BET Surface Area and Porosity: Degassing and Reading Isotherms

A BET measurement determines the specific surface area of a solid, expressed in m² per gram, by physical adsorption of a gas. The sample is normally exposed to nitrogen at the temperature of liquid nitrogen (77 K), and the amount of gas held on the surface is recorded at increasing relative pressures (p/p0). From the gas quantity needed to cover the surface with a single layer, and the area one gas molecule occupies, the total surface area follows. For solids with very low surface area, nitrogen may give too small a signal to be reliable, and krypton is then used instead. The general framework of the method is described in ISO 9277.

The BET equation is only meaningful over a limited range of relative pressure, traditionally about 0.05 to 0.30 p/p0 for nitrogen. For microporous materials such as zeolites, activated carbons and MOFs this fixed window can be misleading, because the micropores fill at very low pressures. This is why the Rouquerol criteria are applied: the BET constant C must be positive, the quantity n(1-p/p0) must increase with p/p0 over the chosen range, and the p/p0 corresponding to monolayer capacity must lie inside that range. In these cases the number is best read as an apparent surface area that is comparable with others obtained by the same protocol, rather than as the true geometric area of the material.

A single number in a results table says far less about a material than the raw adsorption isotherm does. IUPAC groups isotherms into six main types. Type I rises steeply at low pressure and then levels off, which is characteristic of microporous solids. Type II is found for non-porous or macroporous materials and reflects unrestricted progression from monolayer to multilayer. Type IV is the signature of mesoporous materials: capillary condensation produces a sharp uptake at higher pressure and usually a hysteresis loop. Types III and V appear when the interaction between adsorbate and adsorbent is weak. Type VI is a stepped curve from layer-by-layer adsorption on a very uniform, non-porous surface, and it is rare.

Hysteresis means that the adsorption and desorption branches do not follow the same path, and it is usually associated with capillary condensation in mesopores. The shape of the loop gives only general hints about pore geometry: a narrow loop with near-vertical branches can point to fairly uniform cylindrical pores, while other shapes suggest slit-like spaces between plates or ink-bottle networks with narrow necks. An abrupt closure of the desorption branch near 0.4 p/p0 for nitrogen at 77 K is often a property of the liquid itself becoming unstable, not of a particular pore size, and it should not be read as a pore diameter. The isotherm should always be inspected as a graph, not only as a table of values.

The surface and pores of a porous solid hold moisture, carbon dioxide and other vapours from the surrounding air. These molecules block adsorption sites and lower the measured surface area. Degassing removes them before the measurement by heating the sample under vacuum or under a flow of dry, inert gas. Heating rate, final temperature and hold time depend on the material, and the process is finished when the pressure is confirmed to be stable, not simply when a fixed time has passed.

Temperature is the most critical decision. It must stay below the point at which the sample changes: decomposition, sintering, phase transformation, loss of crystal water or collapse of the framework can alter the surface area irreversibly. Polymers, materials with organic content, MOFs, clays and some hydrated oxides are especially sensitive. Where needed, thermogravimetric analysis (TGA) can help choose a suitable temperature. Vacuum degassing is usually more effective at lower temperature, while a flow method can be more practical for some powders. Finally, the sample mass must be determined after degassing, because the area per gram is calculated from that mass, and the loss of moisture can change it considerably.

Quoting a sample amount in grams can be misleading, because what matters is the total surface area inside the cell. The amount of gas the instrument can measure is proportional to that total area, so a small quantity of a high-area material can give the same signal as a large quantity of a low-area one. As a general rule of thumb, a reliable nitrogen measurement aims for at least several to a few tens of square metres of total surface area in the cell, while far less can be enough with krypton. This is rough guidance only, and the exact figure depends on the instrument, the cell and the material.

If you do not know the surface area in advance, simply say so; related materials, the synthesis route or even an estimated particle size are useful information. Too little sample gives noisy and poorly reproducible results, while far too much lengthens degassing and crowds the cell. The NAKTEM team reviews the suitability of the sample and agrees the preparation with you before the measurement, so that the amount, the degassing temperature and, if needed, the analysis gas are settled from the start.

The same isotherm also yields pore volume and pore size distribution, but these are more model-dependent than the surface area. The BJH method rests on the Kelvin equation, which describes capillary condensation, and is applied to mesopores; it tends to underestimate the size of small mesopores and cannot describe micropores reliably. DFT and related methods assume a defined pore geometry and surface chemistry and can cover a wider range, but the result is only as good as the match between the chosen model and the real material.

A pore size distribution should therefore be presented and read as an estimate. Which branch (adsorption or desorption) is used, which kernel is chosen and which pore shape is assumed can all change the outcome. When comparing pore sizes from different laboratories or methods, check whether these assumptions are the same. Where possible, supporting the result with direct imaging such as SEM or TEM raises confidence in the interpretation.

Insufficient degassing is the most frequent problem: the sample still carries moisture or volatile material, and the surface area comes out too low. An unstable pressure during degassing is sometimes the warning sign. Conversely, too high a temperature changes the sample and spoils the result. Leaving a sample in open air for a long time, even if it was degassed, or transporting the cell without sealing it under inert gas, lets moisture return. Weighing the sample before degassing instead of after is another common error.

Other mistakes concern interpretation: sending too little sample, keeping the p/p0 window fixed for a microporous material without questioning it, reading the closure point of the hysteresis loop as a pore size, and comparing BET values obtained by different methods as if they were equivalent. A different analysis gas, temperature or p/p0 range can give a different number even for the same material. The question to ask is whether two values were obtained with the same protocol. When in doubt, talking to the laboratory about the thermal behaviour and the expected surface area of your material before sending it is often the cheapest solution.