How to Read an Infrared Spectrum: Key Features and Interpretation Basics

Updated

An infrared (IR) spectrum plots wavenumber (cm⁻¹) on the x-axis against transmittance (%) or absorbance on the y-axis. To read one, locate the functional group region (≈4000–1500 cm⁻¹) for identifiable stretches, then the fingerprint region (≈1500–400 cm⁻¹) for the compound's overall pattern. The most common misreading is forgetting that transmittance peaks point down (low transmittance = strong absorption), while absorbance peaks point up. This guide covers axes, regions, band assignments, and how IR differs from Raman — enough to interpret a routine spectrum correctly.

Axes and Units: What the Numbers Mean

Axis Typical unit Direction Meaning
X wavenumber (cm⁻¹) High → low, left to right Proportional to photon energy; 4000 cm⁻¹ is higher energy than 400 cm⁻¹
Y (transmittance) % T 100% at top A dip means light was absorbed
Y (absorbance) A (unitless) 0 at bottom A peak means light was absorbed

Wavenumber is the reciprocal of wavelength (cm⁻¹ = 1/λ in cm), so it scales linearly with energy. This is why IR spectra are plotted in cm⁻¹ rather than wavelength: equal spacing on the axis corresponds to equal energy spacing.

The transmittance trap: On a %T plot, an absorption band appears as a downward valley. A strong band might drop to 20–30% T. If you mentally flip the axis, you will assign bands to the wrong wavenumbers. Absorbance spectra avoid this because peaks point up, but many instrument outputs default to %T.

The Two Main Regions

Functional Group Region (4000–1500 cm⁻¹)

This region contains stretching vibrations of light atoms and is where you identify which functional groups are present. Bands here are relatively few and well separated.

Fingerprint Region (1500–400 cm⁻¹)

This region contains coupled vibrations — stretches, bends, and skeletal modes mixed together. Individual bands are hard to assign, but the overall pattern is highly characteristic of a specific compound. Two spectra matching across the fingerprint region is strong evidence of identical structure.

Common Absorption Bands and How to Assign Them

Group Approx. range (cm⁻¹) Appearance Notes
O–H (alcohol) 3200–3600, broad Broad, rounded Broadness from hydrogen bonding
O–H (carboxylic acid) 2500–3300, very broad Very broad, overlaps C–H Often the widest band in the spectrum
N–H 3300–3500 Medium, sharper than O–H Primary amines show two bands; secondary amines one
C–H (alkane) 2850–3000 Sharp, medium-strong Usually the clearest reference bands
C–H (alkene/aromatic) 3000–3100 Sharp, weak-medium Above 3000 cm⁻¹ distinguishes from alkane C–H
C≡C / C≡N 2100–2260 Weak to medium Often weak; may be absent if symmetric
C=O (aldehyde/ketone) 1700–1750 Strong, sharp One of the most reliable bands
C=O (ester) 1735–1750 Strong Higher than ketone due to resonance
C=O (amide) 1630–1690 Strong Lower than ketone
C=C (alkene) 1600–1680 Variable Weak if symmetric
Aromatic C=C 1450–1600 Medium, often multiple Ring breathing modes

Reading strategy: Start at the high-wavenumber end. Look for broad O–H/N–H around 3300, then C–H just below 3000. Move to the 1700 region for C=O. Then scan the fingerprint region for the overall pattern rather than assigning every band.

IR vs. Raman: Why the Same Sample Gives Different Spectra

Both are vibrational spectroscopies, but they obey different selection rules:

  • IR absorbs light when a vibration changes the dipole moment. Polar bonds (C=O, O–H, N–H) give strong IR bands.
  • Raman scatters light when a vibration changes the polarizability. Symmetric, non-polar bonds (C–C, S–S, aromatic rings) give strong Raman bands.

This is why water is a strong IR absorber but a weak Raman scatterer, and why C=O is strong in IR but often weak in Raman. The two techniques are complementary: a band weak or absent in one may be strong in the other. If you only have an IR spectrum, do not expect to see symmetric skeletal vibrations clearly.

Common Misreadings to Avoid

  1. Inverted transmittance peaks. Always check the y-axis label. A downward dip in %T is an absorption, not an emission.
  2. Atmospheric CO₂ and water. CO₂ gives a doublet near 2349 cm⁻¹; water vapor gives sharp lines around 1300–1900 and 3500–4000 cm⁻¹. These are instrument artifacts, not sample bands. If they appear in a background-corrected spectrum, the background was likely collected under different humidity.
  3. Baseline drift. A sloping baseline can make a weak band look like a shoulder or hide one entirely. Baseline correction before interpretation is standard practice.
  4. Over-interpreting the fingerprint region. Do not assign every small band to a specific group. Use the pattern for identification, not individual peak-by-peak assignment.
  5. Ignoring sample preparation effects. KBr pellet spectra can show broad O–H from absorbed moisture; ATR spectra may shift band positions slightly and distort intensities at low wavenumbers.

Practical Checkpoints Before You Interpret

  • Confirm the y-axis is transmittance or absorbance, and note which direction means absorption.
  • Check for atmospheric CO₂ (≈2349 cm⁻¹) and water bands; if present, consider re-collecting the background.
  • Apply baseline correction if the baseline is not flat.
  • Identify the strongest, most unambiguous bands first (C=O, C–H, O–H).
  • Compare the full fingerprint pattern against a reference spectrum rather than assigning bands in isolation.

Platforms such as RDRS SpectraLib provide visualization and database matching for IR, Raman, and XRD spectra, which can help confirm assignments against reference data. For a raw spectrum, the workflow is: load the data, correct the baseline, inspect the functional group region, then match the fingerprint pattern against a library.