Three-dimensional close-to-substrate trajectories of magnetic microparticles in dynamically changing magnetic field landscapes

Rico Huhnstock, Meike Reginka, Claudius Sonntag, Maximilian Merkel, Kristina Dingel, Bernhard Sick, M. Vogel, Arno Ehresmann

TL;DR

Researchers have developed a label-free method to track the three-dimensional "hopping" trajectories of magnetic microparticles in microfluidic environments using standard optical microscopy. By correlating image sharpness (the Tenenbaum gradient) with vertical position, the team experimentally confirmed theoretical predictions of micrometer-scale vertical jumps, paving the way for high-precision, low-cost biosensing in Lab-on-a-chip (LOC) devices.

What Problem Does This Paper Address?

Lab-on-a-chip (LOC) systems rely on the precise manipulation of magnetic microparticles (MPs) to transport, mix, and detect biological analytes. While lateral (2D) tracking of these particles is routine, measuring their vertical ($z$) distance from the substrate is significantly more challenging.

This vertical coordinate is crucial because the interactions between a particle and the substrate—governed by magnetic, electrostatic, and van der Waals forces—change when an analyte (like a disease marker) binds to the surface. Currently, resolving these tiny vertical shifts often requires expensive or complex setups like confocal microscopy or magnetic force microscopy. The authors sought a simpler, "label-free" way to quantify 3D motion using only bright-field microscopy and high-speed video.

Methods

The authors utilized a combination of hardware and software innovations to achieve 3D tracking:

  • Magnetic Actuation: They used a substrate with "exchange-biased" magnetic thin films patterned into 5 µm-wide stripe domains. By applying external magnetic field pulses, they induced a "hopping" motion where particles jump from one domain wall to the next.
  • The Tenenbaum (TB) Gradient: To find the $z$-coordinate, the team measured image sharpness. They employed the Tenenbaum gradient, a metric that sums the squares of image intensity gradients.
  • Calibration: Using a piezo-controlled stage, they swept particles through the microscope's focal plane in 25–50 nm steps to create a calibration curve relating TB gradient values to specific heights.
  • Tracking Software: Lateral ($x, y$) positions were tracked using AdaPT, a machine learning-aided software designed for spherical particles.
  • High-Speed Imaging: Data was captured at 1,000 frames per second (fps) to resolve the rapid dynamics of the particle "hops."

Datasets

While this paper does not introduce a large-scale machine learning benchmark, it utilizes specific standardized superparamagnetic particles (SPPs) for evaluation:

  • micromer®-M: Particles with diameters of 3 µm and 4 µm (featuring a non-magnetic core).
  • Dynabeads™ M-270: Particles with a diameter of 2.8 µm (featuring a homogenous distribution of magnetic material).
  • Simulated Data: The experimental results were compared against a numerical model generated using the OOMMF (Object Oriented Micromagnetic Framework) to simulate magnetic field landscapes.

Key Insights and Ideas

  • Sharpness as a Ruler: The most significant insight is that the Tenenbaum gradient provides a high-resolution proxy for vertical distance. Interestingly, the calibration curve is asymmetric; the "increasing flank" (as the particle moves toward focus) is steeper and more sensitive than the "decreasing flank," making it the ideal regime for tracking.
  • Experimental Confirmation of "Hopping": Theory has long predicted that particles in these magnetic landscapes don't just slide; they "hop" in a 3D arc. This paper provides some of the first quantitative experimental evidence of this specific trajectory.
  • Sensitivity to Optical Response: The method proved robust across different particle types (micromer® vs. Dynabeads™), even though their internal structures and optical responses differ significantly.

Key Results

  • Quantified Vertical Motion: For 4 µm particles, the researchers measured vertical jumps of up to 8.7 µm above the initial equilibrium position.
  • Statistical Agreement: Across multiple particles, the average maximum height jump was $6 \pm 2$ µm, which aligns well with the theoretical estimate of 6.8 µm.
  • High Precision: The system achieved a $z$-coordinate uncertainty of approximately 500 nm for 4 µm particles, which is sufficient to detect significant binding events or surface variations.
  • Temporal Resolution: The 1,000 fps capture allowed the team to observe that the vertical jump occurs almost simultaneously with the lateral shift, though with a slight lag due to fluidic drag.

Limitations and Future Work

The authors acknowledge that their current theoretical model is static—it calculates equilibrium positions based on forces but does not fully account for inertial forces or the complex dynamics of friction in a viscous medium. This explains why the experimental particles appear to move more "smoothly" and with a slight time delay compared to the "abrupt" jumps predicted by the model.

Future Work: 1. Bio-detection: The next step is to use this 3D tracking to detect analyte binding. If a target molecule binds to the particle, the change in the particle-substrate separation distance should be detectable via these 3D trajectories. 2. Surface Analysis: The method could be used as a "multiplexed" version of magnetic force microscopy, using hundreds of particles simultaneously to map the magnetic characteristics of a surface.

Summary updated

Abstract

The transport of magnetic nano- or microparticles in microfluidic devices using artificially designed magnetic field landscapes (MFL) is promising for the implementation of key functionalities in Lab-on-a-chip (LOC) systems. A close-to-substrate transport is hereby instrumental to use changing particle-substrate interactions upon analyte binding for analytics and diagnostics. Here, we present an essential prerequisite for such an application, namely the label-free quantitative experimental determination of the three-dimensional trajectories of superparamagnetic particles (SPP) transported by a dynamically changing MFL above a topographically flat substrate. The evaluation of the SPP sharpness within defocused video-recorded images, acquired by an optical bright-field microscope, was employed to obtain a vertical z-coordinate. This method applied to a prototypical transport scheme, using the static MFL of parallel-stripe domains superposed by a particular magnetic field pulse sequence, revealed a hopping-like motion of the magnetic particles, previously predicted by theory. Maximum vertical particle jumps of several micrometers have been observed experimentally, corroborating theoretical estimates for the particle-substrate distance. As our findings pave the way towards precise quantification of particle-substrate separations in the discussed transport system, they bear deep implications for future LOC detection schemes using only optical microscopy.
BibTeX
@inproceedings{huhnstock2022three,
  author = {Rico Huhnstock and Meike Reginka and Claudius Sonntag and Maximilian Merkel and Kristina Dingel and Bernhard Sick and M. Vogel and Arno Ehresmann},
  title = {{Three-dimensional close-to-substrate trajectories of magnetic microparticles in dynamically changing magnetic field landscapes}},
  booktitle = {Scientific},
  year = {2022},
  doi = {10.1038/s41598-022-25391-z},
}

Loading related content...

Figure