Surface plasmon resonance, from the physics up.

Surface plasmon resonance (SPR) is an optical technique for measuring biomolecular binding in real time without labelling either partner. Under total internal reflection at a thin metal film, p-polarised light can couple into a collective oscillation of the metal’s conduction electrons — a surface plasmon — but only at one specific angle of incidence. That resonance angle depends on the refractive index of the medium within roughly 200–300 nm of the metal. When molecules bind to receptors immobilised on the surface, the local refractive index rises and the resonance angle shifts. Plotting that shift against time gives a sensorgram, from which association and dissociation rate constants can be extracted.

Refractive index within the evanescent field — roughly 200–300 nm of solution adjacent to the sensor surface. Binding is inferred from that measurement, because bound material displaces buffer and raises the local refractive index. The distinction is not pedantic: a temperature fluctuation, a mismatch between sample and running buffer, or non-specific adsorption to the matrix all raise refractive index too, and the detector cannot distinguish them from specific binding. Only experimental design, reference channels and blank subtraction can.

They are the parameters a kinetic experiment returns. The association rate constant ka (M⁻¹ s⁻¹) is the second-order rate constant for complex formation; the dissociation rate constant kd (s⁻¹) is the first-order rate constant for its breakdown. Their ratio kd/ka is the equilibrium dissociation constant KD, in units of concentration, equal to the analyte concentration at which half the surface sites are occupied at equilibrium. Because KD is a ratio, two interactions with identical KD can differ by orders of magnitude in complex lifetime — which is ln2/kd — so reporting KD alone discards information the experiment already provided.

Flow in an SPR flow cell is laminar, so analyte reaches the surface by diffusion across an unstirred layer rather than by convective mixing. If the immobilised density is high and association is fast, the surface depletes analyte faster than diffusion resupplies it, and the observed rate reflects transport rather than chemistry. This is mass transport limitation. Its signatures are a near-linear association phase, a dissociation that decays quickly then drags because released analyte rebinds, and a fitted ka that increases when you lower the ligand density. The standard remedy is to target a saturation response of roughly 20–100 response units for kinetic work.

They address overlapping questions but rest on different physics. SPR excites a surface plasmon in a metal film and detects a shift in the resonance condition, with sample delivered through a microfluidic flow cell. BLI measures interference between light reflected from two closely spaced surfaces on a fibre-optic tip, which is dipped into a stirred well; no plasmon is involved. The practical consequences differ accordingly: BLI parallelises trivially and needs no fluidics, but has higher noise and a less controlled transport regime, particularly during dissociation. Describing a BLI measurement as SPR, which happens in the literature, is simply an error of naming.

A sensorgram is the response-versus-time trace an SPR instrument produces. A standard analysis cycle has five phases: a flat baseline in running buffer, an association phase during analyte injection, an optional steady-state plateau, a dissociation phase once buffer flow resumes, and a regeneration pulse that strips remaining analyte. For a simple 1:1 interaction both binding phases are exponential — association approaches its plateau with observed rate constant kobs = ka·C + kd, and dissociation decays with rate constant kd alone. Departures from exponential behaviour are diagnostic rather than incidental, and identifying them by eye is faster and more reliable than discovering them from a poor fit.

No. The interactive figures compute real optics and integrate the binding rate equations in your browser, so you can explore how the parameters interact without hardware. The reflectivity figure runs a transfer-matrix Fresnel calculation over the full layer stack; the kinetics figures numerically integrate a two-compartment model that includes mass transport. Many readers use the site to plan before booking instrument time, or to interpret data a collaborator has sent them.

Free, with no account and no paywall. On reliability: every substantive statement carries an inline citation to the paper that first reported the result rather than to a review repeating it, and each reference is labelled with whether its bibliographic record has been machine-verified against Crossref — where it has not been, the site says so rather than implying otherwise. Numerical claims in the text are recomputed from the governing equations each time the site is built, and the build fails if any of them drifts.