The MACS Matchmaker is a benchtop instrument for label-free biomolecular interaction analysis. It is built on Focal Molography (FM), an optical biosensing technology that detects molecular binding events in real time and provides quantitative data on binding affinity, kinetics, and specificity. Compared with purely refractometric sensors, FM strongly suppresses common-mode non-specific adsorption and bulk refractive-index changes when ridge and groove conditions remain matched, making it well suited to measurements in complex biological matrices such as serum, plasma, and cell lysates (Reichmuth et al. 2021).

Principles of Focal Molography

Focal molography translates molecular binding events into optical signals through diffractive interference. The key component is the mologram: a nanostructured diffraction grating covering a 400 × 400 µm area, fabricated directly on the sensor chip by photolithography (Frutiger et al. 2019). Across that width the pattern alternates on a sub-micrometre period — roughly a thousand ridges interleaved with as many grooves:

• Ridges (about 200 nm wide, half of the period; Gatterdam et al. 2017) — functionalized with specific capture molecules (ligands). This is where specific analyte binding is intended to occur.
• Grooves (the other half of each period) — the built-in reference region, interdigitated at the nanoscale with every ridge. What fills them depends on the chip: a passivating PEG layer on the [Oligo|PEG] and [Me-Tz|PEG] variants, or a chemically matched, non-target-binding backfill probe on the backfill-capable [Oligo|Oligo] and [Me-Tz|Tz]. The reference works because the groove resembles the ridge to everything except the analyte — which a passivated groove achieves only in simple media, and a backfilled one is built to achieve in complex ones.

A laser beam coupled into the Ta₂O₅ dielectric waveguide propagates as a guided mode along the chip, generating an evanescent field that reaches about 80 nm above the surface — the sensing volume, and the reason only what binds close to the chip is seen at all (Frutiger et al. 2019). Light diffracted by mass on the ridges converges constructively at a single focal point, forming a bright focal spot recorded by a camera. The grooves diffract out of phase and cancel destructively at the same point. The MACS Matchmaker chip contains 64 individually read molograms arranged in an 8 × 8 grid, each yielding a separate real-time sensorgram. Individual molograms are named SM_RxCy (Single Mologram at Row x, Column y) — for example, SM_R3C4 is the mologram at row 3, column 4.

How molography measures binding: CMD and MD

Every mologram simultaneously produces two independent signals from the same optical readout. They come from different physical properties of the focal spot — its intensity and its position — and carry fundamentally different information.

The diffractometric detection principle

The focal spot responds differently to different surface events, and this is the physical basis of focal molography's selectivity.

Specific analyte binding adds mass to the ridges but not the grooves. This increases the optical contrast between ridges and grooves: more mass on the ridges means more coherent scattering in register with the grating period. The focal spot becomes brighter — its intensity rises. This intensity change is the diffractometric signal, quantified as Coherent Mass Density (CMD). The principle is set out in Fattinger 2014, and the scattering argument behind it in Frutiger et al. 2019.

Non-specific adsorption deposits molecules randomly across the entire surface — on ridges and grooves equally. In the ideal matched case this does not change the optical contrast between the two regions: the diffracted fields remain balanced, so focal spot intensity does not change. Instead, the uniform refractive-index increase shifts the position of the focal spot laterally. To first order only the positional (refractometric) channel responds; CMD does not.

Bulk refractive-index changes — from temperature fluctuations, buffer exchanges, or high-concentration matrix components such as serum proteins and regeneration reagents — follow the same logic. They affect ridges and grooves similarly, shift the focal spot position, and change its intensity only slightly. CMD therefore strongly suppresses these perturbations without an external correction step — suppressed, not absent: a regeneration pulse still leaves a small CMD excursion, and it is the size of that excursion, not its absence, that says whether the surface came back.

Coherent Mass Density (CMD)

CMD is the primary output of a focal molography measurement. It is derived from the intensity of the focal spot captured by the camera and reported as a surface mass density in units of pg/mm². CMD preferentially reports mass changes that are spatially coherent with the mologram pattern — principally analyte binding to the active ridge ligands, under a properly matched assay design.

A real-time CMD trace — the sensorgram — reflects the three phases of a binding experiment:

• Baseline: running buffer only; CMD is stable because no analyte is present and there is no coherent mass change.
• Association phase: analyte is injected; CMD rises as analyte binds to the immobilized ligand. The rate of rise is governed by both $k_{\mathrm{on}}$ and the analyte concentration.
• Dissociation phase: buffer replaces the analyte solution; CMD decays as analyte leaves the surface, governed by $k_{\mathrm{off}}$.

Because a temperature change reaches ridges and grooves alike, CMD stays comparatively stable across one — far more so than a refractometric signal, which tracks the bulk index directly. In practice this shortens the thermal equilibration a run needs before it can be trusted, and it is what makes kinetics in a complex matrix such as serum practical without a separate correction channel. How much drift remains is assay- and instrument-specific; establish it for your own setup rather than assuming a figure.

Suppression is conditional, not absolute. Chemistry that differs between ridges and grooves, a disturbance that is not spatially uniform — a bubble, a local temperature gradient, contamination that lands in register — or a change in focal-spot quality can all leave a residual CMD response.

Mass Density (MD)

MD is the refractometric channel of the same instrument, derived from the position of the focal spot. When the local refractive index at the sensor surface changes — for any reason — the focal spot shifts laterally. The magnitude of this shift is converted to a surface mass density, also reported in pg/mm².

MD is physically analogous to the signal from a surface plasmon resonance (SPR) instrument: it responds to all refractive-index changes regardless of their origin — specific binding, non-specific adsorption, temperature drift, and buffer composition changes alike. The two channels are not on a common scale: the same surface reads several times higher in MD than in CMD, because MD counts everything present and CMD counts only what follows the pattern.

The sensitivity of MD to bulk effects is also its principal limitation in complex media. A temperature ramp moves MD by an amount that can rival a full protein monolayer, and pairwise sensor subtraction removes only part of it. In a matrix such as diluted serum the matrix components dominate the MD trace, and regeneration pulses produce excursions large enough to mask a specific binding response altogether.

MD is nonetheless useful as a diagnostic channel. Because it captures every refractive-index event at the surface, it can confirm sample delivery, detect air bubbles, identify gross surface fouling, and flag thermal instability — all from the same measurement as CMD.

Sub-micron self-referencing vs. a separate reference channel

In conventional refractometric biosensors such as SPR, non-specific signals are subtracted using a reference spot located millimetres away from the active spot on the same chip. This approach assumes both spots experience identical perturbations — an assumption that breaks down for spatially non-uniform events. Diffusing serum components, localised temperature gradients, and regeneration pulses that arrive at slightly different times at the two spots all produce residual artefacts that can be comparable in magnitude to the binding response itself.

In focal molography, referencing occurs at the 200 nm scale: every ridge is immediately flanked by a groove half a mologram period away (about 200 nm). Both experience identical local conditions — the same evanescent field, the same flow, the same temperature — because they are separated by far less than the diffusion length of any perturbation. Non-specific contributions are suppressed without a spatially separated reference. What survives is whatever the two regions do not share.

CMD and MD at a glance

What “coherent” means here

Rejecting non-specific binding

Rejection is not automatic in a crude matrix: it holds only while matrix proteins actually bind ridges and grooves alike. Where the two chemistries differ, the matrix adsorbs preferentially to one of them and lands in the coherent channel like an analyte would.

Backfilling is the step that keeps them matched — a non-binding probe in the grooves chosen to resemble the ridge chemistry. It is what makes serum, plasma and lysate measurable, and the one preparation step a complex-matrix assay cannot skip. Backfilling and NSB suppression covers how to choose and tune the probe, and how to verify a matched chip.

Mass Transport

Analyte must diffuse from the bulk solution across an unstirred boundary layer before it can bind. When the surface binds faster than diffusion can replenish analyte ($k_t \lesssim k_{\mathrm{on}}\,[\text{ligand}]$), the measurement becomes diffusion-limited and the fitted k_on underestimates the true on-rate. Raising the flow rate and lowering the ligand density push the assay back into the reaction-limited regime.

For the full treatment — symptoms, mitigations, and an interactive simulator — see Mass transport limitation in Biomolecular Interaction Analysis; see also Recommended Flow Parameters and Ligand surface density.

Ligand surface density

The amount of ligand immobilized on the ridges sets the maximum signal $R_{\max}$, but density also strongly shapes the apparent kinetics — so more is not better.

• Too high. A dense surface consumes analyte faster than diffusion can supply it, producing mass-transport limitation and an underestimated $k_{\mathrm{on}}$. Dissociated analyte is also more likely to rebind a neighbouring site before leaving, which flattens the dissociation phase and underestimates $k_{\mathrm{off}}$. For multivalent analytes, high density promotes avidity and steric crowding.
• Too low. The signal approaches the noise floor, giving poor signal-to-noise and unstable fits.

For clean kinetics, use the lowest density that still gives an adequate signal — a low-$R_{\max}$, transport- and rebinding-free regime. On DDI chips, density is tuned by mixing the active capture strand with an inactive one. Shortening the immobilization or stopping on a signal level also lowers it, but on a backfilled chip — where both regions are driven to saturation so their chemistry matches — set the active fraction by the strand mix rather than by under-loading. See Density tradeoffs for the strand-mix recipe, and per-assay target densities.

Limit of detection and sensitivity

The limit of detection is the lowest analyte concentration reliably distinguishable from a blank. What sets that floor is where focal molography departs most sharply from a refractometric sensor.

On a refractometric sensor in a crude matrix, non-specific binding raises and broadens the blank, so the LOD is dominated by NSB rather than by the instrument. Focal molography suppresses this term: as long as affinity-matched backfilling keeps ridge and groove chemistries comparable, coherent NSB stays out of the CMD channel and the blank stays far flatter, even in undiluted serum or plasma. The detection floor is then set by instrumental and photon noise and by mologram-to-mologram reproducibility — not by the sample. How closely the LOD in plasma approaches the buffer result is assay-dependent and must be established from matrix-matched blanks and standards. Frutiger et al. 2019 derives what sets that floor; Reichmuth et al. 2021 measures it in undiluted samples.

Levers that improve the LOD: average more replicate molograms (random noise falls as $1/\sqrt{N}$), extend the association phase to accumulate more bound mass, and run a sufficient number of blanks (≥ 5) to characterize the blank scatter well. The guided LOD workflow is described under Evaluations and the LOD phase template under Assay Phases.

For term definitions, see the Glossary.