagnostic bands, such as the strong C=O of a carbonyl or the C≡N of a nitrile, has no compatible peak in its region. That logic is only valid if you listed every significant peak, so turning off "This list includes every significant peak" stops anything from being ruled out.

How do I tell an alcohol O–H from a carboxylic acid O–H?

An alcohol gives a broad band centred near 3300–3400 cm⁻¹ with no C=O. A carboxylic acid gives a very broad band spanning 3300–2500 cm⁻¹ that overlaps the C–H stretches, together with a C=O near 1700–1725 cm⁻¹ and a C–O near 1300 cm⁻¹. Mark broad bands with "br" or enter them as a range like 3300-2500 so the tool can tell them apart.

Which spectrum files can I import?

Two-column CSV, TSV or TXT files (wavenumber or µm against %T, transmittance or absorbance) and JCAMP-DX files with XYDATA in plain or compressed (SQZ/DIF/DUP) form, XYPOINTS or a PEAK TABLE. Files up to 5 MB are read in your browser and never uploaded; picked peaks are written into the peak list for you to edit.

Why does the fingerprint region give low-confidence assignments?

Below about 1500 cm⁻¹ many C–C, C–O, C–N and bending vibrations overlap, so a single band there rarely proves a group. Use it to confirm what the 4000–1500 cm⁻¹ region suggests, or to read substitution patterns from out-of-plane C–H bands.

Can this tool determine the full structure of my compound?

No. IR shows which functional groups are likely present, not how they are connected. Combine it with the molecular formula (which adds element and DBE filters here), NMR and mass spectrometry, or compare against a reference spectrum.