Near-field microwave imaging offers a way to investigate materials in liquid without exposing them directly to an electron beam. In a feasibility study, researchers used an atomic force microscopy (AFM) platform to detect polystyrene particles inside a glycerol-filled cavity beneath a thin silicon nitride membrane. The result points to an imaging approach for sensitive or reactive samples that need to remain in their native environment.
For industrial automation and materials teams, the broader relevance is measurement access: some processes and materials are difficult to examine when a sample must be removed from a liquid, sealed environment or operating condition. Microwave probing through a membrane could help address that challenge, although the demonstrated configuration is an early proof of concept rather than a production inspection method.
Why image through a membrane?
Liquid-cell electron microscopy can reveal nanoscale structures in liquid, but the electron beam may also alter sensitive specimens. Beam exposure can trigger chemical reactions in the liquid and damage materials such as soft matter. This complicates interpretation when the measurement itself changes the sample.
Optical imaging avoids electron bombardment, but conventional far-field optics are constrained by diffraction, limiting their ability to resolve features at the nanoscale. Near-field methods take a different approach: a probe concentrates and senses a field close to the sample, so spatial resolution depends largely on the probe rather than on the wavelength alone.
Microwaves are useful in this context because they have low photon energy and can penetrate materials that are opaque to visible light. The study explored whether a microwave probe could detect objects on the far side of a thin dielectric membrane while the objects remained immersed in liquid.
How near-field microwave imaging works
The system combined a metallic AFM probe with a microwave reflectometer. As the tip scanned the membrane, the instrument measured the amplitude and phase of reflected microwave signals. These signals depend on the impedance of the tip–sample region, which is influenced by local electrical properties.
In scanning microwave impedance microscopy, the measured response is commonly separated into capacitive and resistive channels. The capacitive channel is sensitive to dielectric properties, while the resistive channel relates to electrical losses and conductivity. Because the electric field is concentrated near the probe tip, the measurement is most responsive to the material within that localized field.
The experiment used a 50-nanometer-thick silicon nitride membrane supported by a silicon frame. Polystyrene particles were placed in a cavity beneath the membrane, which was filled with glycerol and sealed. Imaging was performed with the AFM probe in contact with the membrane.
What the experiment demonstrated
The membrane remained stable enough for contact-mode scanning with the relatively soft AFM cantilever used in the work. Microwave images distinguished the polystyrene particles from the surrounding glycerol. Their different dielectric properties produced contrast in the capacitive imaging channel, even when the particles were not evident in the corresponding topographic image.
This distinction matters for subsurface measurement: a feature can produce an electrical or dielectric signal without creating a visible change in the surface profile. The simultaneous acquisition of topography, deflection error and microwave data also helped researchers assess how the probe interacted with the membrane and the objects beneath it.
Some particles appeared in both the microwave and deflection-error images, suggesting that the membrane deformed locally as the probe passed over particles in contact with its underside. The membrane nevertheless remained intact during scanning. These observations indicate that mechanical behavior is part of the measurement—not merely a packaging concern—and should be considered when designing thin-window imaging experiments.
Interpreting depth and resolution
The results support detection beneath the membrane, but they do not establish a single, definitive probing depth. Modeling showed that the microwave field coupled to the particle, while also being screened by the membrane. In the images of agglomerated particles, the apparent microwave contrast was smaller than the particle diameter. Based on particle geometry, the authors estimated a probing depth of roughly 50 to 110 nanometers in that configuration.
Images of two isolated particles produced a broader signal profile. The researchers suggested that the instrument might be sensitive to objects at depths of around one micrometer or more, but noted that probe-side fields needed further analysis before that interpretation could be verified. For engineers evaluating the method, the practical lesson is to treat depth as dependent on the complete measurement geometry—including probe, membrane, liquid and target—not as a universal instrument specification.
Implications for in-situ measurement
The study identifies several factors that matter when adapting near-field microwave imaging to other samples:
- Membrane mechanics: Thickness and support geometry must provide adequate strength while allowing the field to reach the target.
- Probe interaction: Tip force and cantilever stiffness affect contact stability and the risk of deforming the membrane.
- Electrical contrast: Differences in dielectric properties or conductivity determine whether the object can be distinguished from its surrounding liquid.
- Signal interpretation: Modeling and controlled measurements are needed to separate target response from membrane screening and probe-side effects.
These considerations are relevant to research on biological materials, electrochemical systems and other samples that are difficult to examine outside their environment. However, the reported tests used polystyrene particles in glycerol; applying the technique to other liquids and materials requires additional validation.
Conclusion
Near-field microwave imaging demonstrated that nanoscale objects can be detected through a thin silicon nitride membrane while immersed in a liquid. The work also showed why successful subsurface measurement depends on balancing membrane strength, probe contact, dielectric contrast and field penetration. For teams developing in-situ characterization methods, the approach is a promising option to investigate further—particularly when electron-beam exposure may disturb a sensitive sample—but its depth and performance must be established for each measurement configuration.