FTIR sample preparation, done properly

1 · How FTIR sampling actually works

An FTIR spectrometer measures how much infrared light your sample absorbs at each wavenumber. Sample preparation exists for one reason: to put the right amount of sample in the beam, in the right physical form, so the absorbance stays in a measurable range. Too much material and strong bands "bottom out" into flat-topped, unusable peaks; too little and the spectrum drowns in noise. As a working rule, aim to keep your strongest bands below roughly 1 absorbance unit.

Three families of technique achieve this:

• Transmission — the beam passes straight through a thin, dilute presentation of the sample: a KBr pellet, a liquid film between IR windows, a pressed polymer film, or a gas in a cell. Pathlength and dilution are your controls. The Beer–Lambert law (absorbance ∝ concentration × pathlength) makes transmission the natural choice for quantitative work.
• ATR (attenuated total reflectance) — the beam totally internally reflects inside a high-refractive-index crystal, and an evanescent wave probes the first 0.5–3 µm of whatever is pressed against it. Effectively a built-in micro-pathlength: no preparation, minimal cleaning, ideal for routine identification.
• Reflectance — diffuse reflectance (DRIFTS) for powders and catalysis, specular and grazing-angle reflectance for coatings and monolayers on reflective substrates.

2 · Choosing a technique for your sample

Start from the physical state of the sample and what you need from the answer. The short version:

For a personalised answer — including the exact accessory and consumables — use the interactive Accessory Selector.

3 · The KBr pellet method — full SOP

The classic transmission preparation for solids, and still the reference method for quantitative and library-quality spectra. Potassium bromide is transparent across the mid-IR (roughly 40,000–400 cm⁻¹), so a sample dispersed at low concentration in a pressed KBr disc behaves like a solid solution in an IR-invisible matrix.

1. Dry your KBr. KBr is strongly hygroscopic. Dry FTIR-grade KBr at ~110 °C and store it in a desiccator. Wet KBr betrays itself as broad bands near 3400 cm⁻¹ (O–H stretch) and 1640 cm⁻¹ (H–O–H bend) that can bury your sample's own O–H and N–H features.
2. Weigh at 0.2–1% w/w. Combine roughly 1–2 mg of sample with 200–300 mg of KBr for a 13 mm die (a ~100:1 dilution). Scale the total mass down with die area for smaller pellets.
3. Grind to fine flour. Grind in an agate mortar until homogeneous and free-flowing. Particle size must fall below the IR wavelength, or scattering produces a steeply sloping baseline. Do not over-grind — excess surface area pulls in moisture.
4. Load the die evenly. Transfer to a 13 mm evacuable die, level the powder, assemble, and pull vacuum to remove trapped air — the main cause of laminated, fragile pellets.
5. Press: 8–10 tonnes, 1–2 minutes. On a 13 mm die, 8–10 tonnes fuses the crystals into a clear disc; a 7 mm die needs only ~2 tonnes, which is why the hand-operated Mini-Pellet Press exists. Release slowly.
6. Eject, inspect, measure. A good pellet is transparent, like a small glass window, roughly 1–2 mm thick. Measure promptly against a fresh pure-KBr background pellet, and store discs in a desiccator.

KBr troubleshooting table

4 · ATR — and how to choose the crystal

In ATR the evanescent wave penetrates only microns into the sample, so the crystal's refractive index sets your effective pathlength — and its hardness and chemistry set what you can safely press against it. Penetration depth increases at lower wavenumbers, which is why ATR spectra show relatively stronger low-wavenumber bands than transmission spectra of the same material (most software offers an ATR correction).

Practical rules: press to the same clamp force every time (the calibrated anvil or digital force gauge exists for exactly this); ensure intimate optical contact for solids; and clean with lint-free wipes and an appropriate solvent between samples — carry-over is the most common ATR error.

5 · Liquids — cells, pathlengths and window materials

A liquid transmission spectrum is defined by two decisions: pathlength and window material.

Pathlength

• Neat organic liquids: 0.015–0.05 mm keeps the strongest bands on scale.
• Aqueous solutions: ≤0.025 mm — or move to ATR, because water absorbs ferociously across the mid-IR.
• Dilute solutions / weak absorbers: up to 1.0 mm, chosen by trial against the ~1 absorbance guideline.

Demountable cells (such as the Omni-Cell) set pathlength with PTFE spacers from 6 µm upwards; sealed cells lock in a fixed path for repeatable quantitative work; fixed-path flow cells (100 µm for oil analysis) make standards-based condition monitoring operator-independent.

Window materials — what survives what

Fogged KBr or NaCl windows are recoverable: a window polishing kit restores optically flat faces in minutes and is far cheaper than replacement pairs.

6 · Mulls — the moisture-proof alternative for solids

Grind 5–10 mg of solid to a fine paste with a drop of Nujol (mineral oil), then sandwich the mull thinly between two IR windows. Because there is no hygroscopic matrix, mulls suit moisture-sensitive samples and salts that would ion-exchange in KBr. The cost: Nujol contributes its own C–H bands near 2950–2850, ~1460 and ~1377 cm⁻¹. Where those regions matter, run a second mull in a fluorinated oil (Fluorolube) and read each region from the mull that is silent there.

7 · Films and polymers

Quantitative polymer FTIR wants a film of known, uniform thickness — which is precisely what a constant-thickness film maker with heated platens produces: 29 mm films at nominal 15, 25, 50, 100, 250 or 500 µm, pressed between platens at up to 300 °C (400 °C in the high-temperature kit). Mount films in card or magnetic holders to keep them flat in the beam.

Very thin, very uniform films sometimes show a sinusoidal ripple across the baseline — interference fringes from reflections at the two parallel faces. Tilting the film slightly, roughening one face, or switching to ATR removes them. (The fringe spacing also measures the film's thickness — a classic teaching exercise.)

8 · Gases

Concentration decides the cell. Percent-level components saturate long paths, so a 10 cm cell with KBr windows is the standard tool. Trace work goes the other way: multipass cells fold the beam through nominal 2.5, 5 or 10 m, and since absorbance scales linearly with pathlength, 10 m buys roughly a hundred-fold sensitivity gain over 10 cm. Heated cells (and heated transfer lines) keep condensable analytes in the vapour phase — essential for evolved-gas and stack-type measurements.

9 · Micro-samples and difficult materials

• Single particles, inclusions, contaminants: flatten the particle in a diamond compression cell and measure it through a 4× beam condenser — no IR microscope required.
• Microlitre liquids: multi-reflection micro-ATR measures from just 10 µL.
• Carbon-filled and black materials: use a germanium ATR crystal — its shallow ~1 µm penetration keeps strongly absorbing compounds on scale where diamond and ZnSe saturate.
• Monolayers and nanometre coatings on metal: grazing-angle reflectance with p-polarised light is the only technique with real sensitivity here.

10 · Lab housekeeping that quietly decides your data quality

• Backgrounds: collect a fresh background under the same purge, accessory and window conditions as the sample. A stale background is the most common source of "mystery" bands.
• CO₂ and water vapour: the sharp band at 2349 cm⁻¹ is atmospheric CO₂; the picket-fence fine structure through ~1400–1900 and ~3500–3900 cm⁻¹ is water vapour. Allow purge time after opening the compartment.
• Humidity — an Indian-lab reality: in coastal and monsoon conditions, KBr, NaCl and hygroscopic consumables degrade fast. Keep them oven-dried and desiccated, run the spectrometer's desiccant maintenance on schedule, and prefer diamond ATR or aqueous-safe windows during the wet months.
• Crystal and window care: lint-free wipes, appropriate solvents, and for ZnSe and Ge, cotton-based materials only — paper wipes slowly abrade soft optics.

11 · Frequently asked questions

What ratio of sample to KBr should I use?

Approximately 0.2–1% by weight — typically 1–2 mg of sample in 200–300 mg of dry KBr for a 13 mm pellet. If peaks flat-top, halve the loading.

How many tonnes do I press a KBr pellet at?

8–10 tonnes on a 13 mm die, held for one to two minutes; about 2 tonnes on a 7 mm die.

Why is my pellet cloudy?

Moisture in the KBr or insufficient load. Dry the KBr at ~110 °C, work quickly, and press within the 8–10 tonne window. Coarse grinding also scatters light — grind to flour.

Diamond, ZnSe or germanium ATR?

Diamond for routine and rough duty. ZnSe when budget matters and samples are soft, mild (pH 5–9) liquids. Germanium when samples are black or strongly absorbing, or when you want to see only the top ~1 µm.

Can I run water-based samples on KBr windows?

No — KBr and NaCl dissolve. Use CaF₂, BaF₂ or ZnSe windows, or ATR.

What is the sharp peak at 2349 cm⁻¹?

Atmospheric CO₂. Refresh your background and allow the purge to settle after opening the sample compartment.

Where can I ask a question that isn't covered here?

Send it to hello@spectrolabsystems.com or WhatsApp +91 88793 99488 — applications questions are welcome whether or not a purchase is attached.

12 · For educators and students

This hub is free to use in coursework and lab manuals with attribution to Spectrolab Systems. Beyond the text, we support teaching laboratories with:

• Syllabus-mapped starter kits — solid, liquid and gas sampling packs matched to Indian university practical courses, at academic pricing.
• The printable compatibility matrix — print it from the Selector page and pin it beside the spectrometer.
• Demonstrations and guidance — talk to us before finalising a departmental purchase; specifying correctly the first time is cheaper than upgrading later.
• Campus ambassador programme — final-year and postgraduate students can represent us on campus; see partner programmes.