# Fileset

[Small Methods - 2026 - Watanabe - Single‐Nanoparticle Detection Using Quasi‐Bound States in the Continuum.pdf](https://mdr.nims.go.jp/filesets/521d7e12-3c56-4fa4-84f4-fcfbcc850b1c/download)

## Creator

[Keisuke Watanabe](https://orcid.org/0000-0002-4285-2135), [Samuel Crowther](https://orcid.org/0000-0002-0545-0348), [Masanobu Iwanaga](https://orcid.org/0000-0002-8930-6940), [Frank Vollmer](https://orcid.org/0000-0003-0565-4671), [Tadaaki Nagao](https://orcid.org/0000-0002-6746-2686)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Single‐Nanoparticle Detection Using Quasi‐Bound States in the Continuum Supported by Silicon Metasurfaces](https://mdr.nims.go.jp/datasets/cf2308cb-79cf-4ae8-b9cc-916ea2c8f07b)

## Fulltext

Small Methods www.small-methods.comRESEARCH ARTICLESingle-Nanoparticle Detection Using Quasi-Bound States in the Continuum Supported by Silicon Metasurfaces Keisuke Watanabe1 Samuel Crowther2 Masanobu Iwanaga3 Frank Vollmer2 Tadaaki Nagao1 , 4 1 International Center for Materials Nanoarchitectonics (MANA), National Institute for Materials Science (NIMS), Tsukuba, Ibaraki, Japan 2 Department of Physics and Astronomy, Living Systems Institute, University of Exeter, Exeter, UK 3 Research Center for Electronic and Optical Materials, National Institute for Materials Science (NIMS), Tsukuba, Ibaraki, Japan 4 Department of Condensed Matter Physics, Graduate School of Science, Hokkaido University, Sapporo, Japan Correspondence: Keisuke Watanabe ( watanabe.keisuke@nims.go.jp) Received: 14 March 2026 Revised: 24 June 2026 Accepted: 22 July 2026 Keywords: biosensors | bound states in the continuum | metasurfaces | silicon | single-molecule ABSTRACT The detection of single particles or molecules represents a critical milestone in the development of biosensing technologies. Optical sensors based on quasi-bound states in the continuum (qBICs) have primarily focused on detecting global refractive index changes, whereas detection of local refractive index perturbations, such as the binding of a nanometer-sized molecule on a surface, remains elusive because of limited quality ( Q ) factors and relatively large mode volumes. Here, we demonstrate low-contrast BIC metasurfaces that can perform sensing with a virus-sized single-nanoparticle resolution. The qBIC resonance operating at the critical coupling condition exhibits an experimental Q factor of 4.5 × 104 in heavy water. The strong interactions between the localized electric field and polystyrene nanoparticles with a diameter of 100 nm enable the experimental observation of step-like resonance wavelength shifts, serving as signatures of individual particle binding events. Furthermore, binding-induced modifications to the qBIC resonance alter the optical confinement and asymmetry factor, inducing changes not only in the resonance wavelength but also in the linewidth and amplitude with single-particle sensitivity. Combined with position-insensitive response and free-space accessible features, low-contrast BIC metasurfaces provide a user-friendly platform for next-generation single-nanoparticle sensing integrated with microfluidic systems. 1Q  s  b  e  i  t  f  i  i  m            To©Sh Introduction uasi-bound states in the continuum (qBICs) have gained con-iderable research interest across a wide range of applicationsecause of their high quality ( Q ) factors and strong electric fieldnhancement [ 1 ]. The resulting enhancement of light–matternteractions has emerged as a central topic of interest [ 2–6 ], par-icularly for sensing applications in aqueous environments. Thusar, several label-free qBIC-based sensor geometries have beennvestigated, primarily by exploiting resonance wavelength shiftsnduced by changes in the refractive index of the surroundingedium [ 7–12 ]. The symmetry breaking of symmetry-protectedhis is an open access article under the terms of the Creative Commons Attribution Licenriginal work is properly cited. 2026 The Author(s). Small Methods published by Wiley-VCH GmbH mall Methods , 2026; 10:e70931 ttps://doi.org/10.1002/smtd.70931BICs emerging at the Γ point of the Brillouin zone offers asignificant advantage, namely, sharp resonances can be accessedunder normal-incidence excitation from free-space. Conventional design strategies for these sensors aim to simul-taneously maximize Q factors and refractometric sensitivity forcapturing global, large-area refractive index variations in thesurrounding medium near the surface. In this framework, theconfinement factor of the optical mode into the external mediumΓext , which is directly proportional to the global refractive indexsensitivity, is a critical parameter. However, increasing Γext typi-cally degrades the Q factor. Recent experimental approaches havese, which permits use, distribution and reproduction in any medium, provided the 1 of 9http://www.small-methods.comhttps://doi.org/10.1002/smtd.70931https://orcid.org/0000-0002-4285-2135https://orcid.org/0000-0002-0545-0348https://orcid.org/0000-0002-8930-6940https://orcid.org/0000-0003-0565-4671https://orcid.org/0000-0002-6746-2686mailto:watanabe.keisuke@nims.go.jphttp://creativecommons.org/licenses/by/4.0/https://doi.org/10.1002/smtd.70931http://crossmark.crossref.org/dialog/?doi=10.1002%2Fsmtd.70931&domain=pdf&date_stamp=2026-08-04a  p  [  i  mT  m  t  v  v  c  b  p  s  p  i  o  m  c  h  m  e  t  f  u  f  d  d  d  Y  u  e  H  e  a  h  t  Q  T  Q  l  l  v  pI  s  s  s  t  d  s  W  n  w  i  v  r  s  f                                                      2 23669608, 2026, 17, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smtd.70931 by National Institute For, Wiley Online Library on [10/09/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creatiddressed this trade-off using nanogap geometries [ 13 ], hybridlasmonic–dielectric metasurfaces [ 14 ], and Bloch surface waves 15 ]. On the other hand, sensors that can resolve local refractivendex perturbations, such as the binding of nanometer-sizedolecules on a surface [ 16 ], remain elusive. he fractional frequency shift induced by a nanometer-sizedolecule follows Δω/ ω ∝ − αex / Veff , where αex and Veff representhe excess polarizability of the molecule and effective modeolume [ 17, 18 ], respectively. Therefore, reducing the modeolume is essential for achieving large resonance shifts. Whenombined with high-resolution wavelength tracking affordedy high Q factors, optical resonators with high Q / V ratios areromising platforms for sensors with single-nanoparticle or eveningle-molecule resolution. Over the past decade, various single-article/molecule optical sensors have been developed [ 19–23 ],ncluding whispering gallery mode (WGM) resonators basedn microspheres [ 24 ], microtoroids [ 25 ], microrings [ 26 ], andicrolasers [ 27 ] exhibiting ultrahigh Q factors. Further, photonicrystal cavities confining light to diffraction-limited volumesave enabled the detection of single viruses [ 28 ] and singleolecules [ 29 ]. In comparison, conventional BIC metasurfacesxhibit lower Q factors because qBIC resonances are susceptibleo fabrication imperfections, and the scattering losses arisingrom the surface roughness restrict the potential to achieveltrahigh Q factors. This limitation prevents BIC metasurfacesrom reaching the signal-to-noise required for virus-sized particleetection. Indeed, most BIC-based sensing studies reported toate have detected global refractive index changes induced byense adsorption of biomolecules on the sensor surface [ 7, 8 ].ang et al. [ 9 ] and Li et al. [ 11 ] investigated single-particle bindingsing simulations based on periodic boundary conditions, whichffectively corresponds to an infinite array of identical particles.owever, this configuration differs substantially from practicalxperiments, where a single particle binds to a finite metasurfacerray. Therefore, experimental detection of virus-sized particlesas not been demonstrated in BIC nanostructures. To addresshis challenge, we recently demonstrated a BIC metasurface with factors exceeding 100,000 using a low-contrast geometry [ 30 ].his ultrahigh Q factor arises from the intrinsically high radiative factors of higher-order qBICs combined with reduced scatteringosses in shallow-etched geometries. These features suggest thatow-contrast BIC metasurfaces can provide reasonably high Q / Values, making them highly sensitive to local refractive indexerturbations. n this study, we demonstrate the real-time detection of virus-ized single nanoparticles in water using low-contrast BIC meta-urfaces with ultrahigh Q factors. We fabricate shallow-etchedilicon metasurfaces with an asymmetry parameter satisfyinghe critical coupling condition, enabling the observation ofiscrete, step-like resonance wavelength shifts as signatures of theurface adsorption events of individual polystyrene (PS) particles.e analyze the overlap between localized electric fields andanoparticles, revealing the largest wavelength shifts for particlesith a 100 nm diameter. Furthermore, we observe binding-nduced changes in linewidth and resonance amplitude. Theseariations can be attributed to the highly sensitive nature of qBICesonances to symmetry perturbation, representing a uniqueensing mechanism enabled by our dielectric BIC metasurfacesor virus-sized particle detection. of 92 Results and Discussion 2.1 Structural and Sensing Characteristics The low-contrast BIC metasurfaces were fabricated on a silicon-on-insulator (SOI) wafer with a 400-nm-thick top silicon layer.Shallow pair-rod nanostructure arrays were formed by dry etch-ing only approximately 52 nm of the top silicon layer with aperiod P of 760 nm. Figure 1a illustrates PS particles dispersedaround a low-contrast BIC metasurface. Tunable laser light isfocused onto the metasurface, and the transmission spectrum isrecorded via a photodiode. When a single nanoparticle lands onthe surface, it interacts with the localized electric field, induc-ing a resonance wavelength shift (Figure 1b ). By continuouslytracking the resonance peak, adsorption and desorption eventscan be observed as discrete, step-like changes in the signal. Thefabricated metasurface is integrated into a polydimethylsiloxane(PDMS) microfluidic channel (Figure 1c ), and the transmissionspectrum is monitored while injecting a PS particle dispersion.Figure 1d shows a scanning electron microscope (SEM) image ofthe fabricated BIC metasurface, wherein each unit cell includes apair of rods with lengths L + 2 ΔL and L − 2 ΔL . The designed rodlength L and width are 640 nm and L − P /2 = 260 nm, respectively.Introducing this length asymmetry transforms the symmetry-protected BIC into a qBIC with an asymmetry parameter α =2 ΔL / L , enabling sharp qBIC resonances to be experimentallyaccessed under normal-incidence excitation. Figure 2a details the custom-built experimental setup. A tunablelaser (TSL210V, santec) with a tuning range of 1500 to 1600 nmwas focused onto the sample through a 10 × objective lens (MPlan Apo NIR, NA = 0.26, Mitutoyo) and a polarizer. The focusedlaser beam diameter was evaluated to be 28.4 μm using the knife-edge method (Supporting Information S1 ). The transmitted beamwas detected using an InGaAs picowatt receiver (WPR-2K-IN-FST, FEMTO), with the background noise minimized using across-polarized configuration implemented with an additionalpolarizer placed in front of the photoreceiver. Figure 2b shows thetransmission spectrum for a metasurface with α = 4% measuredusing the electromechanical wavelength scanning of the tunablelaser. High-resolution wavelength tuning was achieved using thebuilt-in piezoelectric control of the laser wavelength by applyinga 10 Hz triangular voltage generated by a function generator(AFG1022, Tektronix). The resulting resonance spectrum syn-chronized with the triangular modulation was recorded using adata acquisition (DAQ) device (USB-6212, National Instruments)controlled through a LabVIEW interface. The acquired spectrum(inset of Figure 2b ) was fitted with a Fano function expressedas F ( δ) = A0 + F0 ( q + δ)2 /(1 + δ2 ), where A0 and F0 repre-sent fitting constants, and q represents the Fano asymmetryparameter. Here, δ = ( ω − ω0 )/ γ, where ω0 and γ represent theresonance frequency and damping rate, respectively. The fittingyielded an experimental Q factor ( = ω0 /2 γ) of approximately 4.5× 104 . All the measurements were performed in heavy water(D2 O) to suppress unwanted overtone absorption from water(Supporting Information S2 ). The laser wavelength for piezoscanning was calibrated using an optical spectrum analyzer(AQ6374E, Yokogawa), resulting in a slight offset in the displayedwavelengths (horizontal axis) between motor-driven and piezoscanning. Figure 2c shows the Q -factor analysis performed bymeasuring experimental Q factors for the different values of αSmall Methods, 2026ve Commons LicenseFIGURE 1 Low-contrast BIC metasurfaces for single-nanoparticle detection. (a) Schematic of single PS nanoparticle detection using a low-contrast BIC metasurface fabricated on an SOI wafer. (b) Sensing mechanism. The resonance wavelength shift (top) and its temporal evolution (bottom) displaying step-like signal changes. (c) PDMS microfluidic channel integrated on a metasurface array. (d) SEM image of a fabricated BIC metasurface with an etching depth of approximately 52 nm and asymmetry parameter α = 4%. The dashed region indicates a unit cell. (e) Structural parameters of a single unit cell. (  a  Q  Q  t  s  Q  f  c  Q  u  Q  d  l  p  a  m  f  c  α cN  D  c  r                       S 23669608, 2026, 17, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smtd.70931 by National Institute For, Wiley Online Library on [10/09/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable CreatiSupporting information S3 for the transmission spectra for all α)nd comparing them with theoretical fitting curves. The radiative factors of qBIC resonances, Qr , which follow the relationshipr = Q0 α− 2 , were calculated at different α using finite-differenceime-domain (FDTD) simulations, yielding Q0 = 128.9. For amall α, the experimental Q factors deviate significantly fromr . This can be attributed to scattering losses arising fromabrication imperfections. These losses introduce an additionalontribution Qscat − 1 , limiting the total Q factor as Q− 1 = Qr − 1 +scat − 1 . The experimental Q factors fitted well with this equationsing nonlinear least squares curve-fitting method, resulting inscat = 7.8 × 104 . Recently, we theoretically and experimentallyemonstrated that optimal balance between a high Q factor andarge resonance amplitude can be achieved under a critical cou-ling condition, where the radiative and nonradiative Q factorsre equal. Under this condition, wavelength fluctuations areinimized, resulting in the lowest achievable limit of detectionor refractive index sensing [ 31 ]. In the present study, the criticaloupling condition was experimentally determined to occur atCC = 4%, corresponding to the intersection of the Qr and Qscaturves in Figure 2c . ext, unfunctionalized PS particles ( d = 500 nm) dispersed in2 O with 100 mM NaCl were injected into the PDMS microfluidichannel, and the resonance peak wavelength was monitored ineal-time. As shown in Figure 2d , multiple discrete wavelengthmall Methods, 2026steps were observed. Following the initial step-like shift, theresonance wavelength exhibits up-and-down oscillations; thisbehavior is attributed to the particle transiently bouncing near thesurface because of weak trapping under the thermal fluctuations.The subsequent negative step suggests the particle moved awayfrom the surface. A distinct step with reduced noise was thenobserved around 130 s, suggesting the stable adsorption of theparticle onto the surface. The observation of both weak trappingand irreversible binding within a single experiment suggests thepresence of electrostatic repulsion between the anionic sulfonate(R-SO3 − ) groups on the PS particle [ 32 ] and negatively chargednative oxide (SiO2 ) surface in water, even though this repulsiveinteraction is partially screened by reducing the Debye lengththrough the addition of salt (100 mM NaCl) [ 33 ]. 2.2 Size Dependence We conducted FDTD simulations of the electric field enhance-ment under a normally incident x -polarized plane-wave to clarifysensing characteristics associated with PS particles binding tolow-contrast BIC metasurfaces (Figure 3a ). As indicated by theelectric field distribution in the yz -plane, two antinodes appearalong the z -direction. This corresponds to a higher-order mode(TE122 ) qBIC strongly confined within the top silicon layerand exhibits a high radiative Q factor. In the xy -plane at the3 of 9ve Commons Licenseα = 4%α = 4% (a.u.)λ (nm)1592.78 1592.9642~4.5×104Fanob ca01−1234501234567BSObj.Obj.CollimatorTriggerSignalPol.Tunable laserLensPDCameraTubeLensSampleMirrorMirrorPol. 5×10×DAQ PCFunctiongenerator dTime (s)0 50 100 150Δλ  (pm)Wavelength λ (nm)Trapping Binding1592 1593 1594 1595 1596Transmission  (a.u.)104105106 factorsscatr = 0α−2Exp. 1 10α (%)αCCFIGURE 2 Optical characterization and single-particle sensing performance. (a) Schematic of high-resolution transmission spectral measurement setup. (b) Representative transmission spectrum measured in D2 O for a fabricated BIC metasurface with asymmetry parameter α= 4%. Inset shows Fano fitting obtained under fine wavelength tuning based on piezoelectric laser control. (c) Q -factor analysis showing experimental Q factors (open circles) fitted as total Q factors (black dashed line) with simulated radiative Q factor (blue solid line) and scattering Q factor (black solid line). Error bars represent the standard deviation of the experimental Q factors obtained from four different metasurfaces ( N = 4). (d) Real-time resonance wavelength shift measured in a dispersion of unfunctionalized PS particles ( d = 500 nm) showing both weak trapping and discrete step-like binding events. FIGURE 3 Electric field enhancement and interaction mechanisms with nanoparticles. (a) Cross-sectional schematic of the yz -plane (lower left) and simulated electric field intensity enhancement | E2 | at x = 0 nm (right) with the field distribution in the xy -plane at a shallow-etched surface (upper left). (b) Maximum field intensity profile along the z -direction. (c) Schematic of two adjacent unit cells in the xy -plane and classification of nanoparticle binding configurations based on particle diameter d and shallow-etched rod spacings l1 and l2 . 4 of 9 Small Methods, 2026 23669608, 2026, 17, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smtd.70931 by National Institute For, Wiley Online Library on [10/09/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons LicenseFIGURE 4 Size dependence of PS particle binding. (a) Real-time sensing results in the step-dominated region for different PS particle diameters d . Both Fano-fitted and raw spectral peaks are shown. (b) Average step height as a function of d . Error bars represent the standard deviation of the wavelength step heights. The number of detected step events was N = 42, 4, 13, and 8 for d of 100, 200, 500, and 1000 nm, respectively. (c) Histogram of wavelength step heights for d = 100 nm. Red solid and dashed vertical lines indicate the mean step height and the background wavelength fluctuations 3 σ, respectively. s  o  (  t  w  l  p  m  d  d  l  (  r  w  l  i  r  oW  t  w  a  s  s  o  F  t                             S 23669608, 2026, 17, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smtd.70931 by National Institute For, Wiley Online Library on [10/09/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creatihallow-etched surface, the electric field is localized at the edgesf both rods, which are exposed directly to the external mediumFigure 3b ). The localized fields then decay exponentially alonghe z -direction from the silicon surface into the surroundingater (Supporting Information S4 for more details). Such fieldocalization at the interface while preserving a high Q factorresents a key advantage for sensing applications that targetolecular binding events. Based on these results, we classify twoistinct nanoparticle binding configurations based on particleiameter d and gap spacings between adjacent rods, denoted as1 (upper rods) and l2 (lower rods), across neighboring unit cellsFigure 3c ). When d < l1 , l2 , nanoparticles can access the gapegion at the shallow-etched surface and cause strong interactionsith the localized electric field (configuration (i)). When d > l1 ,2 , nanoparticles are excluded from gap regions, which resultsn weaker field interactions (configuration (iii)). In both cases,esonance wavelength shifts can also arise from particle bindingn the top surfaces of the rods (configurations (ii) and (iv)). e compared the peak wavelength shifts for different diame-ers of PS particles, as shown in Figure 4a . Resonance peakavelengths were extracted in real-time from both raw datand simultaneous Fano fitting. When d > l1 ( ∼ 125 nm), thetep height increased with increasing d ; in contrast, the largesttep height was observed at d = 100 nm, where d < l1 . Thisbservation was confirmed by the average step heights shown inigure 4b (Supporting Information S5 for multiple-comparisonests). As discussed above, this behavior can be attributed tomall Methods, 2026the strong interaction between the localized electric field of theqBIC resonance and a single PS particle positioned betweenthe shallow-etched nanorods. Although the Fano and raw peaksexhibited consistent trends, in some cases, only the raw peakshowed a discrete step shift during the binding of small nanopar-ticles, such as those with d = 100 nm. This can be attributedto changes in the Fano fitting parameters induced by localrefractive index perturbations being too small to appear as ameasurable resonance shift in real-time Fano fitting. Althoughthe wavelength fluctuation 3 σ of the raw peak (0.50 pm) waslarger than that of the Fano peak (0.18 pm), local refractiveindex changes were readily identified as discrete steps in theraw spectra. Figure 4c presents a histogram of the step heightsextracted from the raw peaks for d = 100 nm, yielding anaverage step height of 2.78 pm. The wavelength shift distributionfollowed a single-peak normal distribution; this is consistentwith the typical statistics of single-particle binding events [ 34 ].The 3 σ background wavelength fluctuations were well belowthe observed individual step heights, corresponding to a signal-to-noise ratio (SNR = Δλ/ σ) of 16.5. We also confirmed thatthe time interval Δt between successive binding events of PSparticles follows a single-exponential decay, which is consistentwith a Poissonian stochastic adsorption process [ 18 ] of individualbinding events (Supporting Information S6 ). The linear scalingof the event rate with particle concentration provides additionalevidence for individual particle-binding events rather than instru-mental artifacts or random baseline fluctuations (SupportingInformation S7 ). 5 of 9ve Commons LicenseFIGURE 5 Binding-induced changes in qBIC properties. (a) Measured transmission spectrum and corresponding Fano fitting. (b) Real-time detection of single PS particles ( d = 500 nm) monitored via peak wavelength λp , linewidth κ, and peak amplitude A (top to bottom). Red dashed lines indicate particle-binding events. (c) Histograms of step changes in peak wavelength Δλp , linewidth Δκ, and amplitude ΔA (top to bottom). C  t  i  m  p  t  a  B2W  o  i  a  b  r  a  w  l  l  s  a  b  t  t  a  i  i  i  s  (  d                                  6 23669608, 2026, 17, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smtd.70931 by National Institute For, Wiley Online Library on [10/09/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creatiompared with optical cavity sensors reported in the literature,he wavelength steps induced by single-nanoparticle bindingn our BIC metasurface fall between those reported for WGMicrospheres [ 16 ] and those obtained using microring [ 26 ] andhotonic crystal cavities [ 35 ]. This trend is qualitatively consis-ent with the corresponding mode volumes of these platforms,lthough rigorous quantification of the effective mode volume inIC metasurfaces warrants further investigation. .3 Binding-Induced Changes in qBIC Properties e investigated the effect of local refractive index perturbationsn the resonance properties under qBIC conditions. As shownn Figure 5a , we analyzed how the peak wavelength, linewidth,nd amplitude of the Fano resonance evolve upon single-particleinding. Figure 5b,c show the real-time tracking results and cor-esponding histograms of the step heights, respectively, obtainedfter injecting a dispersion of unfunctionalized PS particles. Here,e selected a particle diameter of 500 nm to ensure sufficientlyarge signal changes and to reliably extract the step height,inewidth, and amplitude through real-time Fano fitting. Theolution pH was adjusted to 3.6 by adding HCl while maintaining NaCl concentration of 100 mM to reduce electrostatic repulsionetween the surface and the particles. Consequently, no transientrapping events were observed in peak wavelength, while mul-iple discrete steps were detected, all corresponding to redshifts,s shown in the histogram. Note that the gradual baseline drifts likely caused by thermally induced refractive index changesn the solution (Supporting Information S8 ). The linewidth stepsncreased concurrently with the wavelength steps. The amplitudeteps, on the other hand, exhibited both increase and decreasethe average step change ΔA was − 0.013, indicating a slightecrease). Yang et al. recently demonstrated that variations inof 9permittivity asymmetry can be encoded into the asymmetryfactor of the system without modifying the geometric asymmetry[ 36 ]. Consequently, both the resonance position and amplitudebecome sensitive to changes in the surrounding refractive index.A similar mechanism applies in our case; the binding of singlePS particles can modify the asymmetry factor of the qBICresonance depending on the binding position of the particle.Specifically, when a particle binds at a location that increases(decreases) the effective asymmetry, the radiative coupling tothe external continuum is enhanced (suppressed), leading toan increase (decrease) in the resonance amplitude. In thesescenarios, the linewidth is expected to increase or decrease;however, only positive linewidth steps were observed in ourexperiments. This behavior may be related to weakened opticalconfinement or enhanced scattering induced by the binding ofrelatively high refractive index ( n = 1.59) PS particles. Anothernotable observation is that, in some cases, no amplitude changewas detected even when a wavelength step was observed. Thisbehavior can be attributed to particle binding at locations wherethe change in effective asymmetry is insignificant. These resultsconfirm that low-contrast BIC metasurfaces respond to localrefractive index perturbations through simultaneous changesin three resonance parameters—wavelength, linewidth, andamplitude—all of which exhibit single-particle sensitivity with ahigh signal-to-noise ratio. Accumulating large datasets of suchmeasurements and correlating them with physical parameterscan help extract physical information, including particle size,position, and refractive index, utilizing artificial intelligence. Beyond the sensing performance, silicon-based BIC metasurfacesoffer an additional advantage in their compatibility with est-ablished complementary metal–oxide–semiconductor (CMOS)technologies [ 37 ], enabling low-cost, precision-controlled massproduction. Although this advantage is common to previouslySmall Methods, 2026ve Commons Licenser  a  m  f  i  a  c  v  wA  t  o  i  s  N  m  [  [  c  o  [  [3W  t  w  g  a  l  t  l  i  T  o  e  s  c  a  d  w  w  f  i  p  a  p  f  a  e  a44T  w                                                S 23669608, 2026, 17, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smtd.70931 by National Institute For, Wiley Online Library on [10/09/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creatieported silicon photonic crystals [ 35 ], spiral waveguides [ 38 ],nd silicon nitride microring resonators [ 39 ], our silicon BICetasurfaces enable single-nanoparticle detection using a simpleree-space configuration in which normally incident laser lights transmitted through the device and measured directly. Thispproach eliminates the need for unstable fiber-based opticaloupling systems [ 40 ] and fluorescence labeling [ 23 ], thereby pro-iding a more user-friendly sensing platform readily integratedith microfluidic systems [ 41 ]. t the current stage, however, direct detection of individual pro-eins, which possess substantially smaller excess polarizabilitieswing to their smaller physical volumes, remains challeng-ng. Moreover, the present measurements rely on probabilisticingle-event detection at relatively high particle concentrations.evertheless, detection of trace amounts of molecules or proteinsay be enabled through appropriate surface functionalization 42 ], such as aptamers [ 43 ] and by exploiting trapping potentials 44 ]. Further enhancement of the local electric field intensityan contribute to both signal enhancement and realizationf trapping-assisted sensing based on self-induced back-action 9, 19 ] by introducing nanogaps in paired shallow nanorods 45, 46 ].  Conclusion e demonstrated, for the first time in BIC nanostructures,he experimental detection of virus-sized single PS particles inater by leveraging the ultrahigh- Q nature of a low-contrasteometry. This sensing capability arises from the strong inter-ction between the confined electric fields and nanoparticlesocated within the shallow-etched structures. We further revealedhat the resonance properties of BIC metasurfaces—w avelength,inewidth, and amplitude—were modified by the binding ofndividual nanoparticles, all exhibiting single-particle sensitivity.his ultrahigh sensitivity can be understood as a consequencef binding-induced changes in the optical confinement andffective asymmetry of the qBIC system, representing a uniqueensing feature enabled by BIC metasurfaces. Because of theollective resonant behavior of relatively large-area periodicrrays, BIC metasurfaces are inherently position-insensitive ando not require precise spatial alignment. This feature, togetherith the compatibility with microfluidic integration, makes themell-suited for future high-throughput sensor arrays, whereabrication-induced variations can be compensated throughndividual calibration of each sensing element. In addition, sim-le normal-incidence excitation from free-space provides directccess to sharp qBIC resonances without complex optical cou-ling schemes. These advantages offer promising opportunitiesor simple, cost-effective, and highly sensitive detection of virionsnd other pathogenic nanoparticles. This platform could furthernable real-time investigations of nanoscale binding dynamicsnd label-free biosensing at the ultimate single-entity limit.  Experimental Section .1 Fabrication he low-contrast BIC metasurfaces were fabricated on a SOIafer consisting of a 400-nm-thick top silicon layer and a 2000-mall Methods, 2026nm-thick buried oxide (BOX) layer. A 100-nm-thick resist AR-P6200 was first spin-coated onto the wafer, followed by electron-beam lithography (ELS-BODEN, Elionix) to define metasurfacepatterns (100 x 100 μm). The patterns were transferred intothe top silicon layer using a Bosch process (MUC-21 ASE-SRE,Sumitomo Precision Products) with SF6 and C4 H8 gases. Theetching depth was controlled by adjusting the number of etchingand passivation cycles to achieve the shallow-etched geometry.The resulting etching depth was measured using spectroscopicellipsometry (M-2000, J.A. Woollam) and was approximately52 nm. 4.2 Characterization Unfunctionalized PS particles (Polysciences/Bangs Laboratories)were diluted in heavy water (D2 O, 151882, Sigma-Aldrich) con-taining 100 mM NaCl. The particle concentration was adjustedto 50 μg/mL for d = 100 nm and 100 μg/mL for d = 200, 500,and 1000 nm. The solutions were introduced into a custom-designed PDMS microfluidic channel (Micro TAS Engineering)via a silicone tube with an inner diameter of 1 mm. After sampleintroduction, all measurements were performed under identicalquiescent conditions without external flow. The temperature andilluminated area were kept constant throughout the measure-ments. The output power of the tunable laser (TSL210V, santec)was set to 30–40 μW. The laser linewidth was intentionallybroadened to 500 MHz to suppress interference noise duringtransmission spectrum acquisition. Signal steps of real-timemeasurement results were identified using the window-basedchange point detection algorithm in the ruptures Python library[ 47 ]. The wavelength fluctuation was defined as three times thestandard deviation calculated from 20 consecutive measurementpoints. 4.3 Simulation Electromagnetic field simulations were performed using theFDTD method (Ansys Lumerical). A single-unit-cell model wasconstructed with perfectly matched layer boundary conditionsalong the z -axis and Bloch boundary conditions along the x -and y - axes. The optical constants of silicon and the BOX(SiO2 ) layer were taken from Palik [ 48 ]. Multiple electric dipolesources were randomly positioned at the mid-height of the topsilicon layer ( z = 200 nm) to excite transverse electric (TE)-likemodes. Subsequently, the radiative Q factor Qr was calculated asω0 U ( t )/ P ( t ) [ 49 ], where ω0 is the resonance angular frequency,U ( t ) is the stored electromagnetic energy, and P ( t ) is the radiatedpower absorbed at the simulation boundaries. Spatial electricfield profiles at the qBIC resonance were obtained using athree-dimensional field monitor under x -polarized plane-waveexcitation. Acknowledgements This work was financially supported by JSPS KAKENHI (Grant Nos.JP22K20496 and JP24K17583), Iketani Science and Technology Founda-tion (Grant No. 0361252-A), and “Advanced Research Infrastructure forMaterials and Nanotechnology in Japan (ARIM)” of the Ministry of7 of 9ve Commons LicenseE  N  SCTDT  cR C  C  h2  L  C3  C  l  34  G  S  P5  F  16  U  f7  E  C8  L  D  39  i1  t  M1  V  S  P  a1  BA  a1  R  D  o1  ED  1                                  T         8 23669608, 2026, 17, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smtd.70931 by National Institute For, Wiley Online Library on [10/09/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creatiducation, Culture, Sports, Science and Technology (MEXT). (Proposalumber JPMXP1224NM5259, JPMXP1225NM5235). The authors thankatoshi Ishii for his support with an experimental equipment. onflicts of Interest he authors declare no competing interests. ata Availability Statement he data that support the findings of this study are available from theorresponding author upon reasonable request. eferences 1 . K. Watanabe, H. R. Devi, M. Iwanaga, and T. Nagao, “Vibrationaloupling to Quasi-Bound States in the Continuum Under Tailoredoupling Conditions,” Advanced Optical Materials 12 (2024): 2301912,ttps://doi.org/10.1002/adom.202301912 .  . E. Maggiolini, L. Polimeno, F. Todisco, et al., “Strongly Enhancedight–Matter Coupling of Monolayer WS2 From a Bound State in theontinuum,” Nature Materials 22 (2023): 964–969.  . L. Yuan, Y. Zhao, A. Toma, et al., “A Quasi-Bound States in theontinuum Dielectric Metasurface-Based Antenna–Reactor Photocata-yst,”Nano Letters 24 (2024): 172–179, https://doi.org/10.1021/acs.nanolett.c03585 .  . J. L. Pura, B. Castillo López de Larrinzar, M. Liang, A. García-Martín, J.. Rivas, and J. A. Sánchez-Gil, “Superchiral Light Emerging From Boundtates in the Continuum in Metasurfaces of Si Nanorod Dimers,” ACShotonics 11 (2024): 4100.  . J. Cui, S. Han, B. Zhu, et al., “Ultracompact Multibound-State-Assistedlat-Band Lasers,” Nature Photonics 19 (2025): 643–649, https://doi.org/0.1038/s41566- 025- 01665- 6 .  . L. Sortino, J. Biechteler, L. Lafeta, et al., “Atomic-Layer Assembly ofltrathin Optical Cavities in van der Waals Heterostructure Metasur-aces,” Nature Photonics 19 (2025): 1–8.  . A. Ndao, L. Hsu, W. Cai, et al., “Differentiating and Quantifyingxosome Secretion From a Single Cell Using Quasi-Bound States in theontinuum,” Nanophotonics 9 (2020): 1081–1086.  . Y. Jahani, E. R. Arvelo, F. Yesilkoy, et al., “Imaging-Based Spectrometer-ess Optofluidic Biosensors Based on Dielectric Metasurfaces foretecting Extracellular Vesicles,” Nature Communications 12 (2021):246.  . S. Yang, C. Hong, Y. Jiang, and J. C. Ndukaife, “Nanoparticle Trappingn a Quasi-BIC System,” ACS Photonics 8 (2021): 1961–1971. 0 . Y. Zhou, Z. Guo, X. Zhao, et al., “Dual-Quasi Bound States inhe Continuum Enabled Plasmonic Metasurfaces,” Advanced Opticalaterials 10 (2022): 2200965, https://doi.org/10.1002/adom.202200965 . 1 . Z. Li, M. Xie, G. Nie, J. Wang, and L. Huang, “Pushing Opticalirus Detection to a Single Particle Through a High-Q Quasi-Boundtate in the Continuum in an All-Dielectric Metasurface,” Journal ofhysical Chemistry Letters 14 (2023): 10762–10768, https://doi.org/10.1021/cs.jpclett.3c02763 . 2 . H. Li, Z. Shi, H. Zhang, S. Qiu, and Z. K. Zhou, “Hybrid Metasurfaceased on Si3 N4 Nanopillar for Optical Sensing With Dual Channels,”CS Applied Nano Materials 8 (2025): 2965–2973, https://doi.org/10.1021/csanm.4c06648 . 3 . K. Watanabe and M. Iwanaga, “Nanogap Enhancement of theefractometric Sensitivity at Quasi-Bound States in the Continuum in All-ielectric Metasurfaces,” Nanophotonics 12 (2023): 99–109, https://doi.rg/10.1515/nanoph-2022-0565 . 4 . M. Luo, Y. Zhou, X. Zhao, et al., “High-Sensitivity Optical Sensorsmpowered by Quasi-Bound States in the Continuum in a Hybrid Metal–ielectric Metasurface,” ACS Nano 18 (2024): 6477–6486, https://doi.org/0.1021/acsnano.3c11994 . of 915 . B. W. Lin, M. H. Chen, Y. C. Lee, et al., “Quasi-Bound States inthe Continuum on the Bloch-Surface-Wave Platform for Sensing withSimultaneous High Q-Factor and High Sensitivity,” Advanced OpticalMaterials 14 (2026): 03259, https://doi.org/10.1002/adom.202503259 . 16 . F. Vollmer, S. Arnold, and D. Keng, “Single Virus Detection Fromthe Reactive Shift of a Whispering-Gallery Mode,” Proceedings of theNational Academy of Sciences 105 (2008): 20701–20704, https://doi.org/10.1073/pnas.0808988106 . 17 . L. Kohler, M. Mader, C. Kern, M. Wegener, and D. Hunger, “TrackingBrownian Motion in Three Dimensions and Characterization of Individ-ual Nanoparticles Using a Fiber-Based High-Finesse Microcavity,”NatureCommunications 12 (2021): 6385. 18 . D. Yu, M. Humar, K. Meserve, R. C. Bailey, S. N. Chormaic, and F.Vollmer, “Whispering-Gallery-Mode Sensors for Biological and PhysicalSensing,” Nature Reviews Methods Primers 1 (2021): 83, https://doi.org/10.1038/s43586- 021- 00079- 2 . 19 . M. L. Juan, R. Gordon, Y. Pang, F. Eftekhari, and R. Quidant,“Self-Induced Back-Action Optical Trapping of Dielectric Nanoparticles,”Nature Physics 5 (2009): 915–919, https://doi.org/10.1038/nphys1422 . 20 . W. Chen, Ş. K. Özdemir, G. Zhao, J. Wiersig, and L. Yang, “ExceptionalPoints Enhance Sensing in an Optical Microcavity,” Nature 548 (2017):192–196, https://doi.org/10.1038/nature23281 . 21 . M. Iwanaga, T. Hironaka, N. Ikeda, T. Sugasawa, and K. Takekoshi,“Metasurface Biosensors Enabling Single-Molecule Sensing of Cell-FreeDNA,” Nano Letters 23 (2023): 5755–5761, https://doi.org/10.1021/acs.nanolett.3c01527 . 22 . D. Ludescher, L. Wesemann, J. Schwab, et al., “Optical Sieve forNanoplastic Detection, Sizing and Counting,” Nature Photonics 19 (2025):1138–1145, https://doi.org/10.1038/s41566-025-01733-x. 23 . M. Iwanaga, “Single-MicroRNA Detection on High-Selectivity Meta-surface Fluorescence Biosensors,” ACS Nano 19 (2025): 38841–38848,https://doi.org/10.1021/acsnano.5c15853 . 24 . M. D. Baaske, M. R. Foreman, and F. Vollmer, “Single-moleculeNucleic Acid Interactions Monitored on a Label-Free Microcavity Biosen-sor Platform,” Nature Nanotechnology 9 (2014): 933–939, https://doi.org/10.1038/nnano.2014.180 . 25 . T. Lu, H. Lee, T. Chen, et al., “High Sensitivity Nanoparticle DetectionUsing Optical Microcavities,” Proceedings of the National Academy ofSciences 108 (2011): 5976–5979, https://doi.org/10.1073/pnas.1017962108 . 26 . S. Lin and K. B. Crozier, “Trapping-Assisted Sensing of Particles andProteins Using on-Chip Optical Microcavities,” ACS Nano 7 (2013): 1725–1730, https://doi.org/10.1021/nn305826j . 27 . L. He, Ş. K. Özdemir, J. Zhu, W. Kim, and L. Yang, “DetectingSingle Viruses and Nanoparticles Using Whispering Gallery Microlasers,”Nature Nanotechnology 6 (2011): 428–432, https://doi.org/10.1038/nnano.2011.99 . 28 . K. Watanabe, H. Y. Wu, J. Xavier, L. T. Joshi, and F. Vollmer, “SingleVirus Detection on Silicon Photonic Crystal Random Cavities,” Small 18(2022): 2107597. 29 . M. Mossayebi, A. Parini, A. J. Wright, M. G. Somekh, G. Bellanca,and E. C. Larkins, “Hybrid Photonic-Plasmonic Platform for High-hroughput Single-Molecule Studies,” Optical Materials Express 9 (2019):2511, https://doi.org/10.1364/OME.9.002511 . 30 . K. Watanabe, T. Nagao, and M. Iwanaga, “Low-Contrast BIC Metasur-faces With Quality Factors Exceeding 100,” Nano Letters 25: 2777–2784. 31 . K. Watanabe and M. Iwanaga, “Optimum Asymmetry for Nanofab-ricated Refractometric Sensors at Quasi-Bound States in the Contin-uum,” Applied Physics Letters 124 (2024): 111705, https://doi.org/10.1063/5.0158793 . 32 . J. R. Lopez, E. Treasurer, K. M. Snyder, D. Keng, and S. Arnold,“Whispering Gallery Mode Coulometry of the Nanoparticle-MicrocavityInteraction in Aqueous Solution,” Applied Physics Letters 112 (2018):051109, https://doi.org/10.1063/1.5017041 . Small Methods, 2026ve Commons Licensehttps://doi.org/10.1002/adom.202301912https://doi.org/10.1021/acs.nanolett.3c03585https://doi.org/10.1038/s41566-025-01665-6https://doi.org/10.1002/adom.202200965https://doi.org/10.1021/acs.jpclett.3c02763https://doi.org/10.1021/acsanm.4c06648https://doi.org/10.1515/nanoph-2022-0565https://doi.org/10.1021/acsnano.3c11994https://doi.org/10.1002/adom.202503259https://doi.org/10.1073/pnas.0808988106https://doi.org/10.1038/s43586-021-00079-2https://doi.org/10.1038/nphys1422https://doi.org/10.1038/nature23281https://doi.org/10.1021/acs.nanolett.3c01527https://doi.org/10.1038/s41566-025-01733-xhttps://doi.org/10.1021/acsnano.5c15853https://doi.org/10.1038/nnano.2014.180https://doi.org/10.1073/pnas.1017962108https://doi.org/10.1021/nn305826jhttps://doi.org/10.1038/nnano.2011.99https://doi.org/10.1364/OME.9.002511https://doi.org/10.1063/5.0158793https://doi.org/10.1063/1.50170413  23  “  t  13  C  23  q  (3  t  (3  S  O  a3  P  o  o4  BN4  D  R4  o  S  P4  m  34  “  W  24  m  M  14  S  14  C4  P4  P  PSA  ISS 23669608, 2026, 17, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/smtd.70931 by National Institute For, Wiley Online Library on [10/09/2026]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the app3 . J. N. Israelachvili, Intermolecular and Surface Forces (Academic Press,011). 4 . V. R. Dantham, S. Holler, V. Kolchenko, Z. Wan, and S. Arnold,Taking Whispering Gallery-Mode Single Virus Detection and Sizing tohe Limit,” Applied Physics Letters 101 (2012): 043704, https://doi.org/10.063/1.4739473 . 5 . Q. Quan, D. L. Floyd, I. B. Burgess, et al., “Single Particle Detection inMOS Compatible Photonic Crystal Nanobeam Cavities,” Optics Express1 (2013): 32225, https://doi.org/10.1364/OE.21.032225 . 6 . X. Yang, A. Antonov, H. Hu, and A. Tittl, “Permittivity-AsymmetricBIC Metasurfaces for Refractive Index Sensing,” Nanophotonics 142025): 5311–5321, https://doi.org/10.1515/nanoph- 2025- 0415 . 7 . S. Forouhi and E. Ghafar-Zadeh, “Applications of CMOS Devices forhe Diagnosis and Control of Infectious Diseases,” Micromachines 112020): 1003. 8 . S. J. Tang, S. Liu, X. C. Yu, Q. Song, Q. Gong, and Y. F. Xiao, “On-Chippiral Waveguides for Ultrasensitive and Rapid Detection of Nanoscalebjects,” Advanced Materials 30 (2018): 1800262, https://doi.org/10.1002/dma.201800262 . 9 . Y. Podorova, C. H. Vollbrecht, S. J. Evans, et al., “Single-Particlehotothermal Microscopy Using on-Chip Silicon Nitride Microring Res-nators,” Journal of Physical Chemistry A 130 (2025): 298–307, https://doi.rg/10.1021/acs.jpca.5c06077 . 0 . Y. nan Zhang, T. Zhou, B. Han, A. Zhang, and Y. Zhao, “Opticalio-Chemical Sensors Based on Whispering Gallery Mode Resonators,”anoscale 10 (2018): 13832–13856. 1 . M. C. Estevez, M. Alvarez, and L. M. Lechuga, “Integrated Opticalevices for Lab-on-a-Chip Biosensing Applications,” Laser & Photonicseviews 6 (2012): 463–487, https://doi.org/10.1002/lpor.201100025 . 2 . D. Enders, T. Nagao, and T. Nakayama, “Two-Step Desorption Processf Au Nanoparticles in D2 O Suspension on Amino-Terminated SiO2 /Siubstrate Induced by Small Thiol Molecules,” Japanese Journal of Appliedhysics 46 (2007): 3020, https://doi.org/10.1143/JJAP.46.3020 . 3 . T. Nagao, C. V. Hoang, D. Enders, and M. Oyama, Saito in Nano-aterials for Water Management (Jenny Stanford Publishing, 2015),5–70. 4 . M. C. Houghton, N. A. Toropov, D. Yu, S. Bagby, and F. Vollmer,Single Molecule Thermodynamic Penalties Applied to Enzymes byhispering Gallery Mode Biosensors,” Advanced Science 11 (2024):403195. 5 . G. Sun, L. Yuan, Y. Zhang, X. Zhang, and Y. Zhu, “Q-factor Enhance-ent of Fano Resonance in all-Dielectric Metasurfaces by Modulatingeta-Atom Interactions,” Scientific Reports 7 (2017): 8128, https://doi.org/0.1038/s41598- 017- 07715-6 . 6 . M. Iwanaga, “Comprehensive Analysis of Optical Resonances andensing Performance in Metasurfaces of Silicon Nanogap Unit,”Photonics1 (2024): 1053, https://doi.org/10.3390/photonics11111053 . 7 . C. Truong, L. Oudre, and N. Vayatis, “Selective Review of Offlinehange Point Detection Methods,” Signal Processing 167 (2020): 107299. 8 . E. D. Palik, Handbook of Optical Constants of Solids I-III (Academicress, 1998). 9 . J. D. Joannopoulos, S. G. Johnson, J. N. Winn, and R. D. Meade,hotonic Crystals: Molding the Flow of Light , 2nd ed. (Princeton Universityress, 2008). upporting Information dditional supporting information can be found online in the Supportingnformation section. upporting File : smtd70931-sup-0001-SuppMat.pdf. mall Methods, 2026 9 of 9licable Creative Commons Licensehttps://doi.org/10.1063/1.4739473https://doi.org/10.1364/OE.21.032225https://doi.org/10.1515/nanoph-2025-0415https://doi.org/10.1002/adma.201800262https://doi.org/10.1021/acs.jpca.5c06077https://doi.org/10.1002/lpor.201100025https://doi.org/10.1143/JJAP.46.3020https://doi.org/10.1038/s41598-017-07715-6https://doi.org/10.3390/photonics11111053 Single-Nanoparticle Detection Using Quasi-Bound States in the Continuum Supported by Silicon Metasurfaces 1 | Introduction 2 | Results and Discussion 2.1 | Structural and Sensing Characteristics 2.2 | Size Dependence 2.3 | Binding-Induced Changes in qBIC Properties 3 | Conclusion 4 | Experimental Section 4.1 | Fabrication 4.2 | Characterization 4.3 | Simulation Acknowledgements Conflicts of Interest Data Availability Statement References Supporting Information