Why Gaussian Profiles?¶

The three-shell model treats the lipid bilayer as three rectangular slabs — outer headgroup, hydrocarbon chains, inner headgroup — separated by perfectly sharp boundaries. When you plot this as an electron density profile, each transition is an instantaneous step from one value to the next.

This is mathematically convenient, but it is not physically accurate.

What a real bilayer looks like¶

A lipid molecule is not rigidly fixed in the membrane. At physiological temperatures the bilayer is in a fluid phase — lipid chains are disordered, molecules diffuse laterally, and individual headgroups fluctuate in and out of the water layer on nanosecond timescales.

Three physical effects conspire to make the electron density profile of a real bilayer continuous rather than stepped:

1. Thermal fluctuations of individual lipids. At any moment in time, lipid molecules within a single leaflet are at slightly different axial positions. When averaged over time and across the illuminated sample volume, the sharp edge of the headgroup layer is replaced by a distribution of positions — well described by a Gaussian.

2. Water penetration. Water molecules are not excluded from the headgroup region. They hydrogen-bond to the phosphate and choline groups and penetrate several ångströms into what the step-function model treats as a uniform headgroup slab. The headgroup-water boundary is a gradient, not a wall.

3. Bilayer undulations. The vesicle surface itself fluctuates on longer length scales. These undulations are averaged out by the scattering measurement, and the effect at the level of the electron density profile is again an apparent broadening of every interface.

The cumulative result is an electron density profile that looks like a pair of smooth peaks (the headgroups) flanking a smooth trough (the hydrocarbon chains):

Where the step-function model breaks down¶

The sharp-interface (shell) model works well at small to intermediate \(q\) where the measurement is sensitive to the overall bilayer thickness and vesicle size. At larger \(q\) — where the measurement resolves the internal structure of the bilayer — the two models diverge in a predictable way.

A sharp interface contributes scattering at all \(q\) values with no decay. A Gaussian interface contributes a scattering amplitude that decays as \(\exp(-q^2\sigma^2/2)\), where \(\sigma\) is the interface width. The broader the interface, the faster this decay. At high \(q\), the shell model over-predicts the oscillation amplitude because it assumes perfectly sharp transitions that do not exist in a real membrane.

Three practical situations where the Gaussian model matters:

1. Fitting data at \(q > 0.3\) Å\(^{-1}\) — oscillations in this range are sensitive to interface sharpness. A sharp-interface model will systematically fail here.

2. Extracting interface widths — \(\sigma_H\) and \(\sigma_C\) are physically meaningful parameters that can be compared to MD simulations and reflectometry results.

3. Modeling asymmetric bilayers — the inner and outer headgroup layers can have different widths. The shell model has no way to represent this; the Gaussian model handles it naturally by using independent parameters for each leaflet.

What this model cannot do¶

The Gaussian model is still a model. It assumes:

• The bilayer electron density profile is well described by three Gaussians (adequate for most common phospholipids, but not for all lipid mixtures)
• The vesicle is monodisperse (polydispersity must be handled separately)
• The bilayer is symmetric between leaflets (relaxed in Section 5 page 5)

It does not account for inter-vesicle interactions (structure factor), multilamellarity, or cholesterol-induced changes to chain order that alter the profile shape.

What's next: Building the Electron Density Profile — constructing the POPC bilayer profile in Python and visualizing it.