FTIR sample preparation for ATR, KBr pellets, films, liquids, and powders

The practical role of sample preparation in FTIR

In FTIR, sample preparation is the step that makes a real material measurable without changing the chemistry under investigation. The best route depends on the physical form of the sample, the purpose of the measurement, and whether the method needs to emphasize the surface or the bulk material. ATR is often the quickest option for many solids and liquids because it needs little preparation. Transmission methods, including KBr pellets, thin films, liquid cells, and gas cells, give better control over path length and bulk response. Powders may also be measured by diffuse reflectance or as mulls when pellets are not suitable. For related laboratory workflow topics, see the sample preparation section.

Preparation is not a minor handling detail. It can affect peak intensity, baseline shape, scattering, water interference, reproducibility, and whether a spectrum can be compared fairly with a library record or a previous batch.

Start with the sample form and the analytical question

A good FTIR preparation choice starts with two questions. First, what form is the sample in: hard solid, powder, film, coating, liquid, gel, fiber, gas, or residue? Second, what information is needed: identity confirmation, comparison with a reference spectrum, surface contamination check, bulk composition trend, or quantitative measurement?

These questions matter because FTIR sampling modes do not probe the same volume of material. ATR measures a shallow region close to the crystal-sample interface, so it is commonly treated as a surface-sensitive method. Transmission measures radiation that passes through a prepared thickness of sample, making it more directly affected by the total absorbing path. Diffuse reflectance collects light scattered from a powder bed and can emphasize weak bands differently from transmission unless the spectrum is transformed mathematically and the sample is prepared consistently.

For routine identification, speed and repeatability often matter more than capturing every weak absorption band. For method-driven quality control, however, the preparation route may be fixed by a validated procedure or standard method. ASTM D7948-20 for respirable crystalline silica, for example, shows how a regulated IR method can specify different routes, such as KBr disc analysis, indirect redeposition, or direct filter analysis, depending on the measurement design. Those routes are not universal for every FTIR sample, but the example shows why laboratories should not treat preparation methods as interchangeable after validation.

Common FTIR sample preparation routes

ATR for fast measurement of solids and liquids

Attenuated total reflectance, or ATR, is widely used because many samples can be placed directly against a diamond, zinc selenide, or germanium crystal. Manufacturer guidance commonly describes ATR as requiring minimal or no sample preparation for many solid and liquid samples. This makes it useful for plastics, coatings, rubbers, pastes, oils, tablets, fibers, and unknown residues.

The main limitation is contact. A hard, rough, curved, or granular sample may leave microscopic air gaps, reducing signal and changing relative peak intensities. Contact pressure should be consistent, especially when spectra are compared across batches. ATR spectra also have different relative intensities from transmission spectra because penetration depth changes with wavelength, crystal type, sample refractive index, and angle of incidence. Many software packages offer ATR correction, but correction cannot compensate for poor crystal contact.

KBr pellets for powdered solids and bulk transmission

KBr pellet preparation is a classic transmission approach for powdered solids. The sample is finely ground, diluted in dry infrared-grade potassium bromide, and pressed into a transparent or translucent pellet. Potassium bromide is used because it is largely transparent in the mid-infrared region commonly used for organic and many inorganic FTIR measurements.

Published laboratory and manufacturer procedures often use low sample loading, commonly around 0.1 to 1 percent sample in KBr, because pellets are thicker than liquid films and can easily become over-absorbing. Too much sample, large particles, or poor mixing can produce noisy spectra, flattened peaks, and baseline distortion from scattering. KBr is also hygroscopic, so absorbed moisture can add broad water-related bands and raise the background. Many protocols therefore call for dry KBr, rapid handling, desiccator storage, and a blank KBr pellet or appropriate background when pellet-related scattering or moisture must be corrected.

KBr pellets are not appropriate for every material. Samples that react with the halide matrix, contain moisture, are unstable under pressure, or have critical bands overlapping with water interference may require another approach. Some hydrochloride or chloride-containing materials may need special attention because halide exchange has been reported as a possible source of spectral change in certain pellet preparations.

Thin films, solution casting, and liquid cells

Films are useful when a polymer, coating, adhesive, or soluble solid can be made thin enough for transmission. The objective is to create a uniform path length so that strong and weak bands remain within the detector’s useful response range. If the sample is too thick, intense bands may absorb nearly all the IR light. If it is too thin, weak diagnostic bands may disappear into the noise.

Solution casting can work when the sample dissolves in a solvent that evaporates cleanly and does not react with the material. The solvent should be checked because residual solvent bands can be mistaken for sample peaks. KBr and other salt windows require careful selection because aqueous samples can fog or damage some window materials. For liquids, a demountable or sealed liquid cell provides better path-length control than squeezing a drop between plates, especially when comparing samples quantitatively or monitoring concentration changes.

Mulls and diffuse reflectance for difficult powders

A mineral oil mull is prepared by grinding a solid and dispersing it in a small amount of oil between infrared-transparent plates. This can be useful when a sample is not suitable for KBr pellet preparation. The trade-off is that mineral oil has its own C-H absorption bands, especially around the 3000 to 2800 cm^-1 region and related deformation regions. If those bands are important to the analysis, a mull may obscure the answer.

Diffuse reflectance, often called DRIFTS in FTIR workflows, is another route for powders and highly scattering materials. The sample may be measured as a powder bed or diluted in KBr or another infrared-transparent matrix. Preparation is usually easier than pressing pellets, but particle size, packing density, dilution ratio, and surface reflection can strongly affect the result. For quantitative work or cross-method comparison, diffuse reflectance spectra often need controlled preparation and appropriate mathematical treatment rather than visual comparison alone.

Preparation variables that change the spectrum

Several practical variables explain why two spectra from the same material can look different. The first is particle size. Large particles scatter infrared light, causing sloped baselines and distorted bands. Grinding improves homogeneity, but aggressive grinding can heat, oxidize, hydrate, or physically change some samples. The goal is not maximum grinding; it is controlled, reproducible particle reduction without altering the material. See also: buying guides.

The second variable is water. Moisture can come from the sample, the air, salt plates, KBr powder, or an insufficiently purged instrument. Because KBr is hygroscopic, poor storage can create background features that interfere with interpretation. Water can also damage or cloud salt windows. If water is part of the sample, a water-compatible window or ATR method may be more appropriate than forcing the sample into a KBr workflow.

The third variable is thickness or concentration. In transmission measurements, path length and concentration control absorbance. Overloaded samples create saturated bands that cannot be interpreted reliably. Underloaded samples may miss weak functional-group bands. For a useful routine spectrum, the strongest diagnostic peaks should be intense but not completely saturated.

The fourth variable is surface contact. In ATR, the measured signal depends on intimate contact with the crystal. Soft samples usually contact well; hard powders, textured plastics, and curved parts often do not. Cutting a flat face, applying consistent pressure, or using a different crystal can improve repeatability.

The fifth variable is contamination. Fingerprints, polishing residues, previous samples, solvents, tape, cardboard fibers, and cleaning wipes can all produce infrared bands. Preparation tools, crystal surfaces, pellet dies, mortars, liquid cells, and windows should be cleaned in a way that is compatible with the material and documented in the laboratory procedure.

A sample-type decision guide

Sample type	Common preparation choice	Why it is used	Main caution
Flat polymer, rubber, coating, or residue	ATR	Fast measurement with little preparation	Surface contact and ATR intensity differences
Fine organic powder	KBr pellet, ATR, or diffuse reflectance	Multiple routes depending on surface or bulk objective	Particle size, moisture, and dilution consistency
Hard or irregular solid	ATR after creating a flat contact area, or micro-sampling	Minimizes destructive preparation	Poor contact can weaken the spectrum
Thin polymer film	Transmission film or ATR	Transmission can show bulk film response	Film thickness and interference fringes
Soluble solid	Solution cast film or liquid cell	Can create a uniform thin layer	Residual solvent and window compatibility
Non-aqueous liquid	ATR or liquid cell	ATR is quick; liquid cell controls path length	Volatility, viscosity, and cleaning
Aqueous liquid	ATR or water-compatible liquid cell	Avoids damage to unsuitable salt windows	Strong water bands may mask analyte peaks
Gas or vapor	Gas cell	Provides defined optical path length	Condensation, pressure, and resolution needs

Quality control before collecting spectra

Before collecting a final spectrum, confirm that the preparation route matches the intended comparison. A library spectrum collected by ATR may not match a transmission spectrum perfectly, even when the chemistry is the same. A KBr pellet spectrum may not match a diffuse reflectance spectrum without appropriate correction. If the method is used for release testing, contamination investigation, or long-term trend analysis, the sampling accessory, pressure, dilution, particle size target, background method, and cleaning steps should be consistent.

A practical pre-run checklist helps reduce avoidable errors:

• Confirm that the sampling mode matches the reference spectrum or validated method.
• Check whether the sample is moisture-sensitive, pressure-sensitive, volatile, reactive, or heterogeneous.
• Use clean tools, clean windows or crystals, and compatible solvents.
• For pellets and mulls, grind enough to reduce scattering but avoid changing the sample.
• For ATR, ensure full contact and document pressure settings where possible.
• For transmission, avoid complete absorption in major peaks by adjusting thickness or dilution.
• Run a suitable background and repeat the measurement if baseline shape or peak intensity looks abnormal.

Good FTIR preparation is a balance between speed, sample preservation, and spectral quality. The fastest method is not always the most informative, and the most elaborate method is not always necessary. The best choice preserves the sample chemistry, produces a spectrum within the instrument’s reliable response range, and answers the analytical question with the least avoidable distortion.

Frequently asked questions

Is ATR always better than KBr pellet preparation?

No. ATR is faster and usually easier, but it is more surface-sensitive and depends on crystal contact. KBr pellets can be useful when a bulk transmission spectrum of a powdered solid is needed, when weak bulk bands are important, or when the comparison method was built around transmission data.

How much sample is used in a KBr pellet?

Many laboratory procedures use low loading, often around 0.1 to 1 percent sample in dry KBr. The exact ratio should follow the laboratory method, material behavior, and target absorbance range. If peaks are saturated or the baseline is noisy, reduce sample loading, improve grinding, or prepare a new pellet.

Why does the same material look different by ATR and transmission FTIR?

ATR and transmission interact with different effective sample depths. ATR intensity depends on penetration depth and contact with the crystal, while transmission depends on the prepared path length. Relative peak intensities can therefore differ even when peak positions confirm the same chemistry.

Can water-containing samples be prepared on KBr plates?

Usually this is not recommended because KBr and several salt windows are moisture-sensitive and can fog or dissolve. ATR or a water-compatible cell material is often a better option, although water absorption bands may still overlap with analyte bands.

What is the most common cause of poor FTIR spectra from powders?

Common causes include large particle size, uneven mixing, too much sample, moisture in KBr, poor packing in diffuse reflectance cups, or weak ATR contact. Repreparing the sample with controlled particle size, lower concentration, dry materials, and a clean background often improves the result.