Notice: Wiley Online Library will be unavailable on Saturday 27th February from 09:00-14:00 GMT / 04:00-09:00 EST / 17:00-22:00 SGT for essential maintenance. Apologies for the inconvenience.
 A dual-beam intersection apparatus for microwave breakdown plasma spectroscopy (MBS) has been developed and tested for real-time, simultaneous, and in situ monitoring of the trace gases in the air. The dual microwave beams, which are supplied by a 9.4G Hz pulse magnetron (feed power: 0–80 kW, pulse width: 5 μs, pulse repetition frequency: 10 Hz) using a power divider, are focused and superposed on a center of a reaction chamber through two dielectric lenses. The superposed electric field brings about the gaseous breakdown phenomena on the focusing point, where the photons emitted from the proper gas components can be detected and assigned using two-dimensional wavelength/time-resolved photoemission spectroscopy. The preliminary experiment has demonstrated that the emission intensities of breakdown plasma strongly depended on the phase difference between two microwave beams. Also, almost all the plasma spectra of the dual-beam system were more intensified than that of the single-beam system developed previously by us, and the relative continuum background decreased remarkably as well. For the 81 ppm of CCl2F2 gas mixed with the model atmosphere, the Cl atomic line of 837.6 nm (4p4D − 4s4P5/2) could clearly be detected at room temperature in total pressure of 10Torr. Present experiments have proved that the beam controllability of the dual-beam system is superior to that of the single-beam system to realize the appropriate breakdown conditions.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 Clear understanding of the spatiotemporal evolution of the concentrations of air composition is critically important in elucidating quantitative links between them and the atmospheric environmental problems such as air pollution in urban areas, meso-scale climate changes, and global warming effect [Brasseur et al., 1999; Finlayson-Pitts and Pitts, 2000]. To accomplish this object, the accurate, real time, remote, and in situ measurement of many chemically active species in the air is of quite importance both qualitatively and quantitatively. In the field of radio science, one of the ambitious attempts for this aim is the application of microwave technology to the detection of atmospheric trace gas components. One salient feature of microwaves (especially, X-band region) is its good permeability through the air even in bad weather conditions such as rain, fog, and mist.
 Recently, we have developed a new measurement system for detection of the atmospheric compositions, using a high-intensity pulsed microwave single beam breakdown method combined with a wavelength/time-resolved plasma spectroscopy technique [Suto et al., 2001; Iinuma et al., 2002]. The basic concept of this system is as follows: A pulsed microwave single beam is focused by a dielectric lens and irradiated onto a region of interest in a test chamber (or open space), where the breakdown plasma is generated. The photons emitted from the plasma can be measured to identify a proper gas component of present concern and quantify its concentration. It is noted that this Microwave Breakdown Spectroscopy System (abbreviated to MBS) is basically capable of satisfying four requirements above mentioned, and has a potential to be useful for qualitative and quantitative detections of active species in air. By applying this single beam MBS technique to model atmosphere, we have developed the basic MBS technology and demonstrated that ppm orders of halide impurities (Cl and F) will possibly be detectable.
 However, the single beam technique has some drawbacks which make its practical usage rather inconvenient and difficult: (1) Because of the geometrically fixed form of the dielectric lens, it is essentially impossible to focus the microwave beam on an arbitrary point of interest in open space. (2) Irregular or accidental breakdown at unintended points as well as undesirable areas often occurs during the transmission of highly intensified microwave power. (3) The continuum background of the photo emission spectra (background noise) increases with increasing the impurity metals and chemicals. The background photons may probably be emitted from the surfaces of the wave-guide as well as of the dielectric materials. To overcome these shortcomings of single-beam MBS and to realize more flexible remote diagnostics system, we have developed, more recently, a dual-beam intersection MBS system, the original idea of which was proposed by Gurevich .
 Basic idea of radio wave application to geophysical research might go back to the conception of artificial remedy of ionosphere proposed by Bailey in 1937 [Gurevich, 1980]. Based on this idea, the Gurevich group in the Russian Federation and several U.S. research groups have proposed various theoretical models of microwave breakdown phenomena in the air and developed the related radio science technology [Armstrong et al., 1990; Gurevich et al., 1997]. Recently, the Gurevich group has published a comprehensive review paper which deals with microwave breakdown theory, several experimental research works, and its application to the formation of the artificially ionized region as well as to the remedy of ozone depletion in the stratosphere [Gurevich et al., 2000]. The present dual-beam intersection method is founded on two intriguing ideas proposed by three research groups cited in this paper: the beam focusing technique [Askar'yan et al., 1992] and the superposing technique of two-microwave beams [Vikharev et al., 1984, 1988; Armstrong et al., 1990]. This paper deals with the technical details of design and construction of the dual-beam intersection apparatus along with its preliminary spectroscopy measurement data.
2. Comparison With Other Measurement Methods
 Currently, several measurement methods of the atmosphere are widely used for qualitative and quantitative detections of various kinds of chemical compositions as well as particulate matters. We, at first, compare their characteristics with that of MBS technique. ICP-AES (Inductively coupled plasma atomic emission spectroscopy), ICP-MS (Inductively coupled plasma mass spectroscopy), GC (Gas Chromatography), and NDIR (Non-Dispersive InfraRed) are standard measurement techniques for detecting inorganic elements, organic compositions, and carbon monoxide, etc. in the air [Finlayson-Pitts and Pitts, 2000]. However, all these techniques do not satisfy above four requirements (accurate, real time, remote, and in situ) simultaneously, because these techniques basically need sampling/batch system for air composition analysis. Also, LIDER (Light DEtection and Ranging) [Brasseur et al., 1999] and LBS (Laser Breakdown spectroscopy) [Dudragne et al., 1998] techniques are standard remote sensing and in situ techniques for detection of trace gas components, particulate matters, explosives, and chemical warfare agents. While these techniques are widely used in meteorological/artificial satellite observation as well as in defense/military activity, the influence of weather condition such as moisture and clouds in the target space must be considered. The major features of the MBS method and other competitive systems are shown in Table 1.
Table 1. Comparison of Air Pollutants Detection Techniquesa
A, applicable; D, difficult; N, nonapplicable. ICP AES, Inductively Coupled Plasma Atomic Emission Spectroscopy; ICP MS, Inductively Coupled Plasma Mass Spectroscopy; LIDER, Light Detection and Ranging; LBS, Laser Breakdown Spectroscopy; MBS, Microwave Breakdown Spectroscopy.
In situ measurement
Detection of molecular compositions
3. Dual-Beam Intersection MBS System
 The basic idea for design and construction of dual-beam intersection MBS system is as follows: (1) Two-microwave beams (beams 1 and 2) are irradiated from a microwave power source and superposed on a target point. (2) By controlling the direction of two microwave beams, we can appropriately select the target point, where breakdown plasma including the ionized/excited gases of present concern will be generated. (3) The breakdown plasma can perfectly be separated from boundary surfaces such as those of waveguide and dielectric materials to avoid the unfavorable contamination of the target plasma.
 Based on this designing guideline, we have realized a small laboratory system consisting of the beam intersection circuit, the reaction chamber, and the plasma emission spectroscopy system. The followings are the technical detail of the dual-beam intersection MBS apparatus, its breakdown characteristics including the effect of phase difference between two beams on breakdown conditions, and preliminary spectroscopy data for model air compositions as well as for chlorine atom.
 The block diagram of the MBS apparatus is shown in Figure 1. Technical details of the microwave generating system, the power monitoring system, the optical measurement system, and the pressure measurement system are listed in Table 2. Before designing the MBS apparatus, we have assembled a simple test bench equipped with a low power microwave source and a power sensor for performing a preliminary experiment to test the microwave focusing characteristics of two dielectric lenses. We have measured the spatial distribution profile of the microwave radiation with/without lens using this test bench.
Table 2. Microwave Generating System, Power Monitoring System, Spectroscopy System, and Gas Pressure Measurement System
Microwave generating system
EEV: M5042S Magnetron
9.4 GHz (X-band)
Peak output power
External TTL(0–5V) from delay/pulse generator
WRJ-10: 22.9 mm × 10.2 mm
Microwave power monitoring system
Nihon-koshuha: Crystal mount DM SMP-61H
−74 dB (−44 dB directional coupler and −30 dB attenuator)
Tektronix: TDS540 digitizing oscilloscope
Bunkoukeiki: CT-25C Czerny-Turner monochromator
30 μmm/2 mm
Hamamatsu: M2548 temporal disperser
Hamamatsu: C3140 CCD camera
8.8 mm × 6.6 mm
CCD pixel size
510 × 492 pixels
Hamamatsu: C2280 temporal analyzer
Data acquisition mode
External TTL(0–5V) from delay/pulse generator
Partial pressure gauge
H2, He, H2O, N2 + CO, O2, Ar, CO2
Stanford Research System: DG535
MKS: 127AA-0010A (10Torr head)
4.1. Experimental Apparatus
 A high-power microwave beam (maximum output power: 80 kW) was generated by magnetron (EEV:M5042S) and injected into the power divider and divided into two low-power beams through a phase shifter. The power ratio of two beams was fixed at 1:1 in the present experiment. A dielectric lens was set in front of each phone antenna to focus the microwave beam. The dielectric lens consists of a pair of two convex lenses made of Teflon (the relative dielectric constant is 2.10 at 1 MHz). Two convex lenses are 112 mm and 93 mm in diameter and the maximum thickness of 30 mm and 15 mm, respectively. To design the lenses, the Maxwell equations with curvature boundary conditions were solved by FDTD numerical method. After passing through the two dielectric lenses, these two beams were focused and superposed on the center of the reaction chamber which was made of SUS-304 stainless steel (volume: 76.7 liters; i.d.: 312.4 mm; length: 1000 mm). For emission spectroscopy, the microwave breakdown plasma of target gas was generated on the center of the chamber. The microwave frequency was kept at 9.4 GHz, with the pulse length of 5 μs and the repetition rate of 10 Hz. The intersection angle between the beam axes was fixed to be 90 degrees. The pulsed input power was properly controlled by a variable attenuator. Both the feeding power (input power) and the reflected power were monitored by a crystal mount connected to a digitizing oscilloscope. The divided power was controlled in the range of 15–40 kW in each beam. The effect of phase difference between two beams on breakdown condition is an intriguing present research subject. To perform this experiment, we carefully controlled the phase difference in the range between 0 and 360 degree using a phase-shifter device.
 In the MBS method the photoemission spectroscopy is essentially important to detect the chemical compositions contained in the target gas. The emitted photons from the plasma were collected and transmitted into a 250 mm focal length spectrometer (diffraction grating: 300 grooves/mm). The temporal evolutions of emission lines of interest have been measured by a streak camera with 0–12 μs duration time. The two-dimensional wavelength/time-resolved photo emission spectra have been stored by a CCD camera (charge-coupled-device camera) with 510 pixels width (wavelength) and 492 pixels height (duration time). All the spectroscopy equipment has been synchronously controlled by TTL-level trigger signals generated from a delay/pulse generator. The calibration of the wavelength of emission spectra was performed using a calibrated mercury, neon, and argon lamp source for corresponding spectral region in order to assign the wavelength to the corresponding pixels of CCD camera. For the emission intensity calibration, a calibrated tungsten lamp source was utilized. These calibrations allow us to determine the relative concentration of two chemicals by comparing the emission intensity between two different emission lines.
 A typical example of two-dimensional wavelength/time-resolved photo emission spectra measured for pure nitrogen gas (99.9999%) is shown in Figure 2. For this kind of raw MBS photoemission data, two different analyses are possible: (1) the analysis of wavelength distribution (assignment of chemical compositions), and (2) the analysis of temporal evolution for a specific wavelength (information on their reactive process). Main purpose of the present study is to test the usability of beam intersection MBS method for determination of chemical compositions by detecting their emission lines in the target plasma. Therefore, analysis 1 has priority over analysis 2, and the emission spectra were precisely assigned by referring the several compiled reference data [Striganov and Sventitskii, 1968; Pearse and Gaydon, 1976; Lofthus and Krupenie, 1977; Randzig and Smirnov, 1985] (NIST Atomic Spectra Database Lines Data, available at http://physics.nist.gov/cgi-bin/AtData/).
4.2. Experimental Conditions
 The model atmospheric gas-mixtures containing nitrogen (99.9999% of purity), oxygen (99.99%), and CCl2F2 (99.9%) were prepared for MBS emission spectroscopy in the present experiment. The mixing ratio of nitrogen gas and oxygen gas was fixed at 4:1. The gas pressure was measured by using a Baratron pressure gauge of 10Torr head. The total gas pressure was kept constant at 10Torr, while the partial pressure of CCl2F2 was changed from 81 ppm to 3vol %. The model gas was directly introduced into the test chamber without any purification technique (e.g., dehumidification, etc.).
4.3. Beam Profile Measurement System
 The spatial profiles of focused microwave beam power were measured by using a simple test bench. The test bench consists of a microwave oscillator, a lens mount, a movable open waveguide receiver, a microwave power meter, and a movable carriage. It is possible to change the distance between the dielectric lens and the receiver on a movable carriage by moving it toward x-direction and y-direction. We can, therefore, measure the spatial distribution on and off the axis of the direction of electromagnetic field. The reflected power through a circulator was also monitored simultaneously. A low power FET oscillator (SPC: 14T007: 9.4 GHz, 12 mW) was used in order to measure the power profile in open space, and the low power microwave propagated through the dielectric lens was detected by the open waveguide receiver (detection probe). A transducer at the back of the probe was connected with a microwave power meter (HP: 438A) through a power sensor (HP: 8481A). The measurement was made in the ambient air at room temperature (25°C).
5. Results and Discussion
5.1. Measurement of Beam Profile
5.1.1. Axial Distribution of Microwave Power Intensity
 The distribution of microwave power intensity on the beam axis was measured as a function of the distance from the horn aperture. The axial distribution of relative power intensity with/without dielectric lens is shown in Figure 3. It is found that the focused power intensity is about four times larger than the nonfocused power intensity at the maximum beam power position (∼220 mm from the horn aperture). The numerical analysis of the Maxwell equations with curvature boundary condition gives us a theoretical curve (a dashed curve in Figure 3) of the focused power intensity (“with lens” in Figure 3). The agreement between the measured intensity profile and the calculated values is fairly well.
5.1.2. Two-Dimensional Distribution of Microwave Power Intensity
 The focusing efficiency of dielectric lens must be strongly influenced by its geometric skew in shape as well as its dielectric constant. In order to check this efficiency, we have examined the homogeneity of two dimensional radiation pattern of the focused TE10 mode perpendicular to the beam axis. The measurement was made at the beam focusing point, 240 mm, apart from the horn aperture. The measured spatial distributions of relative power intensity attenuated without/with lens-system are shown in Figures 4a and 4b, respectively. The power intensity is depicted as a contour map on which lines of constant attenuated power are plotted with horizontal and vertical axes. Figure 4b (with lens system) shows that the axial symmetry of radiation pattern appears to be fairly well, indicating that the microwave beam is efficiently focused on the circular area of 2 cm in radius (−3 dB attenuation region).
5.1.3. Determination of Input-Power Range for Breakdown
 Next, we have measured the magnitude of threshold breakdown power of the superposed beams for effective production of breakdown plasmas. The relation between the input power and the electromagnetic field intensity per each beam at the focusing point is shown in Figure 5. The threshold input power for breakdown was determined to be 15 kW/each beam, corresponding to the minimum electromagnetic filed intensity of 0.6 kV/cm. The upper limit of working range (breakdown region) was set to be 40 kW because of the maximum power rating of the microwave generator. The corresponding field intensity limit was 1.0 kV/cm. The simplest estimation (neglecting the phase effect) of the superposed electromagnetic filed intensity is, therefore, ranging between 1.2 kV/cm and 2.0 kV/cm.
where Ethr is the critical electric field for breakdown in air, C0 and C1 are the constants as functions of ω/ν; ω is the angular frequency of the microwave, and ν is the characteristic electron collision frequency. For the case of ω ≒ ν, the magnitude of C0 ≒ 4 ∼ 6, and C1 ≒ 1.0 are usually employed, and p is the air pressure in Torr.
 Under the present experimental condition (ω = 9.4 GHz, p = 10Torr), equations (1)–(3) obtain the estimated values of Eopt ranging between 1.5–2.3 kV/cm, the mean value of 1.9 kV/cm. Intriguingly, the doubled value of present minimum field, 2 × 0.6 = 1.2 kV/cm, is smaller than the estimated mean value, 1.9 kV/cm. This discrepancy may be due to the presence of a few kindling electrons created by one beam ignition, together with the effect of phase difference for dual-beams. This favorable low-power breakdown technique is very effective and of use for open space MBS method to detect more realistic air compositions.
5.2. Background Emission Spectra of Model Air Breakdown Plasma
 Comparison of background emission spectra between the dual-beam method and the single-beam method may be meaningful for us to select the effective technique for detection of specific air pollutants. Figure 6a shows the background emission spectra of model air breakdown plasma produced by the dual-beam method in the visible region (330 nm–850 nm). The spectra were obtained by integrating the number of photons with respect to duration time of 0–12 μs in each microwave pulse. Also, Figure 6b shows the background spectra of the single-beam method [Suto et al., 2001] measured under the same experimental condition using the same data analysis. A salient feature of the spectra in Figure 6a is its simple emission lines, compared with Figure 6b. In Figure 6a, only several lines are observable in the wavelength region of 330–430 nm, which are typical N2 Second Positive Bands (C3Πu → B3Πg: 337.1, 357.7, 380.5, 405.9, and 434.4 nm). All these lines have the same excitation energy of 11.2 eV. Atomic lines of N and O ranging between 650–850 nm are relatively faint. On the other hand, in Figure 6b we clearly observe several atomic lines of N (742.57, 744.44, 747.04, 822.54, and 824.46 nm) and O (777.2, 777.4, and 844.6 nm) together with the N2 Second Positive Bands. The excitation energies of all these lines distribute over 10–13 eV.
 Despite that most of the emission lines detected by the dual-beam method were the same as those of the single-beam method, the intensity of atomic lines both of nitrogen and oxygen of the former, in ranging 650–850 nm, was decreased significantly. This may be attributed to the decrease in the mean electron energy of the plasma as well as in the electron number density during the power injection period of 0–5 μs. It is expected that, for the dual-beam method, the unfavorable situations such as the generation of secondary electrons from the surface of the metal parts or dielectric materials will be reduced noticeably. In addition, because of the relative decrease in the number of high-energy electrons, the excitation of the atoms accompanied by the molecular dissociation will be suppressed significantly. Also, the molecular excitation process of rotation and vibration by high-energy electrons might not proceed too much. This is a clear indication of the favorable beam-power controllability of present intersection method. For all these reasons, the dual-beam MBS is expected to be a promising technique for us to do the unambiguous detection of various air pollutants.
5.3. Effect of Phase Difference on Breakdown Condition
 The phase difference between two microwave beams seems to be critical for effective production of breakdown plasma. To examine this effect appropriately, we have detected the intensity of nitrogen emission line, 337.1 nm, by varying the phase difference ranging between 0–360 degrees. The measured correlation between the phase difference and the emission intensity was shown in Figure 7. We find that the emission intensity strongly depends on the phase difference, and the maximum value reaches about three times as large as the minimum value under the same input power condition. This suggests that the phase tuning will become a useful technique to control the emission intensity of target gas plasma appropriately.
 This phase dependence of the emission intensity can simply be explained by using the following expression of the superimposed power for two-beams intersection. Suppose that there are two beams of microwaves, A and B, with phase difference θ;
Then, the superimposed beam E is expressed as
Using equation (5), we can calculate the mean microwave power density, Pave, as
where Z is the characteristic impedance (376.7 Ω). It may be quite natural to deduce that the number density of excited particles, which will emit the proper photons, is proportional to Pave. Then, equation (6) can reasonably explain the overall tendency of measured emission-phase dependence shown in Figure 7. However, in this measurement, the phase difference at the maximum emission was tuned to ∼315 degrees, which did not agree with the complete in-phase position (360 degrees). This discrepancy may be caused by the unfavorable reflection of two microwave beams on the inner-wall of reaction chamber.
 Another noteworthy point in Figure 7 is that the emission of photons of low input-power microwave breakdown is more intensive than that of high input-power breakdown. To examine this fact more precisely, we have measured the time dependence of the emitted photons of 337.1 nm line by varying the input-power. Figure 8 shows the emission intensity with respect to the elapsed time. These curves make it clear that the 337.1 nm line by low input-power breakdown has much longer emission tail than that of high input-power breakdown. We can attribute it to the reflection phenomena of the input-power in the generated plasmas. Especially, at higher input-power experiment, a significantly large part of the injected power will reflect from rather high-density and high-temperature plasma.
 The reflection process can be evaluated with the help of two plasma parameters: the mean electron temperature and the electron number density. Two-lines-emission method [Armstrong et al., 1990; Gurevich et al., 1997] was employed for estimation of mean electron temperature. In the present experiment, the comparison between the N2 337.1 nm line and the N2+ 391.4 nm line gave us the mean temperature of ∼3.5 eV. Also, the electron density was evaluated from the reflected microwave power and the skin depth of the breakdown plasma. The skin depth δ is expressed as [Roth, 1995]
where n is the electron number density, c is the light velocity, and ωpe is the plasma angular frequency (or plasma cutoff frequency).
 The approximate radius of the breakdown plasma was estimated by a video monitor camera to be ∼2.5 cm. When we assume that the skin depth is nearly equal to the plasma radius, the electron number density is evaluated from equations (7)–(9) to be ∼1011 cm−3. Therefore, it appears that the microwave power cannot permeate the breakdown plasma, because the evaluated electron number density (∼1011 cm−3) is almost the same order of the plasma cutoff density. Under the present power feeding condition, the ratio of the reflected power to the total power would increase with increasing in the input power. This may be the reason of low emission intensity of photons at high input-power MBS experiment.
5.4. Application to Detect Chlorine Atom
 Halogen gases and their many derivatives are still receiving our attention for the issues of environmental atmosphere. The attainment of low background continuum level is one of the favorable characteristics of the dual-beam intersection MBS method for us to detect such kind of trace halogenated chemicals in air. Specifically, in open air, the dramatic decrease of the emission line of oxygen atom, 844.6 nm, will make it possible to detect the 837.6 nm emission line of the chlorine atom generated from the dissociation of CCl2F2 molecules. Preparing the mixture of nitrogen/oxygen/CFC-12 gas (total pressure of 10Torr, vol% = 80: 20: 81 × 10−6) using gas chromatography, we have performed the MBS experiment in the wavelength range of 830 nm to 850 nm. The detected emission lines of chlorine atom (837.6 nm) and oxygen atom (844.6 nm) are shown in Figure 9.
 For halogenated chemicals, our present detectable limit was estimated to be little higher than the sub-ppm level that is our goal target. To raise the detectable limit, one simple but useful technique is to increase the repetition rate of microwave pulse for the accumulation of the spectrum data. Also, the employment of more highly optically sensitive devices will be recommended for better MBS method that is now underway. Based upon the probability analysis of photon counting method [Iinuma et al., 2002], we have estimated the desired number of pulsed microwave shots was about 1.4 × 105 shots for 1 ppm detection. When the pulse repetition rate will be changed from the present rate of 10 Hz to 1 kHz, all the data acquisition/analysis procedure for detection of 1 ppm target gas will be completed within 3 min.
 The authors would like to thank N. Sasaki at Yamagata University for fruitful discussion on the phase effect. The authors wish to acknowledge K. Furukawa, T. Takahashi, K. Komatsu, T. Nagaya, and T. Akama for helping us to assemble the experimental apparatus. The authors also thank T. Matsui of Hitachi, Ltd., and T. Ozaki of High Energy Accelerator Research Organization for valuable suggestions on measurement techniques. This research was supported partly by the Grant-in-Aid for Scientific Research in Priority Areas grant 14048204.