Identifying the polymer in a cured rubber by infrared spectroscopy

Which bands in an FTIR spectrum point to which elastomer, how ATR and pyrolysis sampling differ, and why filled, cured rubber spectra need care.

Intermediate3 min readArticleUpdated 2026-10-23NBRHNBREPDMFKMVMQNRSBR
Cover image: "Identifying the polymer in a cured rubber by infrared spectroscopy", with an analysis trace with peaks.

Infrared (IR) spectroscopy is the workhorse for polymer identification. Molecular bonds absorb infrared light at characteristic wavenumbers (cm⁻¹), producing a pattern that behaves like a fingerprint. For rubber, the difficulty is that the sample is cured, filled with black and often heavily extended.

Three ways to get a spectrum

  1. ATR (attenuated total reflectance). Press a clean cut surface onto a crystal. Quick and easy, but carbon black absorbs strongly and weakens the spectrum, and the surface may not represent the bulk.
  2. Transmission on a film. Swell or dissolve the extracted rubber and cast a thin film. Better for black-free or lightly filled rubber.
  3. Pyrolysate. Heat a small piece, collect the vapours and condense them on a salt plate or analyse them directly. This breaks the crosslinked network into fragments that are easy to identify and works well with carbon black. See Pyrolysis for crosslinked rubber.

Bands that point to a polymer family

These are approximate positions and the intensity pattern matters as much as any single band.

Polymer Characteristic features
NBR and HNBR nitrile C≡N stretch near 2237 cm⁻¹. In HNBR the double bond bands (near 970 cm⁻¹ for trans) are much weaker and a methylene rocking band near 720 cm⁻¹ grows.
EPDM / EPM strong CH₂ and CH₃ bands, methylene rocking near 720 cm⁻¹, no nitrile, no carbonyl
NR and IR double bond near 835 cm⁻¹, bands around 1660 cm⁻¹
SBR phenyl bands near 700 and 760 cm⁻¹ together with butadiene bands near 966 and 911 cm⁻¹
BR cis, trans and vinyl bands near 740, 966 and 911 cm⁻¹
CR (polychloroprene) C–Cl features in the lower region, and its double bond bands
ACM / AEM strong ester carbonyl near 1735 cm⁻¹
FKM and FEPM very strong, broad C–F absorptions roughly between 1000 and 1400 cm⁻¹
VMQ / silicones very strong Si–O–Si between about 1000 and 1100 cm⁻¹, Si–CH₃ near 1260 and 800 cm⁻¹
IIR / halobutyl gem-dimethyl doublet near 1365 and 1390 cm⁻¹

Why spectra of real parts can mislead

  • Fillers and plasticisers add bands. Esters in a plasticiser can mimic an acrylate. Run the extracted rubber, not the whole part.
  • Cure and ageing change the spectrum. Oxidised rubber shows carbonyl and hydroxyl bands that were not in the polymer.
  • Blends hide. Two polymers overlap. Ratios are estimated by comparing band intensities to a calibration curve made from known blends.
  • Fluoroelastomer grades look alike. FKM, FEPM and perfluoroelastomers all absorb strongly in the C–F region. IR alone gives the family. Fluorine content and monomer ratio need other evidence such as elemental analysis, NMR, DSC and the physical profile.

Good practice

Always back an IR identification with a second method. Glass transition by DSC and pyrolysis products are the usual partners. The combination usually removes any doubt.

Sources and further reading

  • ASTM D3677, rubber: identification by infrared spectrophotometry.
  • Band positions are approximate and shift with composition and sampling method. Confirm against reference spectra.
About this article. It is original writing prepared from public sources, standards and general industry practice. It is not a copy of any book or standard and it is not engineering advice. Confirm values against the current edition of the standard and test in your own application. Only analyse material you own or are entitled to analyse.

Have a part to decode?

Send a part you own or may analyse. We estimate what it is made of and profile how it behaves. You get a profile to match, not a copied recipe.