ATR vs transmission FTIR: which sampling technique should you choose?
Every FTIR laboratory eventually settles into a default — usually ATR, because it is fast. The trouble starts when the default is applied to samples it doesn't suit. Here is the decision in full, so you choose deliberately rather than habitually.
What each technique actually measures
Transmission sends the beam straight through a dilute presentation of the sample — a KBr pellet, a thin liquid film between windows, a pressed polymer film, or a gas in a cell. You control pathlength and concentration, and absorbance follows the Beer–Lambert law. That linearity is why transmission remains the reference for quantitative work and library-grade spectra.
ATR reflects the beam inside a high-refractive-index crystal; an evanescent wave probes only the first 0.5–3 µm of whatever touches the crystal face. In effect the technique carries its own micro-pathlength. That is why a neat liquid or an intractable solid that would saturate any sensible transmission cell sits comfortably on an ATR — and why sample preparation largely disappears.
Choose ATR when…
- Speed and throughput dominate. Press, scan, wipe, next — routine identification and incoming-goods QC are ATR's home ground.
- The sample is awkward. Pastes, gels, rubbers, cured coatings, tablets, irregular solids — things you cannot dilute or press thin.
- Water is involved. Water's ferocious mid-IR absorbance suits ATR's micron-scale effective pathlength; transmission of aqueous samples demands paths of 0.025 mm or less with water-insoluble windows.
- You care about the surface. The evanescent wave sees the top microns only — a feature when analysing coatings or surface modification, and a germanium crystal narrows it to roughly a micron.
Choose transmission when…
- Quantitation matters. A fixed-pathlength cell or a weighed KBr pellet gives Beer–Lambert behaviour that ATR's contact- and wavelength-dependent pathlength cannot fully match.
- You need the bulk, not the skin. ATR of a laminated film reports the surface layer; transmission reports the whole thickness.
- You're building or matching against libraries. Classical reference libraries are transmission spectra; ATR spectra show relatively enhanced low-wavenumber bands and slight peak shifts unless software-corrected.
- The analyte is a gas. There is no ATR shortcut here — a 10 cm cell for percent-level components, multipass metres of pathlength for trace work.
- Bands sit at very low wavenumbers. ZnSe ATR cuts off near 650 cm⁻¹ and germanium near 600 cm⁻¹; a KBr pellet transmits to roughly 400 cm⁻¹, and diamond ATR reaches lower still.
What this means for your accessory list
Most working laboratories end up with both: a single-reflection diamond ATR (a Quest or Golden Gate class instrument) as the daily workhorse, and a transmission bench — press, 13 mm evacuable die, demountable liquid cell — for the quantitative, archival and gas work ATR cannot do. Neither replaces the other; they divide the workload.
Unsure which side your samples fall on? Run them through our Accessory Selector, or send the sample description to hello@spectrolabsystems.com — specifying correctly is the part of the job we enjoy most.