Pore dimension determines the transport regime by comparison with the gas molecular mean free path, not by pore size alone. For gas-filled membrane pores, the key quantity is the Knudsen number, (Kn=\lambda/d_p), where (\lambda) is the molecular mean free path and (d_p) is the pore diameter. When (d_p) is much smaller than (\lambda), wall collisions dominate and transport is Knudsen diffusion; when (d_p) is much larger, intermolecular collisions support continuum transport, including molecular diffusion and, where a pressure gradient drives flow, viscous flow.

The dividing principle is collision frequency: gas molecules colliding mainly with pore walls produce Knudsen transport, while molecules colliding mainly with one another produce continuum behavior. In liquid-filled water-treatment pores, transport remains generally continuum because the liquid molecular mean free path is extremely small; pore dimensions instead control hydrodynamic resistance, steric hindrance, and effective solute diffusivity.

How Pore Size Selects Gas Transport

The Relevant Length-Scale Comparison

The gas mean free path, (\lambda), is the average distance a molecule travels between collisions with other gas molecules. The pore diameter, (d_p), defines how much space is available before a molecule encounters a membrane wall.

Their ratio is expressed as:

[ Kn=\frac{\lambda}{d_p} ]

A large (Kn) means the gas is more rarefied relative to the pore. A small (Kn) means the pore is large enough for ordinary intermolecular collisions to dominate.

When Knudsen Diffusion Dominates

Knudsen diffusion becomes important when the pore diameter is comparable to or smaller than the molecular mean free path. In practical terms, (Kn) greater than approximately 1 indicates that wall collisions are more frequent than intermolecular collisions; (Kn) substantially greater than 10 represents a strongly wall-collision-dominated regime.

The Knudsen diffusivity is commonly represented as:

[ D_{Kn}=\frac{d_p}{3}\sqrt{\frac{2k_BT}{m}} ]

where (T) is temperature, (m) is molecular mass, and (k_B) is Boltzmann's constant.

This relationship shows that Knudsen diffusivity increases with pore diameter and molecular speed, and decreases with the square root of molecular mass.

How Knudsen Selectivity Appears

Because each species interacts independently with the pore walls, the Knudsen flux ratio follows approximately:

[ \frac{J_1}{J_2}\approx\sqrt{\frac{m_2}{m_1}} ]

Thus, lighter gases permeate faster than heavier gases. For example, the idealized Knudsen selectivity of a lighter species over a heavier species is governed by the square-root ratio of their molecular masses.

This selectivity is distinct from ordinary molecular diffusion, where transport depends on binary diffusivities and interactions between the gas species.

When Continuum Transport Dominates

Large Pores Relative to Mean Free Path

When (d_p) is much larger than (\lambda), gas molecules collide primarily with one another rather than with the pore walls. The gas can then be treated approximately as a continuum.

For diffusion driven by concentration or partial-pressure gradients, transport is described using ordinary molecular diffusion. The relevant diffusivity depends on gas composition, temperature, pressure, and molecular interactions.

Pressure-Driven Viscous Flow

A pressure gradient can also produce continuum viscous flow through sufficiently large pores. In this regime, the Hagen-Poiseuille description is appropriate for idealized cylindrical pores.

For a porous membrane, the pressure dependence helps distinguish mechanisms:

Knudsen diffusion: flux is approximately proportional to the pressure difference.
Continuum viscous flow: flux depends strongly on pressure, often appearing proportional to the mean pressure multiplied by the pressure difference, or to a pressure-squared difference depending on the formulation.
Continuum molecular diffusion: flux is governed primarily by the species concentration or partial-pressure gradient.

The exact observed behavior also depends on membrane structure, gas compressibility, and whether the pores are straight, tortuous, or interconnected.

The Transition Regime

The boundary between mechanisms is not a single exact pore diameter. A useful engineering classification is:

(Kn < 0.01): continuum flow is generally valid.
(0.01 < Kn < 0.1): slip-flow corrections may be needed.
(0.1 < Kn < 10): transition behavior occurs.
(Kn > 10): Knudsen or free-molecular behavior becomes dominant.

These ranges are guidelines rather than universal discontinuities. A pilot-plant model should account for mixed transport when the pore diameter and mean free path are of similar magnitude.

Why Operating Pressure Matters

Pressure Changes the Mean Free Path

At fixed temperature, increasing gas pressure generally decreases the molecular mean free path. The same membrane can therefore operate in different transport regimes as feed or permeate pressure changes.

A small pore may exhibit Knudsen behavior at low pressure and move toward transition or continuum behavior as pressure increases.

Pilot-Plant Identification

A gas-permeation pilot plant can test the mechanism by measuring steady-state flux while varying feed pressure and pressure difference. Plotting flux against the appropriate pressure variables can reveal whether the membrane is operating predominantly in Knudsen diffusion, viscous flow, or a mixed regime.

This diagnostic is important because fitting all data to a single transport equation can produce an apparently reasonable but physically incorrect pore-size estimate.

What Pore Dimension Means in Water Treatment

Liquid-Filled Pores Are Not Usually Knudsen Pores

Knudsen diffusion is primarily a gas-phase rarefaction phenomenon. In water-filled membrane pores, liquid molecules collide with one another at length scales far smaller than typical membrane pores, so the liquid normally behaves as a continuum.

Consequently, reducing a water-treatment pore from the micrometer to nanometer scale does not automatically create Knudsen flow.

Solute Size Controls Hindered Diffusion

For nanofiltration, ultrafiltration, and related processes, the important comparison is often solute diameter to pore diameter:

[ \lambda_s=\frac{d_s}{d_p} ]

As this ratio increases, the solute experiences greater steric exclusion and hydrodynamic drag near the pore wall. Its effective diffusivity becomes lower than its bulk-solution diffusivity.

A solute that is small relative to the pore may diffuse nearly as it does in bulk water. A solute approaching the pore dimension can experience substantial reduction in diffusion and convective passage.

Dense Membranes Use a Different Model

Reverse osmosis and many dense nanofiltration membranes do not transport species through open pores in the conventional capillary sense. Their behavior is more appropriately described by solution-diffusion: water and solutes dissolve into the membrane material and diffuse through it according to their respective chemical potentials and mobilities.

For these membranes, apparent pore diameter is not the primary criterion for selecting Knudsen versus continuum transport.

Combining Transport Resistances

Mixed Gas Diffusion

When gas molecular collisions and wall collisions are both important, the effective diffusivity can be represented using a combined resistance relationship:

[ \frac{1}{D_{\text{eff}}}

\frac{1}{D_m} + \frac{1}{D_{Kn}} ]

Here, (D_m) is the ordinary molecular diffusivity and (D_{Kn}) is the Knudsen diffusivity.

This form is commonly used as an engineering approximation for combined molecular and Knudsen diffusion, though real membranes may require corrections for porosity, tortuosity, pore connectivity, and surface effects.

Porous-Membrane Properties Still Matter

The pore diameter sets the local collision regime, but it does not alone determine total membrane flux. Effective transport also depends on membrane thickness, porosity, tortuosity, pore-size distribution, and the pressure and concentration boundary conditions.

A membrane with small pores may have Knudsen-controlled transport but low overall throughput because its pores are long or highly tortuous.

Understanding the Trade-offs

Small Pores Can Increase Gas Selectivity

Knudsen transport favors lighter gases according to the inverse square root of molecular mass. This can provide useful selectivity when the target gas pair has an appropriate molecular-weight difference.

However, the selectivity is often modest compared with highly selective solution-diffusion membranes or chemically functionalized materials.

Small Pores Reduce Permeability

Because Knudsen diffusivity is proportional to pore diameter, shrinking the pores reduces the intrinsic diffusivity through each pore. A design that increases wall-collision effects may therefore reduce productivity.

Pilot-plant optimization must balance selectivity against membrane area, pressure requirements, and acceptable flux.

Gas and Liquid Criteria Must Not Be Confused

Applying (Kn=\lambda/d_p) directly to a water-treatment pore is usually inappropriate. Gas membranes are classified by the gas mean free path, whereas liquid membranes are evaluated using continuum hydrodynamics, solute-to-pore size ratios, and membrane phase behavior.

Continuum Flow Is Not the Same as Molecular Diffusion

When (Kn) is small, continuum treatment is justified, but the resulting transport may be either pressure-driven viscous flow or concentration-driven molecular diffusion. The driving force and boundary conditions must be identified before selecting the flux equation.

Making the Right Choice for Your Goal

Use the length-scale comparison first, then select the transport model that matches the phase and driving force.

If your primary focus is gas-separation mechanism identification: Calculate (Kn=\lambda/d_p) at the actual feed and permeate pressures and temperatures, and include a mixed molecular-Knudsen model in the transition range.
If your primary focus is gas membrane selectivity: Test whether measured flux ratios follow the inverse square-root molecular-mass relationship expected for Knudsen diffusion.
If your primary focus is water-treatment membrane design: Compare solute size with pore size and model steric hindrance, hydrodynamic drag, and convection rather than applying a gas Knudsen criterion.
If your primary focus is pilot-plant scale-up: Measure flux across the operating pressure range and account for porosity, tortuosity, thickness, concentration polarization, and pressure-dependent regime changes.

Correctly comparing the characteristic molecular length with the membrane pore dimension is the foundation for choosing a defensible transport model.

Summary Table:

Application	Key comparison	Dominant transport	Main design consideration
Gas-filled pores	Knudsen number, `Kn = λ/dₚ`	Knudsen diffusion when `Kn > 1`; continuum behavior when `Kn < 0.1`	Balance gas selectivity against permeability and pressure requirements
Gas transition regime	`0.1 < Kn < 10`	Mixed molecular and Knudsen diffusion	Use combined-resistance models and pressure-dependent testing
Gas continuum regime	Pore diameter much larger than mean free path	Molecular diffusion or viscous flow	Identify whether concentration or pressure drives transport
Water-filled pores	Liquid mean free path versus pore size	Generally continuum transport	Evaluate hydrodynamic resistance, convection, and concentration polarization
Water-treatment membranes	Solute diameter versus pore diameter	Hindered diffusion and steric exclusion	Account for solute size, pore distribution, porosity, and tortuosity
Dense RO/NF membranes	Species solubility and mobility in the membrane	Solution-diffusion	Do not apply an open-pore Knudsen model

Select the Right Membrane Transport Model

Design and validate your pilot plant with LABPARK Educational and Vocational Unit Operations Pilot Plants. Our systems support gas separation, bioprocess and biotech research, and environmental and water-treatment applications for universities, research institutes, and enterprises.

LABPARK can help you evaluate pore-size effects, pressure-dependent gas transport, mixed Knudsen and molecular diffusion, liquid-phase hydrodynamics, solute rejection, and pilot-scale operating performance. Use reliable experimental data to improve membrane selection, teaching outcomes, process development, and scale-up decisions.

Contact LABPARK to discuss a pilot plant matched to your membrane process, operating conditions, research objectives, and training requirements.