This is a side-on cross-section of the same graphene-oxide (GO) laminate as the 3D version, but the transport itself is computed differently: instead of a single "does it fit?" test at the membrane face, each channel carries a real parabolic Poiseuille velocity profile across its height, and each solute's fate is drawn from a continuous Renkin steric-hindrance probability rather than a hard yes/no cutoff.
Velocity profile: u(y) = u_max·(1 − (2y/h)²), u_max = ΔP·h²/(8μL)
Slit flow rate: Q = ∫u dy = ΔP·h³/(12μL) (per unit width, per channel)
Steric ratio: λ = δ_hydrated / h_eff
Renkin partition: Φ(λ) = (1−λ)²·(1 − 2.104λ + 2.09λ³ − 0.95λ⁵), λ<1; else 0
- Particles move faster near mid-channel, slower near the walls — a direct consequence of the parabolic profile, visualised per-particle rather than only quoted as a stat. The 3D sibling sim states this relationship in its theory box but doesn't actually apply it to particle motion; here it drives the animation.
- Φ(λ) replaces the binary cutoff — real hindered transport through a slit falls off smoothly as a solute's hydrated size approaches the opening, not with a sudden on/off switch at one exact diameter. A species is let through (or bounced) with probability Φ each time it reaches the membrane face, so the "Passed / Blocked" tally converges to the theoretical rejection = 1 − Φ over many attempts.
- Water permeance is reported in real units (L·m⁻²·h⁻¹·bar⁻¹, "LMH/bar" — the standard membrane-science unit) from J = h_eff²·ΔP / (12·μ_water·L·N) using water's actual viscosity, not an arbitrary score.
- Water is exempt from the λ test — this is a deliberate correction versus the 3D sim: at its own minimum spacing slider (d = 6 Å), h_eff drops to 2.5 Å, below water's own 2.8 Å diameter, so the 3D model's literal "diam < h_eff" rule would actually reject water there too — contradicting its own legend ("Water — always slips through"). Bulk water permeation through sub-nm GO capillaries is well documented experimentally regardless of this edge case, so this model special-cases water to always pass, matching the physical claim rather than the arithmetic accident.
Drag d down and watch λ climb toward 1 for Na⁺ and Mg²⁺ — Φ collapses non-linearly, so rejection rises steeply rather than snapping on at one threshold.