FTIR PEAK INTERPRETATION

3400 cm-1 in FTIR: what a band there actually means

3400 cm-1 sits in the stretch region of the X-H bonds. An alcohol, a protein, a slightly damp powder and a carboxylic acid all absorb here, and in most of them the band is broad rather than sharp. What separates the candidates is not the number itself but the width of the band, where it sits inside the 3200-3600 cm-1 window, and the bands that travel with it.

15 published spectra quoted below

Direct answer

A band near 3400 cm-1 is an X-H stretching band, and in most samples it is an O-H stretch. Alcohol and phenol O-H groups that are hydrogen bonded absorb as a broad band across roughly 3200-3600 cm-1, while an O-H group that is not hydrogen bonded shows a much narrower band at the top of that window, near 3500-3650 cm-1. N-H stretches from amines and amides occupy the same region and are usually a little lower, so a 3400 cm-1 band cannot be assigned on its own. The bands that travel with it decide the case: an alcohol brings its C-H stretches at 2850-2960 cm-1, a protein brings the amide II band near 1545 cm-1, water brings its bending band at 1630-1650 cm-1, and a carboxylic acid brings a C=O band near 1730 cm-1 together with an O-H so strongly hydrogen bonded that it runs down past 3000 cm-1. That tail below 3000 cm-1 is the one clean way to separate an acid from an alcohol.

What decides the assignment

  1. How broad the band is

    A hydrogen-bonded O-H is broad, and an O-H that is free of hydrogen bonds is narrow. Wood samples with many hydroxyl and phenolic groups show strong absorption between 3200 and 3500 cm-1 with a narrow peak at 3500 cm-1, while the O-H stretch of an alcohol is described as a strong and broad peak. Width therefore separates free from hydrogen-bonded, not O-H from N-H: the N-H stretch is broad as well.

  2. Where it sits inside the 3200-3600 window

    The published ranges put alcohol O-H at 3650-3200 cm-1 and carboxylic acid O-H at 3300-2500 cm-1, so an acid starts lower and keeps going. In heat-treated wood the band at 3430 cm-1 broadened towards lower wavenumbers as the amount of carboxylic acid rose, which is how an acid reads in practice: the band does not stop, it runs down past 3200 and past 3000 cm-1.

  3. Which bands travel with it

    Water adds a bending band at 1630-1650 cm-1: in one oxide the bands around 3400 and 1630 cm-1 accounted for absorbed molecular water, and in a biopolymer the peak centred at 1645 cm-1 was assigned to the H-O-H bending contribution. An alcohol adds C-H stretches at 2850-2960 cm-1. A protein adds the amide II band near 1545 cm-1. Look for these before naming the 3400 cm-1 band.

  4. Whether the sample was dry

    Water adsorbed from the atmosphere produces a broad band centred at 3400 cm-1 on its own, and adsorbed water is described as part of the hydroxyl signal on a solid surface, so a hygroscopic sample can carry a 3400 cm-1 band that has nothing to do with the material. Drying and re-measuring settles it: a band that shrinks was moisture, a band that stays was part of the sample.

Main assignments

Functional group What it means here Cited sources
Hydroxyl (O-H) The largest group at this wavenumber. Alcohol, phenol, polyol and carbohydrate O-H groups all stretch here, broad and hydrogen bonded unless the group happens to be free of hydrogen bonding. 285
Water (H2O) Moisture and water itself. Water gives a broad band centred at 3400 cm-1 and a bending band at 1630-1650 cm-1 with it; in soil the broad OH/NH band at 3400 cm-1 sits next to a free O-H band at 3600 cm-1. 57
N-H (amine and amide) N-H stretches occupy the same window and sit a little lower: a primary amine was found at 3253 cm-1 and another at 3295 cm-1. In a protein the N-H stretch is the amide A band around 3300 cm-1. 41
Amide The N-H part of a peptide. In gelatin the amide A stretch around 3300 cm-1 comes with the amide I band at 1653 cm-1 and the amide II band at 1546 cm-1. 15
Carboxyl (COOH) The acid O-H is the extreme case: so strongly hydrogen bonded that it runs from 3300 down to 2500 cm-1, and it comes with the acid C=O band. In a reaction mixture the carboxylic acid peak at 1730 cm-1 is the one that moves when the product forms. 9
Silanol (Si-OH) On silica and in surface-modified minerals the O-H of silanol groups falls in the same window as the O-H of trapped water. A silane-modified clay showed a broad band between 3450-3290 cm-1 assigned to OH and NH moieties, with the OH coming from the silicate sheet surfaces, trapped water or free silanol groups. 7
Hydrogen bond The reason the band is broad in the first place. Hydrogen bonding spreads an O-H stretch over hundreds of wavenumbers, so a hydrogen-bonded sample gives a wide 3200-3600 cm-1 band and a dry, non-bonded one gives a narrow band near 3500-3650 cm-1. 26

Check the bands around it

  • The H-O-H bending band of water. Where it appears, the 3400 cm-1 band is at least partly moisture or water.

  • Amide II, the band that has to accompany an amide A (N-H) stretch in a protein.

  • The aliphatic C-H stretch of the chain that carries the O-H group.

  • The C=O band of a carboxylic acid, the partner of the very broad acid O-H that runs down below 3000 cm-1.

  • The O-H band of a group that is not hydrogen bonded, for example on a dry mineral surface.

Possible materials

Material Supporting peaks Overlapping groups Cited sources
PET 3400, 1408, 800 Alkyl C-H, N h 1
Polyethylene 3400, 1720, 1080 Alkyl C-H, Hydroxyl (O-H) 1
SnO2 3400, 2800, 1109 Hydroxyl (O-H), Alkyl C-H, Water (H2O) 1

Materials are shown only when the same literature pool supports this band and at least one additional characteristic peak.

Published spectra

Expert reading

The 3400 cm-1 band is where assignments are most easily over-committed, because almost every organic sample and every slightly damp powder has one. Three habits keep it honest. First, read the band as a region rather than a number: the O-H stretch of alcohols and related groups is reported at 3300, 3400, 3430, 3435 and 3500 cm-1 in the literature quoted here, so a measured 3400 cm-1 is one of several that mean the same thing. Second, always look for the partner band: a 3400 cm-1 band with no water bending at 1630-1650 cm-1, no amide II at 1545 cm-1 and no C-H at 2850-2960 cm-1 is not enough to name a group. Third, let the drying test decide between material and moisture, since adsorbed water alone produces a broad band centred at 3400 cm-1. The bands listed above link to the peak pages for the neighboring wavenumbers, where the other assignments reported at the same position are listed.

Have a spectrum with this band?

Upload your FTIR spectrum and get a full interpretation report — peak assignments with literature citations, library matches, and a literature-backed analysis — in seconds.

Submit Requirement Form
AD SCI paper writing and submission guidance; Custom algorithms for research. Email to enquire