Single-Molecule Sizing through Nanocavity Confinement

Abstract

An approach relying on nanocavity confinement is developed in this paper for the sizing of nanoscale particles and single biomolecules in solution. The approach, termed nanocavity diffusional sizing (NDS), measures particle residence times within nanofluidic cavities to determine their hydrodynamic radii. Using theoretical modeling and simulations, we show that the residence time of particles within nanocavities above a critical time scale depends on the diffusion coefficient of the particle, which allows the estimation of the particle's size. We demonstrate this approach experimentally through the measurement of particle residence times within nanofluidic cavities using single-molecule confocal microscopy. Our data show that the residence times scale linearly with the sizes of nanoscale colloids, protein aggregates, and single DNA oligonucleotides. NDS thus constitutes a new single molecule optofluidic approach that allows rapid and quantitative sizing of nanoscale particles for potential applications in nanobiotechnology, biophysics, and clinical diagnostics.

Keywords: biosensing; confocal detection; microfluidics; nanofluidics; protein sizing; single molecules.

Figure 1

Principle of nanocavity diffusional sizing (NDS). (a) Schematic illustration of the experimental implementation of the NDS approach. The detection volume of a confocal microscope is placed inside the nanocavity of the nanofluidic chip, and residence times of single particles are recorded as they diffuse in and out of the nanocavity. The nanofluidic chip is fabricated by hybrid lithography (see Methods). Single particles are shown in red. (b) Schematic of the nanofluidic chip used for NDS measurements. The nanocavities are located adjacent to nanofluidic channels on a microfluidic chip. An SEM image of the chip is depicted in the right panel. Adapted with permission from Vanderpoorten et al. Copyright 2022 ACS. (c) Workflow of the sizing experiment. First, the particles are detected by confocal microscopy as they diffuse into and out of the nanocavity (shown is a 2D representation of the nanocavity with an adjacent nanochannel, as depicted in panel b). Then, the residence times t are extracted from the recorded time trace and binned in a residence time histogram. This histogram is then fit with an exponential function of the type , from which the hydrodynamic radius R_H can be extracted. The coefficient τ is inversely proportional to R_H.

Figure 2

Theory and simulation of diffusion under nanoconfinement. (a) 2D schematic representations of the nanocavity model with particles entering and exiting the cavity (left), the corresponding 1D positions of particles along the y-axis (center panel), and the time trajectories along the yu-axis (right panels). For a particle diffusing within a nanocavity, two diffusive scenarios can be distinguished: (i) the particle enters the cavity and exits it without reaching the bottom wall (top panels) or (ii) the particle enters the cavity and reaches the bottom of the well before exiting it (bottom panels). The right panels show displacement in time along the y-axis of the diffusion processes. The particle enters at time t_in and exits at time t_out. The cavity depth is d_c, and d_E denotes the effective cavity depth (d_E = d_c – R), with R being the radius of the particle. (b) Analytical modeling of the particle diffusion at short time scales (t ≪ t_c, left panel) and long time scales (t ≫ t_c, right panel). Shown are residence time probability plots; t_c denotes the critical time (see main text). At short time scales (t ≪ t_c), the residence time is independent of the size of the particle. At long time scales (t ≫ t_c), the residence time follows an exponential decay that depends on the particle’s size. Modeled were three particles with different radii R′ fixed at R = R′ (blue), R = R′/2 (orange), and R = R′/10 (green). Diffusion was modeled as a 1D random walk. (c) Simulation results for the diffusion of particles within a nanocavity. Shown are residence time probability plots (log–log plot, left panel; linear–log plot, right panel) for particles of different seizes (50 nm, blue; 25 nm, green; 2.5 nm, purple). At short residence times, particles are scale-invariant. At long time scales, particle residence times exhibit an exponential decay, which is dependent on the size of the particle. Data points represent simulation results. Long and short dashed lines depict fits of the simulation data by power law (p-law fit) and exponential functions (exp fit), respectively.

Figure 3

Nanofluidic diffusional sizing (NDS) of single particles in solution. (a) Experimental setup of the NDS experiment. The observation volume of the confocal microscope is placed within a nanocavity. The fluorescence of particles or biomolecules of interest is observed as they diffuse in and out of the confocal volume. Lower panel: SEM micrographs of the nanofluidic device with nanocavity functionalities used in NDS experiments. SEM micrographs were adapted from Vanderpoorten et al. Copyright 2022 ACS. (b) Examples of time traces from single-molecule detection of nanocolloids, α-synuclein oligomers, and a DNA oligonucleotide within nanocavities. Highlighted in red are the times when a molecule or particle was present within the confocal detection volume. The bin time is 1 ms in all traces. (c) Residence time decay histograms. The data were fit with an exponential function of the form . The slope of the curves gives the decay time. The dotted line corresponds to the critical time t_c for each species. The error bars are standard deviations of the Poisson distribution (), which was calculated from the number of events per bin (N). The following number of single-molecule events were probed in order to create residence time histograms: 45-bp DNA, 969 events; α-synuclein oligomers, 1410 events; 50 nm colloids, 760 events; 20 nm colloids, 1137 events. (d) Extracted decay times versus hydrodynamic radii. The dotted line is a linear fit of the data points. Hydrodynamic radii for the colloids and the DNA were measured by dynamic light scattering (DLS) and for the oligomers by analytical ultracentrifugation (AUC). For a description of the error bars, see Table 1.