Brunauer–Emmett–Teller (BET) theory is a model of multilayer physical adsorption on solid surfaces. Its principal practical use is to estimate specific surface area from a gas adsorption isotherm, using an inferred monolayer capacity and an assumed molecular cross-sectional area. The theory and the associated measurement method are closely related but distinct: the theory describes an idealized adsorption process, whereas the method applies its equation to experimental data. (goldbook.iupac.org)
Origins and physical assumptions
Stephen Brunauer, Paul Hugh Emmett, and Edward Teller introduced the theory in their 1938 paper Adsorption of Gases in Multimolecular Layers. Their approach extended the Langmuir adsorption model, which describes monolayer adsorption, to allow molecules to accumulate in successive layers. (doi.org)
The idealized BET model assumes that:
- The solid has energetically equivalent adsorption sites.
- Each adsorbed molecule can provide a site for a molecule in the next layer.
- Interactions between neighboring molecules within a layer are neglected.
- The first layer has distinct adsorption energetics, while subsequent layers have liquid-like properties.
- There is no upper limit to the number of layers.
- Adsorption and desorption are balanced at equilibrium. (doi.org)
These assumptions do not require a layer to become complete before molecules begin occupying higher layers. Consequently, the inferred monolayer capacity is an equivalent amount sufficient to cover the surface, not necessarily an experimentally observed state containing exactly one complete layer and nothing above it. (tau.ac.il)
The BET equation
Define relative pressure as
[ x=\frac{p}{p_0}, ]
where (p) is the equilibrium gas pressure and (p_0) is the saturation vapor pressure at the measurement temperature. The BET equation is
[ n=\frac{n_m Cx}{(1-x)\left[1+(C-1)x\right]}, ]
where (n) is the amount adsorbed, (n_m) is the monolayer capacity, and (C) is the dimensionless BET constant. Amounts may be expressed either for the entire sample or per unit sample mass, provided the convention is consistent. (tau.ac.il)
Its traditional linear form is
[ \frac{x}{n(1-x)}
\frac{1}{n_m C} + \frac{C-1}{n_m C}x. ]
If the plot has slope (s) and intercept (i),
[ n_m=\frac{1}{s+i}, \qquad C=1+\frac{s}{i}. ]
These relations make it possible to obtain both parameters by linear regression over a selected pressure interval. (tau.ac.il)
A traditional approximate interpretation is
[ C\approx\exp!\left(\frac{E_1-E_L}{RT}\right), ]
where (E_1) and (E_L) are positive magnitudes associated with first-layer adsorption and liquefaction, respectively, and (R) is the gas constant. This interpretation involves simplifying assumptions about adsorption and desorption prefactors; (C) is not a direct calorimetric measurement. (tau.ac.il)
Conversion to surface area
When (n_m) is expressed in moles for the whole sample, the BET surface area is
[ A_{\mathrm{BET}}=n_mN_A\sigma, ]
where (N_A) is the Avogadro constant and (\sigma) is the assumed effective cross-sectional area of an adsorbed molecule. For sample mass (m),
[ a_{\mathrm{BET}}=\frac{A_{\mathrm{BET}}}{m}. ]
If (n_m) is already expressed in moles per unit mass, the first expression directly gives specific surface area. (researchgate.net)
For nitrogen, the conventional cross-sectional area is (0.162\ \mathrm{nm^2}). Specific areas are commonly reported in (\mathrm{m^2,g^{-1}}), although the corresponding SI unit is (\mathrm{m^2,kg^{-1}}). The result can include internal pore-wall area rather than merely the external geometric surface. (goldbook.iupac.org)
Experimental measurement
An isotherm records gas uptake at constant temperature over a sequence of equilibrium pressures. Uptake may be determined manometrically, from gas dosing and pressure changes, or gravimetrically, from changes in sample mass. Before measurement, the sample is outgassed to remove previously adsorbed substances without altering its structure. Accurate temperature control, void-volume determination, and sufficient equilibration are important sources of measurement uncertainty. (researchgate.net)
Nitrogen at about 77 K is a conventional probe gas; argon at about 87 K is also used. The choice matters because molecular interactions and access to narrow pores differ between gases. Nitrogen’s quadrupole moment can affect its orientation and the effective molecular cross-sectional area on particular surfaces. (doi.org)
Selection of the fitting range
A high-quality straight line is not, by itself, evidence that a BET analysis is physically appropriate. Consistency tests commonly associated with Rouquerol and collaborators examine whether:
- The fitted (C) is positive.
- (n(1-x)) increases throughout the selected interval.
- The pressure corresponding to the inferred monolayer capacity falls inside that interval.
- The monolayer pressure implied by the model agrees adequately with that obtained from the experimental isotherm. (sciencedirect.com)
Setting (n=n_m) in the BET equation gives
[ x_m=\frac{1}{1+\sqrt C}. ]
This is the model’s pressure at which the total uptake equals the monolayer capacity; it does not imply that the first layer is completely occupied. (pubs.acs.org)
The often-used interval (0.05\lesssim p/p_0\lesssim0.30) is suitable for some isotherms, not a universal prescription. Appropriate ranges can shift to much lower relative pressures. (researchgate.net)
Applications, limitations, and reproducibility
BET analysis is widely used to characterize fine powders and porous materials, including zeolites, nanostructured silicas, and porous coordination polymers. It supplies a common surface-area metric for comparing samples and processing conditions. (nist.gov)
Its interpretation is most problematic when adsorption departs from layer-by-layer accumulation:
- Micropore filling: In pores approximately 2 nm wide or smaller, uptake cannot generally be interpreted solely as surface coverage.
- Pore condensation: At higher pressures, liquid-like filling of pores introduces a process outside the basic BET model.
- Restricted accessibility: The measured internal area depends on the size and shape of the probe molecule. (researchgate.net)
Thus, a BET area is a model-dependent estimate, not an exact geometric measurement. BET analysis also does not itself yield a pore-size distribution; that requires additional models and assumptions. (researchgate.net)
Data reduction introduces a separate reproducibility issue. A 2022 interlaboratory study supplied 18 identical adsorption isotherms to 61 laboratories and obtained widely varying calculated BET areas. The researchers developed automated analysis based on expanded consistency criteria to reduce ambiguity in interval selection. Standardizing calculations improves comparability, but does not eliminate the physical limitations of the adsorption model. (onlinelibrary.wiley.com)