Molecular Weight in Diffusion-Related Mechanistic Predictions

Why Molecular Weight Enters Diffusion Estimates

PBPK models sometimes need a mechanistic estimate of how rapidly a molecule moves through an aqueous medium. Molecular weight is often available early, and it can serve as a rough proxy for molecular size: for molecules of similar density, shape, and hydration, a larger molecular weight tends to correspond to a larger effective radius and a lower diffusivity. This makes molecular-weight scaling useful for an initial estimate or sensitivity analysis.

The relationship is conditional, not a universal law that uniquely maps molecular weight to mobility. Diffusion depends more directly on a molecule’s hydrodynamic size and on the medium through which it moves. A mass-based estimate therefore requires assumptions about molecular compactness, solvent viscosity, temperature, and the molecule’s interaction with its surroundings.

Definition: The diffusion coefficient \(D\) describes the proportionality between the flux of a freely diffusing species and its concentration gradient in a specified medium. Its SI units are \(\mathrm{m^2/s}\). In this tutorial, \(MW\) denotes molar mass, reported in \(\mathrm{g/mol}\) or \(\mathrm{kg/mol}\); the numerical value in \(\mathrm{g/mol}\) is conventionally also called molecular weight.

From Molecular Weight to a Size-Based Correlation

A common first approximation treats a molecule as a spherical particle moving through a continuous, Newtonian solvent. In the Stokes–Einstein model, its diffusion coefficient is

D=\frac{k_B T}{6\pi\eta r_h}

Here \(k_B\) is the Boltzmann constant, \(T\) is absolute temperature in kelvin, \(\eta\) is the dynamic viscosity of the medium in \(\mathrm{Pa\,s}\), and \(r_h\) is the hydrodynamic radius in metres. The hydrodynamic radius is the radius of an equivalent sphere that experiences the same frictional resistance to motion. It can reflect hydration and molecular shape; it is not necessarily the radius of a hard, dry molecular boundary.

To connect radius to molecular weight, make the additional approximation that the molecule is a compact sphere with density \(\rho\), in \(\mathrm{kg/m^3}\). For a molecule with molar mass \(MW\) in \(\mathrm{kg/mol}\), the volume per molecule is its molar mass divided by density and Avogadro’s constant \(N_A\). Equating this volume to the volume of a sphere gives:

V_{\text{molecule}}=\frac{MW}{\rho N_A} =\frac{4}{3}\pi r^3 \qquad\Longrightarrow\qquad r=\left(\frac{3MW}{4\pi\rho N_A}\right)^{1/3}

If density, shape, hydration, temperature, and solvent viscosity are treated as comparable across compounds, the estimated hydrodynamic radius scales with \(MW^{1/3}\). Substituting that radius into Stokes–Einstein yields the inverse-cube-root relationship for diffusivity.

D\propto \frac{1}{r_h} \qquad\text{and, under the compact-sphere assumptions,}\qquad D\propto MW^{-1/3}

This scaling is a relative estimate under specified assumptions. It does not say that every molecule with twice the molecular weight has the same diffusivity, nor does it supply an absolute diffusivity without a reference or additional physical inputs.

Key result: If two molecules diffuse in the same medium at the same temperature and are assumed to have comparable density, shape, and hydration, a convenient first-pass correlation is \(D_2=D_1(MW_1/MW_2)^{1/3}\). The assumptions, not the algebra, determine whether the estimate is useful.

Worked Estimates

Worked Example: Scaling From a Reference Diffusivity

Hypothetical data. A reference small molecule has \(MW_1=300\ \mathrm{g/mol}\) and an aqueous diffusivity \(D_1=6.00\times10^{-10}\ \mathrm{m^2/s}\). Estimate the diffusivity of a second molecule with \(MW_2=600\ \mathrm{g/mol}\), assuming both are compact, similarly hydrated molecules in the same aqueous medium at the same temperature.

Use the size-scaling relationship:

D_2=D_1\left(\frac{MW_1}{MW_2}\right)^{1/3}

Substitute the values. The ratio of molecular weights is dimensionless, so the diffusivity retains the units of \(D_1\):

D_2=6.00\times10^{-10} \left(\frac{300}{600}\right)^{1/3} =6.00\times10^{-10}(0.500)^{1/3} \approx 6.00\times10^{-10}(0.7937) =4.76\times10^{-10}\ \mathrm{m^2/s}

Under the stated assumptions, the heavier molecule’s estimated diffusivity is about \(4.76\times10^{-10}\ \mathrm{m^2/s}\), or 79.4% of the reference value. Doubling molecular weight does not halve the estimated diffusivity: the assumed radius increases by a factor of \(2^{1/3}\), so diffusivity decreases by its reciprocal.

Worked Example: Estimating Radius and Diffusivity From Physical Inputs

Hypothetical data. Estimate the compact-sphere radius and Stokes–Einstein diffusivity for a molecule with \(MW=450\ \mathrm{g/mol}\), assuming molecular density \(\rho=1.20\times10^3\ \mathrm{kg/m^3}\), temperature \(T=310\ \mathrm{K}\), and medium viscosity \(\eta=6.91\times10^{-4}\ \mathrm{Pa\,s}\). Use \(N_A=6.022\times10^{23}\ \mathrm{mol^{-1}}\) and \(k_B=1.381\times10^{-23}\ \mathrm{J/K}\).

First convert molar mass to SI units: \(450\ \mathrm{g/mol}=0.450\ \mathrm{kg/mol}\). The volume per molecule under the density assumption is:

V_{\text{molecule}}= \frac{0.450\ \mathrm{kg/mol}} {(1.20\times10^3\ \mathrm{kg/m^3})(6.022\times10^{23}\ \mathrm{mol^{-1}})} =6.23\times10^{-28}\ \mathrm{m^3}

Treat this volume as spherical. The resulting radius is:

r=\left(\frac{3(6.23\times10^{-28}\ \mathrm{m^3})}{4\pi}\right)^{1/3} =5.30\times10^{-10}\ \mathrm{m}

Now substitute this radius into the Stokes–Einstein equation. Since \(1\ \mathrm{Pa\,s}=1\ \mathrm{kg/(m\,s)}\), the resulting units reduce to \(\mathrm{m^2/s}\):

D= \frac{(1.381\times10^{-23}\ \mathrm{J/K})(310\ \mathrm{K})} {6\pi(6.91\times10^{-4}\ \mathrm{Pa\,s})(5.30\times10^{-10}\ \mathrm{m})} \approx 6.20\times10^{-10}\ \mathrm{m^2/s}

The estimated diffusivity is \(6.20\times10^{-10}\ \mathrm{m^2/s}\) in the specified medium. It is a model-based value: the calculation assumes that the assigned density and spherical radius adequately represent the molecule’s hydrodynamic size.

Worked Example: Equal Molecular Weight Does Not Imply Equal Mobility

Hypothetical comparison. Consider two molecules, each with \(MW=400\ \mathrm{g/mol}\), in the same medium at the same temperature. Suppose their hydrodynamic radii, determined for this illustration, are \(0.45\ \mathrm{nm}\) and \(0.70\ \mathrm{nm}\). Use the Stokes–Einstein equation with \(T=310\ \mathrm{K}\), \(\eta=6.91\times10^{-4}\ \mathrm{Pa\,s}\), and \(k_B=1.381\times10^{-23}\ \mathrm{J/K}\).

The shared numerator is \(k_BT=(1.381\times10^{-23})(310)=4.281\times10^{-21}\ \mathrm{J}\). For the molecule with radius \(0.45\ \mathrm{nm}=0.45\times10^{-9}\ \mathrm{m}\):

D_{0.45}= \frac{4.281\times10^{-21}} {6\pi(6.91\times10^{-4})(0.45\times10^{-9})} \approx 7.31\times10^{-10}\ \mathrm{m^2/s}

For the molecule with radius \(0.70\ \mathrm{nm}=0.70\times10^{-9}\ \mathrm{m}\):

D_{0.70}= \frac{4.281\times10^{-21}} {6\pi(6.91\times10^{-4})(0.70\times10^{-9})} \approx 4.70\times10^{-10}\ \mathrm{m^2/s}

Their predicted diffusivity ratio is \(D_{0.45}/D_{0.70}=0.70/0.45\approx1.56\). Thus, even at identical molecular weight, the molecule with the smaller hydrodynamic radius is predicted to diffuse about 1.56 times as fast under this model. Molecular weight alone cannot distinguish these cases.

Choosing and Checking an Estimate

The appropriate approach depends on the available information and the purpose of the PBPK calculation. The following methods are related, but they do not answer exactly the same question.

Approach	What it uses	Main limitation
Reference-compound scaling	A known or assumed \(D\) and a molecular-weight ratio	Relies on comparable density, shape, hydration, and medium conditions
Stokes–Einstein estimate	Temperature, viscosity, and hydrodynamic radius	Needs a defensible radius and approximates the medium as a continuum
Direct diffusion measurement	Observed movement in a defined experimental medium	Applies to the measured conditions and species; it may not represent another medium or model context

A direct measurement under relevant conditions can avoid the need to infer diffusivity from mass, but it is not automatically transferable to a different temperature, solvent composition, or molecular state. Stokes–Einstein makes the physical dependence on radius and viscosity explicit. Reference scaling is simpler and useful when information is limited, but it inherits all the similarity assumptions of the reference comparison.

Before using a size-based estimate, check that the input units are consistent, that temperature is in kelvin, and that viscosity is expressed in \(\mathrm{Pa\,s}\) if SI units are used. A useful diagnostic is the direction of the result: at fixed temperature, viscosity, and radius, the Stokes–Einstein equation predicts higher \(D\) at higher \(T\) and lower \(D\) at higher \(\eta\) or \(r_h\). An estimate that reverses one of these directions usually signals a substitution or unit error.

Also ask whether the inferred radius is plausible for the molecule and whether a small change in the scaling assumptions materially changes the prediction. A molecular-weight correlation can be used as a baseline, then varied in sensitivity analysis to show how uncertain diffusivity affects model predictions. It should not be presented as a measured property when it is only a correlation-based estimate.

Where Molecular-Weight Scaling Can Fail

Molecules with the same molar mass can differ in shape, flexibility, hydration, and charge distribution. These features affect the hydrodynamic radius and interactions with the medium. A branched or extended structure need not have the same frictional resistance as a compact structure of equal mass. Strong hydration can increase effective size, while association with other molecules can make the diffusing species larger than an isolated molecule.

The medium matters as well. Stokes–Einstein includes viscosity explicitly, but real biological environments can be heterogeneous and may not behave like a uniform Newtonian solvent at every relevant scale. Binding or transient interactions can slow observed transport relative to the motion of freely dissolved molecules. A free-solution diffusion estimate therefore need not equal an effective transport rate in a tissue environment.

Finally, diffusivity is not the same quantity as membrane permeability or tissue-to-plasma partitioning. Diffusivity describes movement within a medium; permeability additionally depends on crossing a barrier, and partitioning describes relative concentrations between phases at equilibrium. A molecular-weight correlation for \(D\) does not, by itself, determine either of those properties.

Common Pitfall

Wrong reasoning: “The second molecule has twice the molecular weight, so it must diffuse at half the rate.” This treats molecular weight as if it were directly proportional to frictional resistance.

Correction: Under the compact-sphere assumptions, radius scales with the cube root of molecular weight, and Stokes–Einstein makes diffusivity inversely proportional to radius. The corresponding estimate is \(D_2/D_1=(MW_1/MW_2)^{1/3}\), not \(MW_1/MW_2\). Even that cube-root result is only a conditional approximation: shape, hydration, charge, binding, and medium conditions can make equal-mass molecules behave differently.

Key takeaway: Use molecular weight to make a transparent, assumption-dependent first-pass estimate of molecular size or diffusivity. For mechanistic interpretation, track the hydrodynamic radius and medium conditions, and state clearly when the estimate may not represent molecular mobility in the relevant biological environment.

Check Your Understanding

Answer each question, then compare with the short solutions below.

Under the compact-sphere assumptions, how does radius scale when molecular weight increases by a factor of 8?
A reference compound has \(MW_1=200\ \mathrm{g/mol}\) and \(D_1=5.00\times10^{-10}\ \mathrm{m^2/s}\). Estimate \(D_2\) for \(MW_2=400\ \mathrm{g/mol}\) under otherwise comparable conditions.
Why can two molecules with identical molecular weights have different Stokes–Einstein diffusivities?
Name two medium or molecular conditions that should be checked before treating a free-solution estimate as representative of transport in a biological environment.
Does a molecular-weight-based diffusivity estimate determine membrane permeability? Explain briefly.

Solutions. (1) Radius scales as \(MW^{1/3}\), so it increases by \(8^{1/3}=2\). (2) \(D_2=5.00\times10^{-10}(200/400)^{1/3}=5.00\times10^{-10}(0.7937)=3.97\times10^{-10}\ \mathrm{m^2/s}\). (3) They may have different hydrodynamic radii because their shapes, hydration, or interactions differ; at the same \(T\) and \(\eta\), Stokes–Einstein gives different \(D\) when \(r_h\) differs. (4) Examples include medium viscosity and temperature, as well as binding or association that changes effective mobility. (5) No. Diffusivity describes movement within a medium; permeability also depends on crossing a barrier and cannot be obtained from molecular weight alone.