New Model Predicts How Well Metasurface Sensors Can Detect Malaria

Malaria still kills roughly half a million people every year, with more than 95 percent of those deaths occurring in Africa, and the diagnostic tools available in low-resource settings often cannot detect infections early enough. A new theoretical study now offers a way to predict, before any experiment is run, how sensitive a plasmonic biosensor could be at spotting one of the disease’s key protein biomarkers. The work, published in the journal Advanced Metamaterials, presents an effective microscopic model that connects the concentration of plasmodium lactate dehydrogenase, or pLDH, in a solution to the optical signal produced by a gold nanohole metasurface sensor.

The model is the work of A. S. Kiyumbi and M. S. Tame, who set out to address a persistent gap in biosensor development. Rapid diagnostic tests used today for malaria typically have a limit of detection for pLDH of around 25 nanograms per millilitre, and submicroscopic infections with parasite densities below 100 parasites per microlitre are not detected with high confidence by immunochromatographic tests. Surface plasmon resonance sensors, which detect molecules by measuring how their binding alters the refractive index of light confined near a metal surface, have long promised faster and more precise label-free detection, but designing them has relied heavily on trial and error or on complex numerical simulations.

The core of the new approach is a combination of three established physical frameworks. Maxwell Garnett effective medium theory describes how a mixture of protein molecules and buffer solution behaves optically as a single effective medium. Langmuir adsorption kinetics describes how proteins bind to and detach from antibody receptors immobilised on the sensor surface until equilibrium is reached. And the physics of surface plasmons, the collective electron oscillations excited at a metal-dielectric interface, determines how changes in refractive index translate into measurable shifts in the transmitted light spectrum. By weaving these together, the authors created a model that takes the bulk concentration of pLDH in a buffer and predicts the sensor’s response without requiring complicated functionalised experiments.

The example sensor the team studied is a metasurface consisting of a 150-nanometre-thick gold film perforated with a periodic array of nanoholes, each 150 nanometres in diameter with a 400-nanometre period. When light shines through such an array, a phenomenon called extraordinary optical transmission occurs: more light passes through than standard aperture theory would predict, because surface plasmons excited at the nanohole edges funnel and concentrate the electromagnetic field. In simulations using the finite element method software COMSOL, the metasurface in phosphate-buffered saline showed two transmission peaks, one near 645 nanometres and another near 715 nanometres, and the researchers demonstrated through selective suppression with thin chrome layers that the lower-wavelength peak arises from surface plasmons at the buffer-gold interface, making it the sensing peak.

Functionalisation is what gives the sensor its power to detect very dilute proteins. Without receptors on the surface, a 250-nanomolar solution of pLDH would change the refractive index of the buffer by only about one ten-thousandth of a refractive index unit, producing a resonance shift of roughly 0.05 nanometres, far too small for standard spectrometers to resolve. When the gold surface is coated with a crosslinker, protein A and IgG antibodies, however, pLDH molecules bind preferentially near the metal surface, creating a thin layer with an elevated effective refractive index exactly where the plasmon field is strongest. The model accounts for this by treating the surface chemistry as stacked adlayers: a 14.5-nanometre functionalisation layer with a refractive index of 1.45, topped by a bound pLDH layer about 7.9 nanometres thick whose effective refractive index depends on how many binding sites are occupied.

That occupancy is governed by the binding affinity between the antibody and the protein, expressed as the equilibrium dissociation constant. Because the precise value for IgG binding to pLDH is not well documented in the literature, the team chose a modest value of 0.25 nanomolar, noting that a value of 0.306 nanomolar has been reported for one antibody interaction, while natural anti-pLDH antibodies might show constants of tens to hundreds of nanomolar. The model also includes a packing parameter that accounts for the fact that proteins cannot fill the entire adsorbate layer; in a well-packed monolayer of spherical proteins this parameter falls around 0.8 to 0.9, and the researchers found that reducing it from 1 to 0.85 increased the predicted limit of detection by roughly 20 percent.

Running the numbers with a realistic white light source spectrum and a transmission noise level derived from it, the model predicted a limit of detection of approximately 0.02 nanomolar of pLDH, equivalent to 0.70 nanograms per millilitre, for both spectral and intensity-based readout. That figure is about 30 times better than current rapid diagnostic tests and, according to the authors’ analysis, would be sufficient to resolve submicroscopic malaria infections with parasite densities around 50 parasites per microlitre, whose pLDH levels are estimated at roughly 1.87 nanograms per millilitre. The simulated metasurface achieved a bulk spectral sensitivity of 237.09 nanometres per refractive index unit and a local sensitivity of 24.58 nanometres per refractive index unit, and the analytical model reproduced the simulated sensitivity to within 0.01 nanometres per refractive index unit.

The researchers are careful to stress that these are theoretical predictions, not experimentally demonstrated results. The simulations model pLDH spiked into phosphate-buffered saline, which the authors note approximates but does not replicate the clinical setting of whole-blood lysate, where biofouling, sensor-to-sensor variability, temperature drift, fluidic instabilities and non-specific binding could all worsen the real-world performance. The single-site Langmuir binding model also neglects mass-transport limitations, rebinding, surface heterogeneity and lateral interactions between bound molecules, and the authors suggest that future refinements could adopt mass-transport-corrected binding models or more sophisticated isotherms. They propose that future experiments could validate the model by functionalising a nanohole array, spiking PBS with pLDH at concentrations from 0.01 to 10 nanomolar, and comparing measured spectral shifts with the model’s predictions.

What the authors emphasise as the real significance of the work is not the low detection limit itself but the predictive framework behind it. Lower limits have been reported elsewhere, including an 18-femtomolar result from a plasmon-enhanced fluorescence immunosensor, but that approach requires specialised low-light imaging equipment that puts it out of reach of most laboratories and clinics. The metasurface sensor modelled here would need only an uncomplicated transmission-based spectroscopy setup, making it more practical and portable for point-of-care use. Because the model requires only a handful of inputs, chiefly the bulk sensitivity of the non-functionalised sensor, which is routinely measured as a first step in sensor characterisation, along with the antibody layer thickness, protein size, buffer refractive index and binding affinity, it can estimate a functionalised sensor’s achievable detection limit without complex functionalised experiments or simulations.

The framework is also designed to be general. By varying the binding constant, layer thicknesses and protein refractive index, the model can be adapted to other malaria biomarkers such as pfHRP-2, to other diseases including coronavirus and cancer biomarkers, and to multiplexed sensors in which different subunits of a metasurface are functionalised with different antibodies to detect several targets in parallel. The authors note that multiplexed detection could eventually help distinguish plasmodium parasite species and help differentiate malaria from other febrile illnesses that are frequently misdiagnosed in high-transmission regions. For now, the model stands as a design and optimisation tool, one that its creators hope will help researchers avoid sub-optimal sensor geometries before committing to the laboratory work of building and testing the next generation of plasmonic diagnostic devices.

Subject of Research: A theoretical model for predicting the performance of plasmonic metasurface biosensors for detecting the malaria biomarker pLDH

Article Title: An effective microscopic model for plasmonic sensing of malaria

Article References: Kiyumbi, A. S., & Tame, M. S. (2025). An effective microscopic model for plasmonic sensing of malaria. Advanced Metamaterials, 1(1), Article 3. https://doi.org/10.1007/s44468-025-00003-z

Image Credits: AI Generated

DOI: 10.1007/s44468-025-00003-z

Keywords: malaria, plasmonic biosensor, metasurface, pLDH, surface plasmon resonance, limit of detection, Maxwell Garnett theory, Langmuir adsorption, nanohole array, diagnostics, gold nanoholes, theoretical model