Open Access
Add to BibSonomy
Abstract
It has been shown that silicon photonic crystal nanocavities excited by spectrally narrow light can be used to detect ionized air. Here, to increase the range of possible applications of nanocavity-based sensing, the use of broadband light is considered. We find that the use of a superluminescent diode (SLD) as an excitation source enables a more reproducible detection of ionized air. When our photonic-crystal nanocavity is exposed to ionized air, carriers are transferred to the cavity and the light emission from the cavity decreases due to free carrier absorption. Owing to the broadband light source, the resonance wavelength shifts caused by the carriers in this system (for example, due to temperature fluctuations) do not influence the emission intensity. SLD-excited cavities could be useful to determine the density of ions in air quantitatively.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Microcavities with high quality (high-Q) factors have been frequently considered in the development of sensors with high sensitivity, because the interaction between light and matter is proportional to the Q-value of the cavity and inversely proportional to the mode volume [1–4]. For example, photonic crystal (PC) nanocavities have small mode volumes on the order of one cubic wavelength [5,6]. Furthermore, various types of nanocavities with Q values ranging from 100 to 1,000,000 have been proposed [7–9] and the possibility of mass production of such nanocavities using complementary metal–oxide–semiconductor (CMOS)-compatible machinery has been demonstrated [10,11]. Therefore, high-Q nanocavities are promising for the development of sensors with high sensitivity. For instance, a significant number of studies aimed at the development of bio-sensors have considered PC cavities [12–14], and nanocavity designs suitable for sensing have been proposed [15,16]. High-Q PC nanocavities are even applicable to situations where gas needs to be detected [17–19]. Here, the comparatively small sizes of nanocavities can be an additional merit. We recently reported the detection of ionized air using a Raman silicon (Si) laser based on a high-Q PC nanocavity [20].
If a material is exposed to ionized air, electrification of the material can occur and this can induce an electrostatic discharge (ESD). ESDs can cause electronic device failures [21], explosions in chemical plants [22], rocket launch failures, and failures of communication satellites [23]. On the other hand, ionizers that provide ionized air are widely used in various industries. In particular, such ionizers are used in semiconductor manufacturing plants to remove static electricity from Si substrates [24]. Hence devices that can quantitatively determine the density of ions in air are important [25]. Compared to existing detection methods, the detection of ionized air using a high-Q nanocavity has merits with respect to size, weight, sensitivity, safety, and accessibility. Furthermore, such passive Si photonic devices can exhibit a higher resistance to ESD. However, ionized-air detection using high-Q nanocavities is still relatively unexplored. For example, it needs to be investigated how ionized air can be measured quantitatively and how these measurements can be made less sensitive to temperature fluctuations.
As shown in Fig. 1, when a Si PC slab with a cavity is exposed to ionized air, the ionized air molecules transfer their extra carriers to the slab. The transferred carriers can absorb the light in the nanocavity via free-carrier absorption (FCA) [26], which corresponds to an additional energy loss in the cavity and results in a reduction of the cavity’s Q factor. This dependence of Q can be used to measure the presence of ionized air with a Raman Si nanocavity laser: if the total loss of the cavity exceeds the Raman gain, the Raman laser oscillation stops [27,28]. It has been shown that the oscillation stop can occur for both positively and negatively ionized air [29]. However, using a Raman laser is not a superior method for quantitative detection of ionized air, although a high sensitivity is expected. Furthermore, in our previous studies [20,29], the nanocavities were excited by using a wavelength-tunable laser diode (TLD) with a narrow linewidth where the TLD wavelength, λin, was set to the resonance wavelength of the cavity, λ.
Fig. 1. Schematic of a Si PC slab exposed to a flux of negative air ions. The ionized air molecules generated at the metal tip with a potential of Vtip move along the electric field lines. The inset illustrates how the extra electrons (shown in blue) are transferred from the ionized molecules to the Si slab. These electrons cause an additional energy loss in the cavity and lead to a change of the cavity resonance wavelength λ via the carrier plasma effect and the thermo-optic effect.
It is known that this excitation scheme is sensitive to temperature fluctuations. This sensitivity to temperature fluctuations is undesirable for a sensor that detects ionized air based on the magnitude of the reduction of the emission intensity, because this approach cannot distinguish between different causes of the intensity reduction. The transferred extra carriers not only decrease the cavity’s Q factor but also induce a change of λ due to the carrier plasma effect and the thermo-optic effect [30], and simultaneously λ can also shift depending on the cavity temperature with a rate of 80 pm/K even if it is not exposed to ionized air. These wavelength variations are difficult to predict and cause a mismatch between λin and λ, which results in a variation of the emission intensity [28,30]. Such a behavior reduces the reproducibility and accuracy of the measurements.
In this paper, we report the response of a PC nanocavity to exposure to ionized air. A superluminescent diode (SLD), which emits spectrally broad light, is used as the excitation source of the cavity. We demonstrate that the emission from the cavity decreases immediately when it is exposed to a flux of ionized air, and the emission gradually recovers after the irradiation has stopped. Due to the broadband light source, the emission intensity is not affected by the resonance wavelength shifts induced by various phenomena, and thus is mainly influenced by the effect of FCA. The observed magnitude of the emission reduction is proportional to the flux of ionized air. The reproducibility of the ion-flux-induced temporal change in the emission of the SLD-excited cavity is higher compared to that of a TLD-excited cavity.
2. Sample structure
Figure 2 presents the scanning electron microscope (SEM) images of the Si PC sample used in the experiments described in the next section. The PC is defined by circular air holes (with a radius r of 106.4 nm) arranged in a triangular lattice with a lattice constant a of 404 nm. The thickness of the PC slab is t = 221.6 nm. The SiO2 with a thickness of 3 µm underneath the Si slab was removed by 48% hydrofluoric acid to form an air-bridge structure.
Fig. 2. The SEM images of the Si PC sample used in the experiment. (a) Magnified view of our L3 nanocavity structure where the two air holes labelled “A” are shifted away from the cavity center by 0.15a with respect to their original position in the PC lattice. (b) SEM image showing the position of the cavity with respect to the excitation waveguide.
For this study, we used a PC nanocavity consisting of three missing air holes in the PC, a so-called L3 nanocavity. As shown in Fig. 2(a), the two air holes labelled “A”, are shifted away from the cavity center by 0.15a on the x-axis [7]. The theoretical Q value of this cavity is Qdesign ≈ 41,400 according to a finite-difference time-domain (FDTD) calculation. The line defect located six rows away from the L3 cavity in Fig. 2(b) is the waveguide used to excite the cavity. The waveguide width was chosen to be $1.05\sqrt 3 a$ (1.05W1) in order to efficiently excite the resonant mode. The theoretical Q value that takes into account the coupling between the excitation waveguide and the cavity, is Qtotal ≈ 24,400.
The size of the Si chip used in the experiments was 800 µm × 2,000 µm in the x- and z-directions, respectively. The PC pattern was fabricated by photolithography and CMOS-compatible processes. The chip was fabricated using the same 300-mm Si-on-insulator (SOI) wafer as that used for our previous study of a Raman laser [31]. Details of the fabrication process are described in [31]. The x-direction of Fig. 2 is parallel to the [100] direction of the top Si layer, and the crystal orientation of the top Si layer in this SOI wafer is rotated by 45 degrees with respect to the substrate [32]. The orientation of the cavity does not affect the response of the cavity to ionized air, but the surface condition of Si affects the response. All experiments in this paper were performed within 10 days after the removal of a thin oxide layer on the Si surface by 1% dilute hydrofluoric acid. The sample was kept in a dry desiccator (relative humidity less than 5%) except during the experiments to prevent water from adhering to the Si surface [33].
3. Experimental setup
Figure 3(a) shows the experimental setup used to investigate the response of the L3 nanocavity to exposure to ionized air. An SLD (Thorlabs SLD1550P-A40) and a tunable CW laser (Santec TSL-510) were used as excitation sources in separate experiments to clarify the effect of the spectral width of the excitation light. The broadband light from the SLD was passed through an isolator (Thorlabs IOT-G-1550A) to reduce the amount of light that returns to the SLD, because it is important to stabilize the emission spectrum and intensity of the SLD. Then, the light was passed through a tunable optical filter (Santec OTF-350, hereafter abbreviated as OTF) to remove the wavelength components that are more than a few nanometers away from the resonance wavelength of the L3 nanocavity; the bandwidth of the OTF was set to 3 nm and the center wavelength was set to 1557 nm. Note that our nanocavity has a broad free spectral range of more than 50 nm, and hence the results presented in Section 4 will not change even if the bandwidth of the excitation light is increased significantly.
Fig. 3. (a) Experimental setup. NA: numerical aperture. NIR: near-infrared. (b) The spectrum of the SLD (black curve; FWHM: 41 nm) and the spectrum measured after the OTF (red curve; linewidth: ∼3 nm). (c) The spectrum of the TLD (linewidth: ∼4 fm).
The black curve in Fig. 3(b) shows the emission spectrum of the SLD measured using a monochromator and a liquid-nitrogen-cooled InGaAs array detector. The spectrum is relatively broad [full width at half-maximum (FWHM): 41 nm] and the peak is located at a wavelength of about 1550 nm. Considering the spectral splitting that has been observed in a previous study using SLD excitation [34], the SLD spectrum in Fig. 3(b) has a better shape owing to the use of the isolator. The red curve shows the spectrum after the OTF. Figure 3(c) shows the emission spectrum of the TLD. We consider that the observed linewidth of 104 pm in Fig. 3(c) is governed by the resolution limit of the monochromator, because the linewidth of the tunable laser is 500 kHz (∼4 fm) according to the manufacturer’s specification.
The light from the excitation source (either SLD or TLD) was modulated by a mechanical chopper at a frequency of about 1 kHz. Then, it was passed through a polarizer to obtain transverse-electric (TE) polarized light, which was focused on the facet of the excitation waveguide (by an objective lens with a numerical aperture of NA = 0.4) to pump the system.
When λin matches λ, the energy of the excitation light is efficiently transferred to the nanocavity by evanescent mode coupling. Under this condition, the intensity of the light emitted from the cavity in the vertical direction of the slab is the highest, and the energy efficiency (η) of this cavity emission (upper side of the slab) is calculated by the following equation [35]:
In our experiment, the light emitted in the y-direction was collected by another objective lens (NA = 0.7), and the emission intensity was measured by an InGaAs photodiode [Fig. 3(a); photodiode 1] and a lock-in amplifier system (NF Corporation LI5630). The emission images were measured by a near-infrared (NIR) camera (FLIR SC2500, 320 × 256 pixels). Similarly, the light transmitted through the excitation waveguide was collected by an objective lens, passed through a polarizer that efficiently transmits TE polarized light, and was detected by a photodiode [Fig. 3(a); photodiode 2].
Either positively or negatively ionized air was generated by a metal tip connected to a high-voltage power supply depending on the polarity (Green techno Co., Ltd., GT20P and GT20, respectively). The high-voltage power supplies were used to control the potential of the tip relative to the ground (Fig. 1; Vtip). The value of Vtip was manually controlled in steps of 1 kV. It is well-known that the edge shape of the metal tip is important for generating ionized air [36]. In this work, we used copper needles with a radius of 0.25 mm for both types of ions, and the needles were polished by a micro grinder to sharpen them. This resulted in both a lower threshold voltage for the corona discharge than that required in our previous studies [20,29] and an isotropic irradiation of the sample with ionized air as shown in Fig. 1. The distance between the metal tip and the Si chip was about 1 cm.
The electrons (or holes) transferred to the Si slab can undergo several processes (e.g. trapping at surface states), but mainly diffuse to the sample stage. We improved the sample stage in order to increase its conductivity: the Si chip was fixed to the copper block with a thermally conductive aluminum adhesive sheet (electrical resistivity: 0.5 Ω/cm2) and the ground wire was directly connected to the copper block and the block was fixed to the sample stage by a stainless-steel screw. These improvements make the electron escape from the cavity to the stage easier and prevent the potential difference between the Si chip and the metal tip from decreasing during the irradiation, which in turn allows us to obtain a stable corona discharge. As a result, the reproducibility of the experiments increased and the number of ions that reach the Si chip within a few seconds increased. On the other hand, when the sample was attached to an insulating block (acrylic plastic), the potential difference immediately dropped to such a low value that the corona discharge stopped.
As explained in the introduction, the transferred carriers shown in Fig. 1 can affect the nanocavity via FCA, the carrier plasma effect, and the thermo-optic effect. This FCA due to the transferred carriers constitutes an additional energy loss in the cavity, and the magnitude of this loss is defined as QFCA−1. The experimental Q value (Qexp) is lower in the case of a larger QFCA−1. The plasma effect induces a blueshift of the resonance wavelength, and the thermo-optic effect induces a redshift by a temperature increase [30]. The magnitudes of the FCA and the plasma effect are proportional to the carrier density in the cavity, while the magnitude of the thermo-optic effect depends on the rate at which the carriers release their excess energy to the cavity.
4. Experimental results
All experiments shown in this section were performed in ambient air at room temperature with a humidity of 27–36% (dew point temperature 3.3–6.7°C). The temperature of the sample stage was stabilized by a Peltier controller. Before presenting the experimental results obtained using the SLD, we prove that the irradiation of the cavity with ionized air actually induces a λ shift and we assess the degree of the reduction of the emission intensity when spectrally narrow light is used. Figure 4 shows the resonance spectra of the PC L3 nanocavity (obtained using the TLD). The measurement procedure used to determine the resonance spectra is given in Appendix A1. The open circles represent the experimental data while the solid curves are the fits of the data to a Lorentz function. The black data show the resonance spectrum of the L3 nanocavity without exposure to ionized air. The blue data show the spectrum obtained when the sample was irradiated with negatively ionized air using Vtip = −5 kV, and the red data show the spectrum for positively ionized air with Vtip = +5 kV. This magnitude of Vtip is not very high since many commercial corona-discharge ionizers use a Vtip of ±7 kV (for example, the bar type ionizers of the SJ-H series from KEYENCE) [37]. Note that the sample was continuously irradiated with ionized air during the acquisition of the latter two spectra (acquisition time: two minutes).
Fig. 4. Resonance spectrum of the L3 nanocavity without exposure to ionized air (black), that of the sample exposed to negatively ionized air (blue), and that exposed to positively ionized air (red).
In the case of no exposure, the resonance wavelength estimated from the fitting curve is λ = 1557.147 nm and the FWHM of the resonance spectrum (Δλ) is 44.9 pm. The additional small peak at 1557.36 nm is due to the Fabry–Pérot (FP) interference of the excitation waveguide. A FP interference with a separation of 0.26 nm is observed in the transmission spectrum in the range from 1556 to 1558 nm. According to the relation ${Q_{\exp }} = \lambda /\Delta \lambda $, the experimental Q value is estimated to be about 34,700. This value is higher than the calculated Qtotal of 24,400. This overestimation is mainly due to the error in Δλ caused by the FP interference, while the variation of Qin by air hole fluctuations in the fabricated cavity may contribute to the overestimation to a lesser extent [38].
When the sample was irradiated with negatively ionized air, the λ of the cavity blueshifted by 0.262 nm and the Δλ increased to 57.8 pm. The spectral shape of the blue resonance spectrum in Fig. 4 is not smooth, because the carrier density in the nanocavity was unstable during the measurement. The magnitudes of the blueshift and Δλ were different in every measurement, and we roughly estimated an Qexp of 27,000. In the case of positively ionized air, the λ of the cavity blueshifted by 0.142 nm relative to the data without exposure, and Δλ increased to 103.9 pm. Therefore, Qexp is about 15,000.
The Qexp and the emission intensity decreased in the case of exposure to ionized air due to the FCA caused by the transferred extra carriers. The additional loss due to FCA, QFCA−1, is estimated using the following relation:
An important result is that the λ of the cavity exhibited a blueshift when we exposed the sample to ionized air. Note that this shift direction is opposite to that in our previous study, and that the magnitude of the shift is about ten times larger [20]. This is explained by the fact that the shift of λ depends on the balance between the thermo-optic effect and the carrier-plasma effect. The blueshift due to the carrier-plasma effect was dominant in this study since NFCA was more than one order of magnitude larger than in our previous study [20]. Furthermore, the temperature rise of the nanocavity due to the thermo-optic effect is not linearly proportional to NFCA. It is noted that the magnitudes of the observed blueshift are more than 104 times larger than the linewidth of the TLD, and are more than three times larger than Δλ. If a TLD is used for the excitation of the cavity, such large shifts in λ reduce the detection reliability. In this perspective, the use of broadband light from the SLD should be useful.
The open circles in Fig. 5 show the temporal change of the emission intensity for SLD excitation. The gray area in the figure indicates that the sample was irradiated with negatively ionized air for 5 seconds (from 30 to 35 s) using a Vtip of −5 kV. To understand the stability improvement that is achieved by SLD excitation, the temporal change for TLD excitation is also presented (closed circles). Note that the carrier density was lower than that determined from the data in Fig. 4 because of the short duration of the exposure. The interval for acquiring data was 1.5 s. The λin of the TLD was fixed to the wavelength at which the emission intensity is maximized in the case of no exposure to ionized air, that is, the λ before irradiation. The intensity of the SLD light behind the objective lens was about 10 µW without the chopper while that of the TLD light was 5 µW. The inset shows a NIR camera image of the cavity in the case of SLD excitation. A clear emission image is observed even in the case of SLD excitation.
Fig. 5. Temporal changes of the emission intensity from the L3 cavity. The data obtained by TLD excitation and SLD excitation are shown by the closed and open circles, respectively. The inset shows the NIR camera image of the cavity during excitation by the SLD.
At the end of the exposure to ionized air, the emission intensity in the case of SLD excitation decreased by 43.6%, while that for TLD excitation decreased by 94.5%. The reduction of the emission intensity due to the FCA occurs for both excitation conditions. On the other hand, the behavior of the coupling efficiency is different: in the case of SLD excitation, the efficiency of coupling the energy propagating through the waveguide to the cavity is not influenced by a λ shift, but in the case of TLD excitation it strongly decreases by a λ shift. As shown in Fig. 3(b), the linewidth of the SLD light is much larger than the magnitude of the λ shift observed in Fig. 4. Therefore, the reduction of the emission in the case of SLD excitation is two times smaller than that for TLD excitation.
Another important characteristic is how the emission intensity recovers after irradiation. In the case of SLD excitation, the intensity recovers to the initial intensity within 10 s after the irradiation stop at the time t = 35 s. This means that, at t = 45 s, the electron density in the cavity has decreased to a low density where FCA is negligible. On the other hand, in the case of TLD excitation, the intensity recovers to only 95% of the initial value within 12 s after the stop and it does not recover to 100% even within 1 minute. This incomplete recovery is due to a slight difference in the cavity temperature compared to the initial temperature: The λ of the Si nanocavity shifts by 8 pm with a temperature change of 0.1°C. Initially, λin is tuned to λ to achieve the optimum excitation condition before irradiation, but if the cavity temperature changes by 0.1°C, the emission intensity of a cavity with Qexp = 34,700 already decreases by 11% due to the mismatch between λin and the new λ. Thus, the emission intensity in the case of TLD excitation is strongly affected even by small temperature changes. Note that the sample temperature often shifts by more than 0.1°C even without exposure to ionized air. The use of an SLD enables us to correctly assess the influence of FCA, because in the case of such a broadband light source, the decrease in the emission intensity is not affected by the λ shift due to such temperature changes.
We repeated this experiment on different days, and responses similar to that in Fig. 5 were always obtained in the case of SLD excitation. In the case of TLD excitation, the reduction of the emission intensity due to FCA was reproducible, but the degree of recovery after the irradiation stop was different every time. This variation in the degree of recovery was probably due to random fluctuations of the sample temperature. Such temperature variations also occur in the case of SLD excitation, but here the emission intensity recovered to the initial intensity (because here the λ shift has no influence on the emission intensity). These results indicate that the reliability of the nanocavity-based detection of ionized air can be improved by using SLD excitation.
A movie of the temporal change of the cavity emission in the case of SLD excitation can be seen in Visualization 1. The experimental conditions are the same as those in Fig. 5 except for the timing of the irradiation; in the movie, the sample was irradiated with ionized air between t = 5 s and 10 s. This movie was recorded using an exposure time of 10 ms, a frame rate of 25 frames per second, and a camera resolution of 320 × 256 pixels. The automatic brightness control of the camera was turned off. Figure 6 shows a few snapshots extracted from Visualization 1. The movie demonstrates that it is possible to detect the emission changes due to ionized air using a NIR camera.
Fig. 6. Representative frames extracted from Visualization 1, which was recorded by the NIR camera. The timestamps correspond to the time in Visualization 1. The sample was continuously excited by the SLD and was exposed to a flux of negatively ionized air during t = 5 to 10 s.
Figure 7 clarifies the dependence of the cavity emission reduction on Vtip in the case of SLD excitation. Figures 7(a) and (b) show the responses for negative voltages, and Figs. 7(c) and (d) show those for positive voltages. In Fig. 7(a), an emission reduction is not observed for Vtip = −3 kV since no corona discharge occurred at this voltage. On the other hand, it occurred for Vtip = +3 kV. The threshold for a corona discharge is slightly lower in the case of positive polarity [36], and thus the reduction of emission due to exposure to positive ions is larger than that for negative ions with the same |Vtip| in Fig. 7. This figure clarifies that the emission intensities for both polarities decrease as |Vtip| increases. This is because NFCA increases with |Vtip| since the flux of the ionized air (or ion wind) increases with |Vtip| [39]. The above results suggest that it may be possible to quantitatively measure the density of ions in air using SLD-excited high-Q nanocavities.
Fig. 7. (a) Time traces of the cavity emission for different negative voltages (Vtip = –7 to –3 kV). The exposure time was 5 s (starting time t = 30 s). (b) The minimum intensity recorded for each time trace in (a) as a function of Vtip. (c), (d) The corresponding results for different positive voltages.
Figure 8 shows the dependence of the cavity response on the exposure time in the case of SLD excitation and |Vtip| = 5 kV. In each measurement, the irradiation starts at the time t = 30 s. Figures 8(a) and (b) show the responses to negatively ionized air and Figs. 8(c) and (d) show those for positively ionized air. The emission intensities for both polarities decrease as the exposure time increases. This is because NFCA increases. As shown in Figs. 8(b) and (d), the minimum intensities recorded for positive ions are lower than those for negative ions if the exposure time is ≤ 6 s. This is because a larger amount of positive charges is supplied to the cavity due to the lower threshold voltage of the positive corona discharge. However, the reduction of the emission due to irradiation with positive ions saturates at longer exposure times. We performed the measurements for Fig. 8 several times on different days and found that the reproducibility of the time trace decreases as the exposure time becomes longer than 10 s. The cause of this lower reproducibility is unknown but we consider that a modification of the Si surface could be involved.
Fig. 8. (a) Time traces of the cavity emission for different exposure times in the case of SLD excitation and Vtip = –5 kV. (b) Minimum intensity recorded for each time trace in (a) as a function of the exposure time. (c), (d) The corresponding results for Vtip = +5 kV.
Finally, we show that FCA in the excitation waveguide also contributes to the change in the cavity emission. Figure 9 shows the time trace of the emission from the cavity and that of the light transmitted through the waveguide with the length of 800 µm. These time traces were simultaneously measured using the two photodiodes shown in Fig. 3. Figure 9 shows that the cavity emission decreases by 43.6% while the transmitted light decreases by 16.4%. This indicates that the FCA in the waveguide is responsible for one-third of the total decrease of the cavity emission. This also suggests that it is possible to detect ionized air by only using a PC waveguide, and the characteristics of this technique will be reported elsewhere.
Fig. 9. Comparison between the time trace of the cavity emission and that of the light transmitted through the excitation waveguide for Vtip = –5 kV. Both signals were simultaneously measured using the SLD for excitation. The open circles are the same data as shown in Fig. 5.
5. Discussion
As shown in Fig. 5, the emission intensity of an SLD-excited nanocavity is less sensitive to variations in the resonance wavelength λ, while it remains sensitive to the FCA loss. Depending on the environment (for example, in semiconductor fabrication plants or in huge data centers), the temperature of a Si chip may change by more than 1°C, which would induce a λ shift of more than 100 pm. Furthermore, although a large number of high-Q nanocavities with the same structure can be fabricated on an SOI wafer by a CMOS-compatible process, their final λ values will be scattered over a few nanometers due to manufacturing inaccuracies [40,41]. Even in such cases, an SLD can excite all nanocavities [34]. As shown in Fig. 6, the ion-flux-induced change in the emission from the SLD-excited L3 cavity can be clearly detected by a NIR camera. The use of SLD excitation can improve the reliability of such a sensing system for ionized air. It can also help us to make the detection system smaller, less expensive, and lighter, because the implementation of a mechanism that automatically tunes the λin of a TLD to the λ of a certain cavity at a given time will increase both system cost and size.
Note that the measurement system presented in Fig. 3 contains components that are not necessary for actual space-charge sensing applications. A possible simple setup for space-charge sensing is briefly described: a Si chip module with a high-Q nanocavity, an SLD, a bandpass filter, and an InGaAs photodiode are needed (size less than 10 cm × 5 cm), and all these components can be connected by optical fibers [42]. Various types of fiber-coupled SLDs, fiber-coupled photodiodes, and fiber-coupled bandpass filters are commercially available. The Si chip module needs to be installed in the location where the detection of space charges is required. The driver controlling the SLD and the computer that monitors the intensity measured by the photodiode are placed in other locations that are not exposed to space charges.
The demonstrated sensing method may be even applicable to an environment where the daily temperature change is larger than 100°C (for example, the surface of a solar panel of a space satellite) [43]: by using a bandpass filter with a bandwidth of 20 nm, the nanocavity can be pumped even in environments with such large temperature variations. It is already known that the Q value of a Si nanocavity is not sensitive to the temperature if the Si surface is kept clean. Further experiments with large temperature variations are considered an important part of future research on SLD-excited cavities.
As shown in Fig. 7, the reduction of the emission from an SLD-excited cavity is proportional to the flux of the ionized air. Therefore, it may be possible to develop a nanocavity-based sensing system that can determine the density of ions in air quantitatively. However, there are several issues that need to be solved. As shown in Figs. 4 and 8, the reproducibility of the cavity response is not high when it is exposed to the ion wind for a long time. We consider that the relationship between the Si surface and the detection sensitivity needs to be investigated.
The detection of ionized air using an SLD-excited nanocavity is sensitive to FCA. Therefore, the detection sensitivity is proportional to the Q of the cavity (since the reduction in the cavity emission by a given QFCA is more significant in cavities with a higher Q). L3 nanocavities with a Qexp larger than 1 million have already been realized [44,45], and they can be fabricated using a CMOS-compatible process. However, the emission of an SLD-excited nanocavity decreases as the Qexp increases (because Δλ decreases), and thus it is important to consider the sensitivity of the detector (e.g. the NIR camera or InGaAs photodiode) when selecting the Q value. On the other hand, the detection using the SLD-excited cavity is not sensitive to λ. By using a chip with a nanocavity array, where each nanocavity has a slightly different λ (for example, a spacing of 1 nm) [46], detection of ionized air with a high spatial resolution becomes possible.
In our device, the FCA loss in the cavity is used as the detection mechanism, and thus it should also be possible to detect other materials that increase the absorption loss in the cavity. For example, by modifying the Si surface, the detection of other gases may be possible. Furthermore, Qexp values larger than 100,000 can be maintained even if the surface condition is changed [47]. It is also important to study the influence of the type of nanocavity on the sensor characteristics. By using a nanocavity with an increased surface area, such as a slotted cavity or a nanobeam cavity, the sensitivity for the detection of charges can be increased.
6. Summary
The responses of a PC L3 nanocavity with Q = 34,700 to the exposure to ionized air in the cases of SLD and TLD excitation have been presented. Owing to the spectrally broad light from the SLD, the emission intensity is not affected by resonance wavelength shifts, which can be induced by the carrier plasma effect and the thermo-optic effect. The reproducibility of the detection of ionized air using the emission change of an SLD-excited cavity is high compared to that using the TLD. The development of a system that can determine the density of ions in air quantitatively could be possible using SLD-excited nanocavities. These results are useful for designing ionized-air detection systems.
Appendix
A1. Method for obtaining resonance spectra
The measurement method used to obtain the resonance spectra in Fig. 4 is the same as that reported previously [20,31,38]: As shown in Fig. 3, the output from the TLD was split into two beams. The weaker beam (intensity: 10% of the total TLD output) was analyzed by a high-precision wavelength meter with an absolute accuracy of ±0.3 pm (Agilent 86122A). The stronger beam (90%) was focused on the facet of the excitation waveguide to pump the device. The laser power incident on the waveguide facet was less than 5 µW without chopping. The sample used in this experiment does not have any special structure that increases the coupling efficiency. Therefore, there may be a loss of 10∼20 dB at the facet. The light intensity coupled to the silicon PC waveguide was small enough to neglect nonlinear optical effects of silicon such as two-photon absorption. To obtain the spectra shown in Fig. 4, the λin of the TLD was scanned from short to long wavelengths to cover the spectral region near λ, and the emission intensity of the nanocavity was measured as a function of λin using a lock-in amplifier system.
A2. Estimation of the carrier density in the nanocavity
The free carrier density NFCA was calculated according to the following procedure. First, the relation between QFCA and the absorption coefficient for FCA, αFCA, is considered [48]:
Here, n is the refractive index at λ. Then, αFCA is estimated from the following relation with NFCA: where σFCA is the absorption cross section for FCA. We used σFCA = 7.27 × 10−18 cm2 for electrons and σFCA = 3,67 × 10−18 cm2 for holes [49]. By using either QFCA = 121,200 or QFCA = 26,400, we obtain αFCA = 1.158 cm−1 for negative ions and αFCA = 5.315 cm−1 for positive ions from Eq. (4). By substituting these values for αFCA in Eq. (5), we obtain NFCA = 1.59 × 1017 cm−3 and NFCA = 1.45 × 1018 cm−3, respectively. The NFCA for irradiation with positive ions is 9.1 times higher than that for negative ions. However, it is not possible to conclude from these results that positive charges easier accumulate in the Si nanocavity. We consider that this large difference is mainly caused by the experimental error due to the following reasons: The Δλ estimated in Fig. 4 has a large uncertainty due to the FP interference and the unstable carrier density during the measurements. Furthermore, the threshold voltage for a corona discharge depends on the polarity and Vtip was manually controlled in steps of 1 kV. Detailed measurements using a high-voltage power supply with a digital control and an improved resolution should be performed.Funding
Japan Society for the Promotion of Science (18H01479, 21H01373); Japan Science and Technology Agency (JPMJST2032, JPMJST2111).
Acknowledgements
Yuki Takahashi was supported by a fellowship from the ICOM Foundation.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. F. Vollmer, S. Arnold, and D. Keng, “Single virus detection from the reactive shift of a whispering-gallery mode,” Proc. Natl. Acad. Sci. 105(52), 20701–20704 (2008). [CrossRef]
2. V. R. Dantham, S. Holler, V. Kolchenko, Z. Wan, and S. Arnold, “Taking whispering gallery-mode single virus detection and sizing to the limit,” Appl. Phys. Lett. 101(4), 043704 (2012). [CrossRef]
3. F. Vollmer and L. Yang, “Label-free detection with high-Q microcavities: a review of biosensing mechanisms for integrated devices,” Nanophotonics 1(3-4), 267–291 (2012). [CrossRef]
4. C. Ciminelli, F. Dell’Olio, D. Conteduca, C. M. Campanella, and M. N. Armenise, “High performance SOI microring resonator for biochemical sensing,” Opt. Laser Technol. 59, 60–67 (2014). [CrossRef]
5. O. Painter, J. Vučković, and A. Scherer, “Defect modes of a two-dimensional photonic crystal in an optically thin dielectric slab,” J. Opt. Soc. Am. B 16(2), 275–285 (1999). [CrossRef]
6. S. Noda, A. Chutinan, and M. Imada, “Trapping and emission of photons by a single defect in a photonic bandgap structure,” Nature 407(6804), 608–610 (2000). [CrossRef]
7. Y. Akahane, T. Asano, B. S. Song, and S. Noda, “High-Q photonic nanocavity in a two-dimensional photonic crystal,” Nature 425(6961), 944–947 (2003). [CrossRef]
8. B. S. Song, S. Noda, T. Asano, and Y. Akahane, “Ultra-high-Q photonic double-heterostructure nanocavity,” Nat. Mater. 4(3), 207–210 (2005). [CrossRef]
9. M. Nomura, K. Tanabe, S. Iwamoto, and Y. Arakawa, “High-Q design of semiconductor-based ultrasmall photonic crystal nanocavity,” Opt. Express 18(8), 8144–8150 (2010). [CrossRef]
10. Y. Ooka, T. Tetsumoto, A. Fushimi, W. Yoshiki, and T. Tanabe, “CMOS compatible high-Q photonic crystal nanocavity fabricated with photolithography on silicon photonic platform,” Sci. Rep. 5(1), 11312 (2015). [CrossRef]
11. K. Ashida, M. Okano, M. Ohtsuka, M. Seki, N. Yokoyama, K. Koshino, M. Mori, T. Asano, S. Noda, and Y. Takahashi, “Ultrahigh-Q photonic crystal nanocavities fabricated by CMOS process technologies,” Opt. Express 25(15), 18165–18174 (2017). [CrossRef]
12. M. G. Scullion, T. F. Krauss, and A. D. Falco, “Slotted photonic crystal sensors,” Sensors 13(3), 3675–3710 (2013). [CrossRef]
13. S. Chakravarty, W. C. Lai, Y. Zou, H. A. Drabkin, R. M. Gemmill, G. R. Simon, S. H. Chin, and R. T. Chen, “Multiplexed specific label-free detection of NCI-H358 lung cancer cell line lysates with silicon based photonic crystal microcavity biosensors,” Biosens. Bioelectron. 43, 50–55 (2013). [CrossRef]
14. T. Baba, “Photonic and iontronic sensing in GaInAsP semiconductor photonic crystal nanolasers,” Photonics 6(2), 65 (2019). [CrossRef]
15. S. H. Kwon, T. Sünner, M. Kamp, and A. Forchel, “Optimization of photonic crystal cavity for chemical sensing,” Opt. Express 16(16), 11709–11717 (2008). [CrossRef]
16. M. Erdil, Y. Ozer, and S. Kocaman, “High-Q slot-mode photonic crystal nanobeam cavity biosensor with optomechanically enhanced sensitivity,” IEEE J. Sel. Top. Quantum 25(2), 1–6 (2019). [CrossRef]
17. T. Sünner, T. Stichel, S. H. Kwon, T. W. Schlereth, S. Höfling, M. Kamp, and A. Forchel, “Photonic crystal cavity based gas sensor,” Appl. Phys. Lett. 92(26), 261112 (2008). [CrossRef]
18. Y. Chen, W. S. Fegadolli, W. M. Jones, A. Scherer, and M. Li, “Ultrasensitive gas-phase chemical sensing based on functionalized photonic crystal nanobeam cavities,” ACS Nano 8(1), 522–527 (2014). [CrossRef]
19. A. K. Goyal, H. S. Dutta, and S. Pal, “Recent advances and progress in photonic crystal-based gas sensors,” J. Phys. D: Appl. Phys. 50(20), 203001 (2017). [CrossRef]
20. S. Yasuda, Y. Takahashi, T. Asano, Y. Saito, K. Kikunaga, D. Yamashita, S. Noda, and Y. Takahashi, “Detection of negatively ionized air by using a Raman silicon nanocavity laser,” Opt. Express 29(11), 16228–16240 (2021). [CrossRef]
21. S. H. Voldman, ESD: Circuits and devices (John Willey & Sons, Ltd, 2015).
22. M. Glor, “Electrostatic ignition hazards in the process industry,” J. Electrostat. 63(6-10), 447–453 (2005). [CrossRef]
23. D. M. Harland and R. D. Lorenz, Space systems failures (Plaxis Publishing Ltd, 2005).
24. H. Imazono, T. Terashige, and K. Okano, “The double jet ionizer for ULSI manufacturing processes,” IEEE Trans. Semicond. Manufact. 15(2), 189–193 (2002). [CrossRef]
25. S. Plong-Ngooluam, N. Jindapetch, P. Wounchoum, and D. Sompongse, “A finite element analysis of multiple ion receiving plates for ionizer balance monitoring,” J. Electrostat. 86, 50–58 (2017). [CrossRef]
26. T. Liang and H. Tsang, “Role of free carriers from two-photon absorption in Raman amplification in silicon-on-insulator waveguides,” Appl. Phys. Lett. 84(15), 2745–2747 (2004). [CrossRef]
27. Y. Takahashi, Y. Inui, M. Chihara, T. Asano, R. Terawaki, and S. Noda, “A micrometre-scale Raman silicon laser with a microwatt threshold,” Nature 498(7455), 470–474 (2013). [CrossRef]
28. D. Yamashita, T. Asano, S. Noda, and Y. Takahashi, “Strongly asymmetric wavelength dependence of optical gain in nanocavity-based Raman silicon lasers,” Optica 5(10), 1256–1263 (2018). [CrossRef]
29. Y. Takahashi, S. Yasuda, M. Fujimoto, T. Asano, K. Kikunaga, S. Noda, and Y. Takahashi, “Oscillation interruption of a Raman silicon nanocavity laser induced by positively ionized-air irradiation,” in Proc. of Nonlinear Optics (2021), paper NF2B.1.
30. D. Yamashita, Y. Takahashi, J. Kurihara, T. Asano, and S. Noda, “Lasing dynamics of optically-pumped ultralow-threshold Raman silicon nanocavity lasers,” Phys. Rev. Appl. 10(2), 024039 (2018). [CrossRef]
31. T. Yasuda, M. Okano, M. Ohtsuka, M. Seki, N. Yokoyama, and Y. Takahashi, “Raman silicon laser based on a nanocavity fabricated by photolithography,” OSA Continuum 3(4), 814–823 (2020). [CrossRef]
32. Y. Yamauchi, M. Okano, H. Shishido, S. Noda, and Y. Takahashi, “Implementing a Raman silicon nanocavity laser for integrated optical circuits by using a (100) SOI wafer with a 45-degree-rotated silicon top layer,” OSA Continuum 2(7), 2098–2112 (2019). [CrossRef]
33. H. Sekoguchi, Y. Takahashi, T. Asano, and S. Noda, “Photonic crystal nanocavity with a Q-factor of ∼9 million,” Opt. Express 22(1), 916–924 (2014). [CrossRef]
34. R. Shiozaki, T. Ito, and Y. Takahashi, “Utilizing broadband light from a superluminescent diode for excitation of photonic crystal high-Q nanocavities,” J. Lightwave Technol. 37(10), 2458–2466 (2019). [CrossRef]
35. C. Manolatou, M. J. Khan, S. Fan, P. R. Villeneuve, H. A. Haus, and J. D. Joannopoulos, “Coupling of modes analysis of resonant channel add-drop filters,” IEEE J. Quantum Electron. 35(9), 1322–1331 (1999). [CrossRef]
36. J. S. Chang, P. A. Lawless, and T. Yamamoto, “Corona discharge processes,” IEEE Trans. Plasma Sci. 19(6), 1152–1166 (1991). [CrossRef]
37. H. J. Kim, B. Han, C. G. Woo, and Y. J. Kim, “Ozone emission and electrical characteristics of ionizers with different electrode materials, numbers, and diameters,” IEEE Trans. on Ind. Appl. 53(1), 459–465 (2017). [CrossRef]
38. T. Kawakatsu, T. Asano, S. Noda, and Y. Takahashi, “Sub-100-nW-threshold Raman silicon laser designed by a machine-learning method that optimizes the product of the cavity Q-factors,” Opt. Express 29(11), 17053–17068 (2021). [CrossRef]
39. M. Rickard, D. D. Rankin, F. Weinberg, and F. Carleton, “Characterization of ionic wind velocity,” J. Electrostat. 63(6-10), 711–716 (2005). [CrossRef]
40. H. Hagino, Y. Takahashi, Y. Tanaka, T. Asano, and S. Noda, “Effects of fluctuation in air hole radii and positions on optical characteristics in photonic crystal heterostructure nanocavities,” Phys. Rev. B 79(8), 085112 (2009). [CrossRef]
41. K. Ashida, M. Okano, M. Ohtsuka, M. Seki, N. Yokoyama, K. Koshino, K. Yamada, and Y. Takahashi, “Photonic crystal nanocavities with an average Q factor of 1.9 million fabricated on a 300-mm-wide SOI wafer using a CMOS-compatible process,” J. Lightwave Technol. 36(20), 4774–4782 (2018). [CrossRef]
42. Y. Saito, T. Asano, S. Noda, and Y. Takahashi, “Design characteristics of a Raman silicon nanocavity laser for efficient emission of light into an adjacent waveguide,” in Proc. of Nonlinear Optics (2021), paper NM1A.6.
43. K. Toyoda, H. Masui, T. Muranaka, M. Cho, T. Urabe, T. Miura, S. Kawakita, Y. Gonohe, and T. Kikuchi, “ESD ground test of solar array coupons for a greenhouse gases observing satellite in PEO,” IEEE T. Plasma Sci. 36(5), 2413–2424 (2008). [CrossRef]
44. T. Nakamura, Y. Takahashi, T. Asano, and S. Noda, “Improvement in the quality factors for photonic crystal nanocavities via visualization of the leaky components,” Opt. Express 24(9), 954–9549 (2016). [CrossRef]
45. K. Maeno, Y. Takahashi, T. Nakamura, T. Asano, and S. Noda, “Analysis of high-Q photonic crystal L3 nanocavities designed by visualization of the leaky components,” Opt. Express 25(1), 367–376 (2017). [CrossRef]
46. Y. Takahashi, T. Asano, D. Yamashita, and S. Noda, “Ultra-compact 32-channel drop filter with 100 GHz spacing,” Opt. Express 22(4), 4692–4698 (2014). [CrossRef]
47. T. Ito, K. Ashida, K. Kinoshita, R. Moriya, T. Machida, K. Maeno, T. Endo, K. Yamada, M. Okano, and Y. Takahashi, “Nondetrimental surface modification of ultrahigh-Q photonic crystal silicon nanocavities,” in Proc. of Conference on Lasers and Electro-Optics Pacific Rim (CLEO-PR) (2018), paper W4D.3.
48. T. Asano, B.-S. Song, and S. Noda, “Analysis of the experimental Q factors (∼1 million) of photonic crystal nanocavities,” Opt. Express 14(5), 1996–2002 (2006). [CrossRef]
49. J. C. Sturm and C. M. Reaves, “Silicon temperature measurement by infrared absorption: Fundamental processes and doping effects,” IEEE T. Electron Dev. 39(1), 81–88 (1992). [CrossRef]
