FTIR
Absorption bands fingerprint the functional groups.
Every bond type absorbs infrared light at its own characteristic wavenumber, so an FTIR spectrum is a functional-group fingerprint: which bonds are present, roughly how much, and whether they changed since the last sample.
What it measures
Infrared light drives molecular vibrations; a bond absorbs where the photon energy matches its vibrational mode. The spectrum is read by region:
- 3200-3600 cm⁻¹: O-H and N-H stretches (broad when hydrogen-bonded: water, alcohols).
- 2850-3000 cm⁻¹: C-H stretches of alkyl backbones.
- 1650-1800 cm⁻¹: C=O stretches: carbonates, esters, ketones, the workhorse region for electrolytes and binders.
- 1000-1300 cm⁻¹: C-O stretches; below ~1500 cm⁻¹ lies the fingerprint region, unique to the whole molecule.
Quantification follows Beer-Lambert, A = ε·l·c: absorbance (converted from transmittance as A = −log₁₀T) scales linearly with concentration, so a calibrated band height or area becomes a concentration measurement.
How to read the output
Correct the baseline before reading anything, drift and scattering tilt the whole spectrum, and ATR measurements need the wavelength-dependent penetration correction before band intensities compare honestly. Then read presence, position, and shape: a band where none should be is contamination or degradation; a band shifted tens of wavenumbers reports a changed chemical environment; broadening reports disorder or hydrogen bonding. For comparisons, overlay spectra with a vertical offset and look for what appeared, vanished, or moved, the differences carry the answer, not the absolute spectra.
A real use case
Incoming electrolyte QC on a cell line: each drum is supposed to be the same EC/DMC-based formulation. FTIR takes minutes per sample, the carbonyl region around 1750-1800 cm⁻¹ resolves the cyclic carbonate from the linear one, and Beer-Lambert-calibrated band areas give the solvent ratio. One delivery shows the C=O envelope shifted toward the linear carbonate plus an unexpected O-H band: a mislabeled blend with moisture pickup, caught at goods-in for the cost of a syringe of liquid, instead of three weeks later as a coulombic-efficiency excursion in formation.
Common mistakes
- Reading raw transmittance for quantification. Beer-Lambert is linear in absorbance, not transmittance, convert first.
- Skipping baseline and ATR corrections, then comparing band intensities across samples measured on different days or accessories.
- Assigning a single band without checking the rest of the group’s pattern, most functional groups announce themselves at more than one wavenumber.
- Ignoring atmospheric CO₂ (~2350 cm⁻¹) and water-vapor bands, which appear and vanish with purge quality and masquerade as sample changes.
- Quantifying outside the calibrated range, Beer-Lambert bends at high absorbance, and extrapolated calibrations overestimate.
Corrected, assigned, quantified, compared
Niobia ingests FTIR exports, applies baseline correction and ATR correction guidance, and converts transmittance to absorbance where quantification needs it. It identifies functional groups by spectral region, O-H, N-H, C-H, C=O, C=C, C-O, and the fingerprint region, logs peak positions and widths (FWHM), and runs Beer-Lambert quantification against your calibration. Multi-sample comparisons come back as offset overlay plots with the band-level differences called out, which is the format the electrolyte-QC decision above actually needs.
Frequently asked
When is FTIR the right tool versus Raman?
They are complements: polar bonds (C=O, O-H) absorb strongly in the infrared, while symmetric and carbon-backbone modes scatter strongly in Raman. Electrolytes, binders, and degradation products with carbonyl chemistry usually read better in FTIR; carbon materials read better in Raman.
Can FTIR be quantitative, or only identify groups?
Quantitative within a calibration: Beer-Lambert ties absorbance linearly to concentration, so a calibrated band quantifies solvent ratios, additive levels, or moisture. The honesty conditions are a corrected baseline, absorbance units, and staying inside the calibrated range.
What does a shifted or broadened band mean physically?
Position shifts report a changed bonding environment, coordination, hydrogen bonding, salt association. Broadening reports a distribution of environments: disorder, mixtures, or hydrogen-bond networks. Both are real signals, not instrument noise, once baseline and purge artifacts are excluded.