1. Introduction

Terahertz (THz) electromagnetic waves exhibit a remarkable sensitivity to water and are considered safe for use in biological tissues, rendering them viable for advancing next-generation biomedical diagnostic technologies [1,2]. The THz electromagnetic spectrum aligns with the energies associated with weak hydrogen bonds and large molecules, facilitating spectroscopy and imaging studies that leverage the unique vibrational frequencies of biochemical molecules [3]. These distinct characteristics enable the analysis of drug components, detection of counterfeit drugs, and screening for narcotics [4]. The THz waves uniquely correspond to the recombination and dissociation energies of water molecular clusters, endowing them with exceptional sensitivity to water [5,6]. This heightened sensitivity allows THz waves to precisely measure the distribution and quantity of water within the bio-samples, enabling their application in diagnostic devices for various diseases, such as tumors and burns [7,8,9,10,11]. Tumors, characterized by the presence of new blood vessels and active metabolism, exhibit distinct water and fat distribution patterns compared to normal tissues. THz waves excel at visualizing these disparities in water and fat content within tissues, positioning them as a promising next-generation technology for cancer diagnosis, with various research endeavors currently underway. For burn injuries or scars, discernible differences exist in terms of tissue density, structure, and water distribution when compared to normal unaffected areas [12]. THz waves offer the capability to characterize the condition of burns or scars based on these variations in tissue properties. However, imaging biological tissues, which contain approximately 70% water, poses challenges when employing THz transmission imaging [13]. Consequently, reflective THz imaging has been favored for most biological tissue measurements. For reflective THz imaging measurements of biological tissues, the tissue was placed on a quartz or silicon substrate, with the resulting image generated from the reflected signal from the samples. The reflection coefficients of wave are related with the difference of refractive indices [14]. Unfortunately, the refractive index difference between the substrate surface and the tissue surface could reduce the intensity and signal-to-noise ratio of the THz signals delivering to the tissues and obscuring both the quantitative and qualitative information. To address these challenges, there is pressing need for a technology that compensates for the refractive index difference between these two surfaces. Ultrasound technology has previously employed special gel type impedance-matching materials to prevent image distortion caused by reflected waves at the skin surface or to prevent skin burns arising from the refractive index disparity between air and skin during high intensity focused ultrasound therapy [15,16]. Hence, the development of THz impedance-matching technology, bridging the refractive index differential between the substrate surface and the tissue surface, becomes essential for enhancing the performance of THz biomedical imaging and facilitating high-power THz wave therapy. The ideal material for THz impedance-matching must exhibit biocompatibility, exceptional transparency to THz waves, and resistance to absorption by tissue, ensuring the preservation of tissue properties. In this study, we investigated the impedance-matching effect of THz waves by employing various biocompatible liquids, including water, glycerol, and petroleum jelly. Our research leverages THz two-dimensional (2D) imaging and THz time-of-flight (TOF) imaging applied to skin phantom with these materials to elucidate the THz wave impedance-matching effect and compare their efficacy. Subsequently, the THz wave impedance-matching effect was confirmed in the skin of mice.

2. Experimental methods

We utilized a reflection-mode THz imaging system for acquiring the images and THz waveforms as shown Fig. 1. The THz pulses were obtained using femtosecond fiber laser and photoconductive antennas. The femtosecond fiber laser had a center wavelength and pulse width of 1.5 µm and 80 fs, respectively. A fiber-coupled antenna (TERA15-TX-FC, Menlo system) served as the photoconductive antenna for THz pulse generation, while another fiber-coupled dipole antenna (TERA15-RX-FC, Menlo system) was employed for detecting THz signals. To rapidly acquire THz signals, we utilized a fast scanner with frequency and amplitude settings of 20 Hz and 30 ps, respectively. The THz signals were amplified via a low noise current free amplifier (SRS570, Stanford Research), and then digitized through a data acquisition board. The generated THz pulses were guided by a polymethyl-pentene (TPX) lens and plate metal mirror. The TPX lens focused the THz signals onto the sample stage with 0 degree of incidence angle, while the vertically reflected signals from the samples were guided to a silicon beam splitter, another TPX lens, and a detector. Notably, all systems excluding the sample, were placed in a dry air chamber to avoid signal distortion due to water vapor absorption. Tissues were placed on a crystallized z-quartz window of 3 mm thickness. The sample was moved along the x and y axes to obtain 2D images. The THz imaging size was 40 × 40 mm² and we obtained the THz imaging 160 × 160 of pixel number with scanning resolution of 0.25 mm. The scanning times took up to 22 minutes. THz band of THz pulses was from 0.2 to 2 THz and the polarization of THz pulses was horizontal.

 

Fig. 1. Scheme of reflection-mode THz imaging system. The THz imaging system was based on the fiber laser.

Download Full Size | PDF

3. Results

3.1 THz image of candidate liquids

We obtained THz reflection images of three materials-water, glycerol, and petroleum jelly to understand their impedance-matching effect with a quartz window, as illustrated in Fig. 2. The order of increasing reflection intensity was as follows: glycerol (lowest), water, petroleum jelly, and air. This result is consistent with our previous findings. The complex refractive index of glycerol, water, and petroleum jelly were 1.9 + 0.143i, 2.25 + 0.764i and 1.5 + 0.009i at 0.5 THz, respectively [17]. The intensity of the reflection image is related to the refractive index difference between the materials and quartz, which has a refractive index of 1.95 + 0.003i at 0.5 THz [18]. This indicates that glycerol has a considerable impedance-matching effect on the quartz window.

 

Fig. 2. Imaging of water, glycerol and petroleum jelly drop on quartz window: (a) Photograph and (b) THz peak to peak images of air (A), water (W), glycerol (G), and petroleum jelly (P) (norm.: normalized amplitude).

Download Full Size | PDF

The THz time-domain waveforms at a single point in the images are presented in Fig. 3. The peak-to-peak THz time-domain waveforms increased in the following order: glycerol, water, petroleum jelly, and air. This result strongly suggests that glycerol can be employed as an impedance-matching material for quartz, since the refractive index of glycerol is similar to that of quartz, even though petroleum jelly exhibits the lowest absorbance among the three materials.

 

Fig. 3. THz waveforms of water, glycerol and petroleum jelly drop on quartz window of Fig. 2 (a.u.: arbitrary unit),

Download Full Size | PDF

3.2 THz impedance-matching effect in a skin phantom

We evaluated the impedance-matching effect using a skin phantom model comprising commercial hydrogel patches, commonly used for wound healing. Figure 4 presents the complex THz optical constants of the commercial hydrogel patch utilized in this study. These values were obtained using the conventional transmission THz time-domain spectroscopy.

 

Fig. 4. THz complex constants of commercial hydrogel patch (a) refractive indices and extinction coefficients and (b) absorption coefficients of commercial hydrogel patch.

Download Full Size | PDF

In this experiment, water, glycerol, and petroleum jelly were applied to the surface of the hydrogel patches, which were placed on a quartz window. The THz pulses irradiated the hydrogel patches, which passed through the quartz window, and subsequently measuring the reflected signals. The images were obtained using a point-to-point scanning method. A THz reflection image is depicted in Fig. 5. Figure 5(a) provides a photograph of the sample, with no differences observed as all materials were transparent to visible light. In contrast, the THz image (Fig. 5(b)) reveals variations among the samples. Specifically, the reflected intensity in the THz image was similar to that shown in Fig. 2 and 3, with increasing intensity in the order of glycerol, water, petroleum jelly, and air. Notably, when the sample was placed directly on the quartz window without any impedance-matching material, the 2D THz images were distorted. The complete image of the sample was not observed, and only a partial THz image was obtained as seen in Fig. 5(b).

 

Fig. 5. Imaging of hydrogel patch on the quartz window (a) Photograph and (b) THz peak to peak images of air (A), water (W), glycerol (G), and petroleum jelly (P) (norm.: normalized amplitude).

Download Full Size | PDF

This distortion can be attributed to the poor adhesion of the hydrogel patch to the quartz window, resulting from the bending of the hydrogel patch and the presence of wrinkles on its surface. However, when impedance-matching materials, such as water, petroleum jelly, or glycerol, were applied, a stable adhesion to the quartz window was achieved. This led to the production of more uniform THz images, as the impedance-matching materials effectively filled the gaps between the wrinkles on the patch’s surface and the quartz window, reducing the impact of the wrinkles. The results can be depicted by TOF images. To further observe the THz pulse penetration characteristics of each material within the skin phantom, we obtained THz TOF images along the horizontal and vertical dotted line in Fig. 5, as presented in Fig. 6 and 7.

 

Fig. 6. THz TOF images of vertical short dot line of Fig. 5(a) Hydrogel patch without impedance-matching materials and applied glycerol, and (b) Hydrogel patch applied water and petroleum jelly. The dot line boxes indicate the location of samples (a.u.: arbitrary unit).

Download Full Size | PDF

 

Fig. 7. THz TOF images of horizontal long dot line of Fig. 4: (a) Hydrogel patch without and with water, and (b) Hydrogel patch applied glycerol and petroleum jelly (a.u.: arbitrary unit).

Download Full Size | PDF

The THz TOF images showed the changes in THz 1^st pulses reflected from the interface between the skin phantom and quartz window as shown in Fig. 6, 7. This distortion attributed to the poor adhesion of the hydrogel patch to the quartz window, resulting from the bending of the hydrogel patch, were observed in TOF images as shown in left part of Fig. 6(a). The THz 2^nd pulses that pass through the skin phantom and are reflected from the top of the skin phantom in contact with the air layer showcase the similar result of 1^st pulses.

The THz TOF images highlight the changes in THz 1^st pulses reflected from the interface between the skin phantom and quartz window as shown in Fig. 6, 7. They also showcase the THz 2^nd pulses that pass through the skin phantom and are reflected from the top of the skin phantom in contact with the air layer. In the case of the skin phantom without any applied impedance-matching material, strong reflections in both 1^st and 2^nd THz pulses were observed. However, when water was used as the impedance-matching material, we observed a 40% reduction in reflectance on the quartz window surface compared to signals without any material as depicted in Fig. 7(a). Additionally, the lowest THz 2^nd pulse intensity was exhibited for water, primarily due its high absorption of THz waves. Conversely, when glycerol was applied, lowest reflectance was observed in THz 1^st pulses, and the intensity of the THz 2^nd pulses intensity was similar to that of petroleum jelly. Petroleum jelly exhibited the highest THz 2^nd pulse intensity owing to its low absorbance in the THz wave. However, it exhibited reflectance that was 3.7 times higher in the first THz pulses compared to glycerol as depicted in Fig. 7(b). This is because the absolute difference of refraction index between the quartz window and petroleum jelly was larger than that between glycerol and water, although the refractive index of petroleum jelly was the smallest among the three substances. These results are further illustrated in Fig. 8, which presents the THz time-domain waveforms obtained along the dotted lines in Fig. 8. The THz time-domain waveform of the sample placed directly on the quartz window without impedance-matching material exhibited the largest reflected signal. The reflected signals at the interface between the skin phantom and quartz window decreased in the following order: air, petroleum jelly, water, and glycerol. The THz 2^nd pulse reflected after penetrating the skin phantom decreased in the following order: glycerol > petroleum jelly > air. These results suggest that glycerol is the most suitable candidate for use as an impedance-matching material. They also confirm that the utilization of impedance-matching materials significantly enhances the efficiency of THz pulse penetration in biomedical imaging and therapeutic applications.

 

Fig. 8. THz time-domain waveforms of the dotted lines of Fig. 7 (a.u.: arbitrary unit).

Download Full Size | PDF

To depict the terahertz impedance-matching effect, we obtained a three-dimensional volumetric image as shown in Fig. 9. Using the Hilbert transform, we extracted tomography signals from the THz time domain waveform. We captured images of 1,000 layers at each time interval, and these layer images were reconstructed using ImageJ to produce a THz 3D volumetric image. The three-dimensional image clearly displayed the shape of a bent skin phantom when no impedance-matching material was applied. In the image using water, the second reflective surface was not measured. However, in the images with applied petroleum jelly and glycerin, the second reflective surface was observed. These results indicate that terahertz impedance-matching materials are proving to be highly beneficial for tomographic imaging of biological tissues or films.

 

Fig. 9. 3-demension volumetric image of hydrogel patch on the quartz window of Fig. 5.

Download Full Size | PDF

3.3 THz impedance-matching effect in skin tissue

To verify the impedance-matching effect, we conducted experiments using skin extracted from BALB/c nude mice, as illustrated in Fig. 10. In these experiments, we applied water, glycerol, and petroleum jelly to the extracted skin and obtained THz images for each case. The THz imaging experiment was conducted 20 minutes after the tissue was excised. To prevent changes in the skin's moisture during the scanning of THz images, we packed the sample with wrap. The scanning time took 22 minutes. As shown in Fig. 10, the application of glycerol resulted in the lowest reflected image. However, due to the high absorption of THz pulses in the skin compared to the measurement range of the system, it was challenging to confirm the results of the THz 2^nd pulse. Nevertheless, these findings provide strong evidence that impedance- matching is a viable approach for real skin tissue. This technology holds significant potential for applications in THz cancer treatments that utilize high-power THz signals. It reduces the interaction between the skin tissue and THz on the skin surface, thereby decreasing heat and other potential side effects.

 

Fig. 10. Imaging of animal skin on the quartz window: (a) Photograph and (b) THz peak to peak images of (A) air, (W) water, glycerol (G), and petroleum jelly (P) (norm.: normalized amplitude).

Download Full Size | PDF

4. Conclusions

In this study, we have demonstrated the impedance-matching effect of THz imaging in skin phantoms and skin. We evaluated three materials, namely, water, glycerol, and petroleum jelly, as potential solutions for THz impedance-matching using THz reflection imaging system. It was found that glycerol exhibited excellent properties as an impedance-matching material for THz imaging on quartz windows and could be useful for practical applications. We further conducted ex-vivo experiments using animal skin tissues to verify THz impedance-matching. The results showed that THz impedance-matching techniques are useful not only for diagnosis but also THz wave therapy when utilizing high power THz sources. Moreover, to apply clinically, the study of various skin condition depending on hydration should be required.