How to Read an FTIR Spectrum: A Practical Guide

Fourier-Transform Infrared (FTIR) spectroscopy is one of the most widely used analytical techniques in chemistry and materials science. By measuring how a sample absorbs infrared light at different frequencies, FTIR produces a spectrum that reveals the molecular structure of the sample — specifically, which functional groups are present and how they are bonded together.

Interpreting an FTIR spectrum is a fundamental skill for organic chemists, polymer scientists, forensic analysts, and quality-control laboratories. Whether you are confirming the identity of a synthesized compound, checking raw materials for contamination, or characterizing an unknown sample, reading the spectrum correctly is the critical first step.

This guide walks you through a systematic approach to FTIR spectrum interpretation — from understanding what the axes represent, to identifying functional groups region by region, to avoiding common mistakes. If you want to automate part of this process, SpectralBench's FTIR Peak Identifier can detect peaks and assign functional groups with confidence scores.

What Does an FTIR Spectrum Show?

An FTIR spectrum is a plot of infrared light absorption as a function of frequency. The x-axis displays wavenumber in reciprocal centimeters (cm⁻¹), conventionally running from 4000 cm⁻¹ on the left to 400 cm⁻¹ on the right. Wavenumber is directly proportional to energy: higher wavenumbers correspond to higher-energy vibrations.

The y-axis shows either transmittance (%) or absorbance. In a transmittance spectrum, absorption bands appear as downward dips — the deeper the dip, the stronger the absorption. In an absorbance spectrum, bands point upward as peaks. Both representations carry the same information; transmittance is more traditional, while absorbance is linear with concentration and therefore preferred for quantitative work.

Each absorption band corresponds to a specific molecular vibration — stretching or bending of chemical bonds. The position (wavenumber), intensity (strong, medium, or weak), and shape (broad or sharp) of each band encode information about the functional groups in the molecule. In this sense, an FTIR spectrum is a molecular fingerprint: no two different molecules produce exactly the same spectrum.

The Four Key Spectral Regions

The mid-infrared range (4000–400 cm⁻¹) is conventionally divided into four regions. Working through each region systematically is the most reliable way to extract structural information from a spectrum.

1. X-H Stretching Region (3600–2500 cm⁻¹)

This is the first region to examine because it immediately tells you whether certain key hydrogen-bearing groups are present. O-H stretches appear as a broad absorption between 3200 and 3600 cm⁻¹ — the breadth results from extensive hydrogen bonding, and a broad band centered around 3300 cm⁻¹ strongly suggests an alcohol, phenol, or carboxylic acid. N-H stretches appear as medium-intensity bands between 3300 and 3500 cm⁻¹; primary amines produce a characteristic doublet (two peaks), while secondary amines show a single band. C-H stretches are sharp absorptions between 2800 and 3100 cm⁻¹. The exact position distinguishes sp³ C-H (below 3000 cm⁻¹, as in alkanes) from sp² C-H (above 3000 cm⁻¹, as in alkenes and aromatics).

2. Triple Bond Region (2500–2000 cm⁻¹)

This region is often quiet, which makes any absorption that does appear highly diagnostic. Nitriles (C≡N) produce a strong, sharp peak near 2250 cm⁻¹. Terminal alkynes (C≡C) absorb near 2100 cm⁻¹ as a weaker band — and internal alkynes may show no absorption at all if the molecule is symmetric. Cumulated double bonds such as isocyanates (N=C=O) and azides (N₃) also absorb in this window. If this region is blank, you can immediately rule out nitriles, alkynes, and related groups.

3. Double Bond Region (2000–1500 cm⁻¹)

The carbonyl stretch (C=O) is arguably the single most diagnostic band in infrared spectroscopy. It appears as a strong, sharp absorption between 1630 and 1850 cm⁻¹, and its exact position tells you what type of carbonyl you are dealing with: ketones near 1715 cm⁻¹, aldehydes near 1730 cm⁻¹, esters near 1740 cm⁻¹, carboxylic acids near 1710 cm⁻¹, amides near 1650 cm⁻¹, and anhydrides showing two bands around 1800 and 1760 cm⁻¹. C=C stretches in alkenes and aromatics appear between 1600 and 1680 cm⁻¹, typically weaker than carbonyls.

4. Fingerprint Region (1500–400 cm⁻¹)

Below 1500 cm⁻¹, the spectrum becomes complex. This region contains C-O stretches (1000–1300 cm⁻¹), C-N stretches, S=O stretches, and various bending modes. While individual band assignments are harder to make here, the overall pattern is unique to each molecule — hence the name “fingerprint region.” This region is most useful for confirming identity by comparison with a reference spectrum rather than for de novo structural determination. SpectralBench's FTIR Peak Identifier database includes fingerprint-region assignments to help with matching.

Step-by-Step Interpretation Workflow

Rather than scanning the spectrum randomly, follow a systematic workflow each time. Consistency prevents you from overlooking important features.

1. Assess the overall spectrum shape. Is the spectrum rich in sharp C-H and carbonyl bands (suggesting an organic compound)? Does it show broad, featureless absorptions (possible inorganic or highly polymeric material)? This first impression narrows the field immediately.
2. Examine the 3600–2500 cm⁻¹ region. Look for O-H (broad), N-H (medium, possibly doubled for primary amines), and C-H (sharp). The presence or absence of these bands tells you about hydrogen bonding and the hybridization of carbon.
3. Check for a carbonyl (C=O). This is the strongest and most diagnostic band in most organic spectra. Its position narrows the compound class. If absent, you can rule out ketones, aldehydes, esters, carboxylic acids, amides, and anhydrides in one step.
4. Scan the triple-bond region (2500–2000 cm⁻¹). Any absorption here is significant because few groups absorb in this quiet zone. Look for nitriles, alkynes, azides, and isocyanates.
5. Examine the fingerprint region. Use reference spectra or SpectralBench's database for matching. Look for strong C-O bands in esters and alcohols, aromatic out-of-plane bending patterns, and S=O stretches in sulfonyl compounds.
6. Consider sample preparation. ATR spectra may show slight shifts in peak positions and relative intensities compared to transmission (KBr pellet) spectra. Thin-film spectra may exhibit interference fringes. Always note the technique used when comparing with reference data.
7. Cross-reference with other data. FTIR is powerful but rarely used in isolation. Combine your interpretation with NMR, mass spectrometry, or elemental analysis for definitive identification.

Common Pitfalls in FTIR Interpretation

Even experienced spectroscopists can be tripped up by these common mistakes. Being aware of them will save you from incorrect assignments.

• Confusing O-H and N-H stretches. Both produce broad absorptions in the 3200–3500 cm⁻¹ region. O-H bands tend to be broader and more intense. Primary amines produce a doublet, which is a distinguishing feature — but a mixture of an alcohol and a secondary amine could look similar at first glance.
• Ignoring band shape and intensity. Position alone is not enough to make an assignment. A strong, sharp band at 1715 cm⁻¹ is a ketone carbonyl. A weak shoulder near the same position could be an artifact, an overtone, or a different functional group entirely. Always consider the full profile of each band.
• Over-interpreting the fingerprint region. Assigning every band below 1500 cm⁻¹ to a specific functional group is unrealistic. Many absorptions in this region arise from coupled vibrations that are difficult to assign without computational support. Focus on the strongest, most isolated bands.
• Forgetting that some groups have multiple bands. Primary amines show two N-H stretches (symmetric and asymmetric). Anhydrides show two C=O stretches. Esters have both a C=O and a strong C-O band. A single band is rarely enough to confirm a functional group — look for the full pattern.
• Not accounting for hydrogen bonding. Hydrogen bonding shifts O-H and N-H stretches to lower wavenumbers and broadens them dramatically. A free O-H absorbs as a sharp peak near 3650 cm⁻¹, while a hydrogen-bonded O-H can appear as a broad hump stretching from 3600 down to 2500 cm⁻¹. If your spectrum was measured in the solid state or as a neat liquid, hydrogen bonding effects will be pronounced.

Band Intensity and Shape

Beyond peak position, intensity and shape carry important structural information that can make or break an assignment.

Intensity reflects the magnitude of the dipole moment change during the vibration. Polar bonds like C=O and C-F produce strong absorptions. Nonpolar or symmetric bonds like C≡C in internal alkynes may produce weak or even absent bands. A strong carbonyl band is far more reliable as a diagnostic marker than a weak C=C stretch.

Broad vs. sharp bands indicate whether hydrogen bonding is at play. O-H and N-H groups involved in hydrogen bonding produce broad absorptions spanning hundreds of wavenumbers. In contrast, C-H stretches and carbonyls are almost always sharp and well-defined. If you see a broad band in the 3000–3500 cm⁻¹ region, think hydrogen bonding immediately.

Doublets— two closely spaced bands — are diagnostic for specific functional groups. Primary amines show a doublet in the N-H stretch region (symmetric and asymmetric vibrations). Anhydrides show a doublet in the C=O region (symmetric and asymmetric stretches separated by about 60 cm⁻¹). The aldehyde C-H Fermi resonance doublet near 2720 and 2850 cm⁻¹ is one of the most reliable aldehyde markers.

Shoulders indicate overlapping bands from similar functional groups or conformational isomers. A carbonyl band with a shoulder may indicate two different carbonyl environments in the same molecule, or a mixture of compounds. Deconvolution techniques or second-derivative spectra can help resolve overlapping bands.

FTIR Interpretation Examples

Let's walk through three real-world spectra to see the systematic approach in action. For each example, we follow the same top-down workflow: X-H region first, then carbonyl, then triple bonds, then fingerprint.

Example 1: Ethanol (CH₃CH₂OH)

Starting at the high-wavenumber end, ethanol shows a broad, strong O-H stretch centered around 3300 cm⁻¹ — the hallmark of a hydrogen-bonded hydroxyl group. Moving down, sharp C-H stretches appear between 2900 and 3000 cm⁻¹ (both CH₃ and CH₂ groups). There is no carbonyl band anywhere near 1700 cm⁻¹, immediately ruling out aldehydes, ketones, esters, and carboxylic acids. The fingerprint region shows a strong C-O stretch near 1050 cm⁻¹, consistent with a primary alcohol. The combination of broad O-H, C-H stretches, no carbonyl, and a C-O stretch near 1050 cm⁻¹ points unambiguously to a primary alcohol.

Example 2: Acetone (CH₃COCH₃)

Acetone's spectrum is dominated by a strong, sharp carbonyl stretch at 1715 cm⁻¹ — right in the ketone range. The X-H region shows only sharp C-H stretches around 2960 and 2925 cm⁻¹ from the methyl groups; notably, there is no broad O-H or N-H absorption, which rules out alcohols, carboxylic acids, and amines. The triple-bond region is empty. In the fingerprint region, CH₃ bending modes appear near 1360 cm⁻¹ (symmetric umbrella) and 1430 cm⁻¹ (asymmetric deformation). The C-C stretches contribute to bands around 1200 cm⁻¹. The spectrum is simple and clean, matching a symmetric dialkyl ketone.

Example 3: Nylon 6,6 (a polyamide)

Nylon shows a moderately broad N-H stretch near 3300 cm⁻¹ — not as broad as the O-H in ethanol, but noticeably wider than a sharp C-H band. Below that, C-H stretches from the methylene chains appear between 2850 and 2930 cm⁻¹. The carbonyl region reveals the amide I band near 1640 cm⁻¹ (C=O stretch shifted to lower frequency by resonance with nitrogen) and the amide II band near 1540 cm⁻¹ (N-H bend coupled with C-N stretch). These two bands together — amide I and amide II — are the signature of an amide bond. The fingerprint region shows strong methylene bending near 1460 cm⁻¹ and the amide III band (C-N stretch mixed with N-H bend) near 1260 cm⁻¹. The spectrum is consistent with a secondary amide (one N-H per repeat unit), confirming a polyamide like nylon.

Tools for FTIR Analysis

SpectralBench provides free, browser-based tools for FTIR analysis — no installation, no account, no data uploaded to external servers.

• FTIR Peak Identifier — automated peak detection and functional group assignment with confidence scores
• Spectral File Viewer — view JCAMP-DX, SPC, and OPUS files directly in the browser
• FTIR Functional Group Table — comprehensive reference table with 80+ functional group entries

Summary

Interpreting an FTIR spectrum is a systematic process: start with the overall shape, work through the four spectral regions from high to low wavenumber, and always consider band intensity and shape alongside position. With practice, you will recognize common functional group patterns at a glance.

For detailed wavenumber ranges and intensity data, consult the FTIR Functional Group Frequency Table. To try automated peak identification on your own spectra, use the FTIR Peak Identifier.