Materials in fluid states can inherently consist of a variety of substances mixed together, and unlike solid state materials, they can be easily contaminated by externally introduced substances. These impurities within fluid substances can significantly alter the composition and properties of materials, leading to issues like degradation of product quality, reduced production output, and wastage of resources. As a result, in fluid-based processes technology, detecting impurities within fluids plays an important role in process management[1-3].

The conventional techniques for detecting impurities within fluids primarily rely on contact-based methods, wherein electrical and optical sensors are inserted into the fluid[4,5]. Contact-based methods offer the advantage of precisely measuring impurities within the fluid, including factors like concentration and chemical composition. However, these methods have a drawback in that they require placing measurement sensors inside the fluid, which can lead to contamination of the fluid material by the sensors themselves. Particularly in cases where the fluid substance exhibits strong chemical reactivity, it is hard to apply the contact-based method because the chemical reactions between the fluid and the sensors can create new contaminants.

On the other hand, non-contact detection techniques can detect impurities within the fluid without direct contact between the sensor and the fluid, making it possible to detect impurities within the fluid without contamination caused by the sensor[6,7]. Non-contact detection techniques primarily rely on optical methods, detecting impurities through optical phenomena such as changes in brightness and scattering of detection light caused by impurities. Although optical techniques can detect impurities and their distribution quickly with high resolution, they cannot be used when the fluid and fluid transport channels are not optically transparent.

To overcome the limitations of these conventional techniques, recent research has been actively focused on utilizing microwave-based impurity detection methods within fluids. Microwave-based technologies utilize high penetration capabilities of microwaves to a dielectric medium and detects impurities by measuring changes in reflection or transmission of irradiated microwaves caused by the interaction between microwaves and impurities. Indeed, various studies have been reported based on non-contact microwave sensors to detect impurities[8,9]. Similar to inductive proximity sensor methods, microwave sensor methods detect metal impurities through interaction with the oscillating electromagnetic field at gigahertz frequencies[8-10]. However, while microwave sensor technologies offer high measurement sensitivity, they have a drawback that the spatial distribution of impurities within the fluid cannot be determined.

Recently, a new technique for imaging the distribution of microwave near-fields has been reported, named Thermo-elastic Optical Indicator Microscopy (TEOIM)[11]. TEOIM optically visualizes the distribution of microwave near-fields through optical indicators, offering advantages such as fast measurement throughput, optical resolution, and a wide field of view[12-18]. Because the microwave near-field structure depends on the spatial distribution of electromagnetic properties of the species under test, one can expect that the microwave near-field structure will change in the presence of impurities within the fluid, depending on their electromagnetic properties and distribution. Indeed, recent applications of TEOIM have demonstrated its capability to detect the spatial distribution of air bubbles within tubes filled with water[19].

In this study, we report a new method for detection of impurity within fluids based on microwave near field imaging using TEOIM. By using TEOIM, we investigated the change of microwave near field distribution of a thin glass box (depth: 1 mm) filled with water when a metallic impurity was inserted. The measurement results showed that a strong microwave near-fields appeared around the metallic impurity, and the structure of microwave near-fields depended on the position and length of the metallic impurity. By analyzing the changes in near-field distribution, we showed that the size and position of the metallic impurity can be determined from the microwave near-field distribution. This study demonstrates that microwave near-field imaging through TEOIM can be applied to detect metallic impurities within a fluid and measure their size and position.

Figure 1 illustrates the experimental setup of TEOIM for microwave near-field imaging. The signal generator generated microwaves with a frequency of 10 GHz and a power level of -10 dBm. These microwaves were amplified to a power level of 30 dBm by the power amplifier (gain: 40 dB), and then transmitted through a coaxial cable to a waveguide. Finally, the transmitted microwaves were radiated through the open end of the waveguide antenna to the specimen. The waveguide antenna was designed to be rotatable, allowing for changes in the orientation of the electric field direction of the radiated microwaves.

Figure 1. (Color online) Illustration of measurement system. The yellow arrows indicate propagation direction of probing light, and blue arrows indicate coaxial cables. The incident light is polarized circularly by a circular polarizer and propagates to the optical indicator. The reflected light from the optical indicator passes through a linear polarizer, and then propagates to the CCD camera. During the measurement, the optical indicator is irradiated by microwaves through an open-ended waveguide. The microwave signal is generated by a signal generator and amplified by a power amplifier connected to the power supply.

Microwave near-field imaging was conducted through an optical microscope system consisting of a light source, a circular polarizer, an optical indicator, a linear polarizer, and a CCD camera. In this study, a white LED surface light source with green-colored filter (λ = 530nm) was used as the light source, and a soda lime glass substrate (100 mm × 100 mm × 1 mm) coated with an ITO (Indium Tin Oxide) thin film (thickness: 200 nm) was used as the optical indicator. The emitted light from the light source is polarized into left-circular polarization (LCP) state by the circular polarizer and then directed onto the optical indicator. The incident light passing through the optical indicator undergoes polarization changes depending on the microwave near-field distribution acting on the indicator. Subsequently, the reflected light passes through the linear polarizer and is finally captured by the camera.

Figure 2 illustrates structure of the sample used in the experiment and the measurement principle of TEOIM. A thin glass box was fabricated by attaching two thin glass plates (width: 75.5 mm, height: 55 mm, thickness: 1.2 mm) with a separation of 1.2 mm using silicon adhesive. A copper wire (radius: 1 mm) was affixed at the center of the box, followed by filling it with distilled water. The prepared sample was attached to the optical indicator, and microwaves were radiated to the sample to visualize the distribution of microwave near-fields.

Figure 2. (Color online) Illustration of sample structure and measurement principles. A thin glass box was fabricated by attaching two thin glass plates with a separation of 1.2 mm using silicon adhesive. A copper piece was affixed at the center of the box, followed by filling it with distilled water. The prepared sample was placed between the waveguide and optical indicator, and it was irradiated by microwaves. As the incident light passes through the glass substrate of the indicator (indicated by yellow arrows), its polarization state is changed from circular to elliptical due to the photo-elastic effect induced by thermal stress in the glass substrate. The thermal stress is caused by the Joule heating generated by microwave current in the ITO thin film.

Microwaves (frequency: 10 GHz) emitted from the open end of the waveguide propagate into the thin glass box, and these propagated microwaves interact with the metallic impurity within the box. When microwaves interact with the metallic impurity, microwave near-fields are formed around the impurity, and the structure of near-fields depends on the spatial distribution and position of the impurity. When an optical indicator is placed within the area of the microwave near-field, strong microwave currents can flow within the ITO thin film of the indicator due to the magnetic component of the microwave near-field. This results in joule heating of the ITO film due to the microwave currents, and the generated heat is diffused into the glass substrate of the indicator, creating thermal stress within the substrate. The spatial distribution of thermal stress is determined by the distribution of the heat source, which in turn mimics the distribution of the microwave near-fields. Therefore, by measuring and analyzing the thermal stress distribution of the optical indicator, one can visualize the microwave near-fields distribution[11].

The thermal stress distribution of the optical indicator can be measured through photo-elastic effects. When circularly polarized light passes through the indicator, the polarization state of the incident light changes from circular state to elliptical state due to the photo-elastic effects induced by the thermal stress within the indicator. The intensity of the elliptically polarized light after passing through the linear polarizer can be expressed as follows[20]:

where, I is the intensity of light passing through the linear polarizer, x and y are spatial coordinates, E_i is the magnitude of the electric field of the incident light, φ is the angle between the optical axis of the linear polarizer, β is the linear birefringence due to the photo-elastic effects, and θ is the angle between the optical axis of the linear polarizer and the principal axis of stress. When the thickness of the indicator is significantly smaller than its area, the thermal stress within the glass substrate can be divided into normal and shear stress components. Then, the distribution of linear birefringence induced by each stress component can be expressed as follows[11, 20]:

where, β1 and β2 are the linear birefringence induced by perpendicular stress and shear stress, I is the measured intensity, MWON and MWOFF are the states with and without microwave irradiation, respectively. Finally, the heat source distribution can be expressed as the spatial derivatives of β1 and β2, as shown below[11, 20]:

where, q represents the heat source density, x and y are spatial coordinates, and C denotes a constant dependent on the wavelength of light and the photo-elastic coefficients of the substrate. Therefore, from the intensity distribution measured by the CCD camera, one can calculate the heat source distribution to visualize the microwave near-fields.

Figures 3(a) and (b) show the microwave near-field distribution of a thin glass box filled with water according to horizontal (horizontal configuration) and vertical (vertical configuration) orientations of the electric field of the microwaves, respectively, where no metal impurity was inserted in the glass box. The measurement results showed that microwave near-fields were formed at the center of the waveguide for both horizontal (horizontal configuration) and vertical (vertical configuration) orientations of the electric field of the microwaves. Furthermore, it was observed that in the horizontal configuration, the near-field distribution was elongated in the vertical direction, whereas in the vertical orientation, the distribution was elongated in the horizontal direction. These results indicate that the microwave near-fields are elongated in the vertical direction with respect to the electric field direction of the microwaves, suggesting an influence from the distribution of microwaves radiating from the waveguide. The TE-mode waveguide used in the experiment is rectangular type, where the electric field of the microwaves aligned parallel to the short axis of the waveguide. Moreover, the intensity of microwaves within the waveguide is at its peak at the center and gradually diminishes as one moves away from the center along the longitudinal axis of the waveguide. Consequently, when microwaves are emitted from the waveguide antenna, they manifest elongation perpendicular to the direction of the electric field.

Figure 3. (Color online) (a–b) Microwave near-field distributions of a water-filled glass box without metal impurities according to microwave electric field (MW-E) direction: (a) MW-E was parallel to the horizontal direction (horizontal configuration); (b) MW-E was perpendicular to the horizontal direction (vertical configuration). The bi-directional red arrows indicate the MW-E direction.

The observed microwave near-field distribution of the water-filled glass box was different from the previous observation where the microwave near-field appeared along the length direction of the water-filled tube[19]. This distinction can be attributed to the isotropic spatial distribution of water in the box regarding the microwave electric field. Therefore, these measurement results indicate that the water-filled glass box exhibits isotropic properties with respect to the microwave electric field, and that microwaves can penetrate through a fluid with high conductivity, such as water.

To validate the detection of impurities within the fluid through microwave near-field imaging, a copper piece was placed in a thin glass box filled with water, and the microwave near-field distribution was visualized by TEOIM. The copper piece was made by cutting a copper wire with a diameter of 1 mm, and it was attached at the center of the glass box with its length direction aligned horizontally. Figures 4(a–d) show the microwave near-field distribution in the water-filled glass box containing copper pieces of lengths 3 mm (a–b) and 5 mm (c–d) for both horizontal and vertical configurations. The measurement results showed that in the horizontal configuration, a strong ring-shaped microwave near-field appeared around the copper piece, whereas in the vertical configuration, two regions showing a strong microwave near-field appeared above and below the copper piece. These microwave near-field distributions were distinctly different from the case without a metal piece, and as the length of the copper piece increased, these structural characteristics became more pronounced. Therefore, these experimental results indicate that one can determine the presence and size of a metal impurity in a fluid from the change of microwave near-field distributions. Furthermore, these results imply that in the presence of metallic impurities within the fluid, strong microwave magnetic fields form around the metal, indicating that the metal alters the electromagnetic properties of the surrounding water. These intriguing results have sparked further ongoing research to understand the underlying mechanisms.

Figure 4. (Color online) Microwave near-field distributions in a water-filled glass box with metal impurities of varying lengths: (a–b) 3 mm long copper piece measured with horizontal and vertical configuration; (c–d) 5 mm long copper piece measured with horizontal and vertical configuration. The bi-directional red arrows indicate the MW-E direction.

The most important advantage of the TEOIM-based microwave near-field imaging technique is that it can visualize the distribution of microwave near-fields over a wide range with a fast measurement throughput[19]. Indeed, studies have already reported using these advantages to visualize the distribution of electromagnetic defects present in samples[13, 15]. Thus, one can expect that the spatial distribution of impurities within a fluid can be detected with a fast measurement throughput by TEOIM. To demonstrate this idea, the microwave near-field distribution was visualized according to the position of the copper piece in the water-filled glass box. Figures 5(a–b) show the microwave near-field distribution when the copper piece was moved to the left and right from its center, respectively. From the measurement results, one can see that around the copper piece, ring-shaped microwave near-field distributions appeared, and regions with the lowest intensity of the microwave near-field appeared at the copper piece. Therefore, it can be concluded that one can determine the position of metal impurities by identifying the lowest intensity region in the ring-shaped microwave near-field distributions. These experimental results indicate that the spatial distribution of microwave near-fields varies with the size and position of metallic impurities, and by analyzing these spatial structural changes, it is possible to measure not only the presence of impurities but also their size and position.

Figure 5. (Color online) Microwave near field distributions of water-filled glass box with metal impurities with different position: (a) copper piece placed left side from the center of the optical indicator; (b) copper piece placed right side from the center of the optical indicator. The bi-directional red arrows indicate the MW-E direction, and the dashed black line indicates the center of the optical indicator.

In this study, a new method for non-destructive detection of metallic impurities within a fluid using TEOIM-based microwave near-field imaging was presented. We visualized the microwave near-fields distribution of a water-filled glass box containing metal impurity with different sizes and positions. From the experimental results, we showed that the metal impurities change the microwave near-field distribution of the water-filled glass box, and one can determine the position and size of the metal impurities by analyzing the change of the near-field distribution. Therefore, it can be concluded that TEOIM-based microwave near-field imaging can be utilized for non-destructive technology for detection of impurities within fluid substances.

This work was supported by the National Research Foundation of Korea (NRF) Grant funded by the Korean Government [Ministry of Science and ICT (MSIT)] under Grant 2020R1C1C1004556.