Open Access
Add to BibSonomy
Abstract
A ground-based system for measuring tropospheric OH radical based on laser-induced fluorescence (AIOFM-LIF) was developed in this work. In this system, ambient air is expanded through a 0.4 mm nozzle to low pressure in a detection chamber, where OH radical is irradiated by the 308 nm laser pulse at a repetition rate of 8.5 kHz. Then, the resultant fluorescence corresponding to the transition at 308 nm is detected using gated photon counting. The AIOFM-LIF system was integrated into a mobile observing platform for the field observation following the series of laboratory characterization. A portable standard OH radical source by water photolysis-ozone actinometry was established and optimized for accurate system calibration. The factors affecting the system sensitivity were quantified. It was shown that the ultimate system sensitivity is 9.9 × 10−8 cps (molecules cm−3)−1 mw−1; the minimum detection limits are (1.84 ± 0.26) × 105 cm−3 and (3.69 ± 0.52) × 105 cm−3 at night and noon, respectively; and the whole error of AIOFM-LIF system is about 16%. Then, the system was deployed in Shenzhen, China, during the “A comprehensive STudy of the Ozone foRmation Mechanism in Shenzhen” (STORM) campaign. Valid OH radical concentrations for 31 days were obtained, and the peak of the daily average concentration was 6.6 × 106 cm−3 around 12:00. And a high correlation (R2 = 0.77) between OH and j(O1D) was also observed in this field campaign. The relationship between OH concentration and NOx was attentively discussed. The deployment of AIOFM-LIF system in STORM campaign has demonstrated its capability of measuring tropospheric OH radical with high sensitivity and accuracy in a polluted environment.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
OH radical is the most important oxidant in the atmosphere, and plays a central role in the removal of most trace gases in the tropospheric atmosphere [1,2]. The global and annual mean concentrations of OH in the troposphere are around 1 × 106 cm−3 according to the analyses of several decades of CH3CCl3 concentration data [3–5]. According to its high reactivity, OH radical atmospheric chemical life is extremely short, about 1 s in clean areas and even shorter in polluted urban areas and forests, and its daytime peak concentration is basically less than 1 ppt, which imposes extremely stringent requirements for the sensitivity of the measurement instrument [6–8].
Several groups have built their own OH radical measuring instruments. The main methods for measuring OH radical are laser-induced fluorescence (LIF) [9–14], differential optical absorption spectroscopy (DOAS) [15,16], chemical ionization mass spectroscopy (CIMS) [17], and superconducting-magnet-based Faraday Rotation Spectrometer (FRS) [18]. Since its first application of the method of fluorescence assay by gas expansion for the determination of OH radicals [11,12], LIF has been demonstrated as an effective tool for detecting atmospheric OH radicals owing to its high sensitivity, good selectivity, and low detection limit. In this method, ambient air is drawn into a low-pressure cell directly through the gas expansion, and laser around 308 nm is utilized to excite OH radical from ground state to the excited state, resulting in the 308 nm fluorescence. It is noteworthy that LIF is not an absolute measurement technology, so the validity of calibration determines the accuracy of OH radical detection.
In the past years, the LIF-based instruments for in situ measurement of OH radical have been used in numerous ground-based field campaigns. Early observations were carried out in clean and moderately polluted regions [2], and the established mechanisms explain well the observed OH concentrations mainly in clean areas and unpolluted rural areas at low VOC reactivity [19–23]. However, recent observations showed that the modeling concentration of OH is even two times as high as the measured one where atmospheric NOx and VOC concentrations were both high in polluted regions [24]. It seems that current models tend to underestimate the measured OH by up to an order of magnitude at low NOx and several ppb isoprene in forested regions, such as North America [25,26], Amazonian rainforest [27] and the tropical forest of Borneo [28,29].
In China, air quality has deteriorated over the past two decades owing to the rapid growth in economy and the increasing urbanization. Owing to the emission form transportation and various industries to support more than one billion people, air pollution in China is characterized by high VOCs, high NOx, and heavy haze. Moreover, pollution sources have become increasingly diversified, secondary pollution has intensified, and the mechanism of radical chemical reactions have become more complicated. Several field campaigns have been carried out on measuring OH radical in China during the past 15 years [30–33]. However, in those studies, it seems that the modeled OH radical concentration is often underpredicted because of the complexity of atmospheric chemistry in urban environments, which indicate the potential unknown reaction mechanisms may be present in polluted environments. Thus, a series of field observations of OH radicals with high accuracy is requisite [34].
In this paper, the development of a ground-based system for the tropospheric OH radical measurement (AIOFM-LIF) has been reported. A portable standard OH radical source by water photolysis-ozone actinometry was established for calibrating the system. The key factors, such as OH radical sampling, laser wavelength stability, solar background deduction, system sensitivity, and error analysis, has been characterized and discussed in detail. At last, the first deployment of the AIOFM-LIF system in the “A comprehensive STudy of the Ozone foRmation Mechanism in Shenzhen” (STORM) campaign in Guangdong, China, has been demonstrated and the corresponding data of OH radical and j(O1D) was compared and analyzed.
2. Instrument description
A set of a ground-based system for the tropospheric OH radical measurement based on LIF (AIOFM-LIF) was developed at Anhui Institute of Optics Fine Mechanics, Chinese Academy of Sciences. As shown in Fig. 1, the AIOFM-LIF system mainly consisted of four parts, laser source, OH detection cell, reference cell, control and measurement unit. For the purpose of being deployed under different ground-based field conditions, a software based on LabVIEW environment has also been developed to link each module of the system for the automatic control and the accurate measurement of OH radical. Each module will be introduced in detail in the following.
2.1. Laser source
The laser source is a combination system. First, a tunable dye laser (Sirah laser, Spectra Physics, USA) pumped by A 532 nm laser produced by a Nd:YVO4 laser (Navigator YHP40, Spectra Physics, USA), will generate a 616 nm laser beam, then 616 nm laser light can be converted to 308 nm laser by frequency doubling BBO (Barium borate) crystals. The 308 nm laser output pulse has a pulse width of 20 ns and spectral linewidth of ~0.9 pm with a high repetition rate of 8.5 kHz. Maximum average laser power produced by this laser system at 308 nm is approximately 120 mW. Regarding to the problem of laser wavelength drift according to the fluctuation of the temperature, laser source is usually placed in a shield for the isolation of the laser in a stable environment, so the inner temperature is not dramatically affected by external temperature.
2.2. OH detection cell
The central part of OH detection cell is mainly composed of a 100 mm × 100 mm × 100 mm aluminum block with laser arms, sampling part, and fluorescence detection unit, which were all anodized and blackened for stray light reduction. Ambient air is drawn through a small nozzle and expanded into the OH detection cell which is pumped by a vacuum pump (XDS35i, Edwards, UK), where a fast collimated gas beam formed. Given the influence of pressure on fluorescence lifetime and the pumping speed of the vacuum pump, the pressure in the cell is set to approximately 400 Pa.
The laser beam is introduced into OH detection cell by an optical fiber, and then expanded and collimated to a diameter of approximately 8 mm. The gas beam intersected with the laser beam and OH radical is excited in this excitation region resulting in the generation of fluorescence corresponding to the transition at 308 nm. The fluorescence lifetime of OH radical is about 152ns under dry air condition. Fluorescence collection unit includes a front lens group similar to the Kohler illumination system, and a rear lens group similar to an infinite objective system, which is assembled perpendicular to the two beams and close to the OH radical excitation region. A narrowband interference filter centered on 307.5 nm with a bandwidth of 5 nm is mounted between the fluorescence collection unit and the detector, which has a rejection of 10−5 throughout a wavelength range of 200–1200 nm. To enhance the efficiency of fluorescence collection, a coated mirror is mounted on the opposite side of the detector.
It is well known that the fluorescence was extremely weak compared with the laser scattering in the cell at low pressure due to Rayleigh and Mie scattering. Thus, detector needs to be in a normal closed state and only to be turned on to detect fluorescence after the laser shots, to prevent interference from laser stray light. Therefore, the Detector should be capable of fast opening and closing function besides high performance, high sensitivity and low noise. A home-made gated PMT was developed in our laboratory [35] and assembled to the AIOFM-LIF system. Given that the afterpulse of the PMT interferes with the fluorescence measurement and limits the improvement of sensitivity of the AIOFM-LIF system, the MCP (Photek Ltd., UK) was selected. It can be used with multiple microchannel plates to gain up to 3 × 107 with pulse fast rise time of 50 ps and pulse FWHM of 80 ps. The fluorescence signal was sent to a gated photon counter (PMS400A, Becker & Hickl GmbH, Germany) firstly for measurement, then integrated into a computer. A photodiode was set at the exit of the fluorescent cell to measure laser energy in real time for the real-time correction of fluorescence.
Solar stray light is the largest noise of the LIF system in field observation and has strong randomicity. It also contributes shot noise and thus leads to the worsening of the signal-to-noise ratio of the system. To accurately measure and subtract solar stray light, a solar stray light measurement channel was added 2.5 µs after each fluorescence measurement, and the measurement time lasted 3 µs. Meanwhile, a strip baffle was placed 20 cm away from the nozzle to shield direct sunlight.
It is believed that the size and the shape of nozzle play an important role in system sensitivity for measuring OH radical. Different kinds of nozzles, with different shapes and various diameters ranging for from 0.4 mm to 1.0 mm have been used by all the current LIF groups (as summarized in Table 1). The effects of different sizes of sampling nozzles on solar stray light counting have also been considered in this work. Two nozzles (Beam Dynamics, USA) with different sizes have been investigated here. The test has shown that qualitatively, under the same cell pressure, the larger the nozzle size is, the stronger the stray light of the sun and the larger the shot noise as shown in Fig. 2 (a), while at the same time, small nozzle leads to the low fluorescence as shown in Fig. 2 (b). So, it seems that to reduce noise is in contradiction to improv sensitivity of the system. To compromise this contradiction and in view of the pumping force of the vacuum pump, the nozzle with 0.4 mm diameter was selected.
Table 1. Summary about the nozzle and detection limit
Fig. 2 (a) During daytime, the black dot line represents the solar stray light count corresponding to 1.0mm nozzle, the red dot line corresponding to 0.4mm nozzle, and the blue dot line represents the dark noise of MCP; (b)under the same OH radical concentration and intracavity pressure, the black dot line represents the fluorescence counts corresponding to 1.0mm nozzle, and the red dot line corresponding to 0.4mm nozzle.
2.3. Reference cell
As the Full Width at Half Maximum (FWHM) of OH radical spectrum line is only about 1.1 pm, a small temperature shift in the wavelength of the laser, such as 2pm/°C, will considerably disrupt the detection of OH radical using LIF technique. Therefore, a reference cell was established in the AIOFM-LIF system for identifying OH radical resonance lines and monitoring wavelength changes in real time. A synthetic gas with a certain humidity was pumped into the reference cell by a vacuum pump, and the pressure in the cell was controlled by valve and maintained at about 400 Pa (same as OH detection cell) . High concentrations of OH radical was produced by the thermal decomposition of water vapor at high temperatures. The power of thermal decomposition was 14 W. A small branch output of the 308 nm laser was introduced into reference cell to excite OH radical. Fluorescence was collected and sensed by PMT perpendicular to the laser beam and the water pyrolysis region. A photodiode was set at the exit of reference cell for measurement of laser energy synchronously in order to correct the fluorescence in real time. Resonance wavelength of OH radical was determined through two times of scans. Both of the wavelength scanning steps were 0.1 pm/s. The excitation spectra of OH radical were obtained by the first scan. During the second scan, when fluorescence reached 95% of its maximum, the scan stopped, then the corresponding wavelength was fixed as the wavelength of OH resonance excitation. Figure 3 shows the process of wavelength scanning for determining the resonance excitation line. The fluctuation of the fluorescence count in reference cell was only 1.2% at wavelength resonance within 15 minutes, which indicates the stability of the concentration of OH radicals by thermal decomposition in reference cell. The laser wavelength deviation from the resonant wavelength can be detected by real-time monitoring of the wavelength in reference cell.
It is well known that the drift of laser wavelength is inevitable in case of drastic changes in ambient temperature. So in this work, if fluorescence in reference cell deviates from more than 90% of maximum value, the laser will scan again and re-determine the resonance wavelength of OH radicals. OH radical fluorescence in OH detection cell must be corrected synchronously by fluorescence of the reference cell. While fluorescence in reference cell deviates less than 90%, fluorescence in OH detection cell needs to be corrected by the reference cell to eliminate the change of the wavelength in a small range synchronously. However, we noticed that the parameters (for example, temperature) of reference cell and OH detection cell were slightly different. Reference cell uses pyrolysis method to produce OH radical, and its detection temperature is higher than that of OH detection cell. Under low pressure, the Doppler broadening is prevalent, and the fluorescence lineshape is conformed to the Gauss distribution and its half-height line width is only related to ambient temperature. It means that FWHM in two cells is different, and if the fluorescence in OH detection cell is corrected directly by the fluorescence in reference cell will bring errors. Therefore, the influence of this difference must to be studied carefully.
To avoid the influence of pressure broadening, the pressure in reference cell is kept similar to that in OH detection cell The wavelength scan and the corresponding Gauss fitting both in OH detection cell and reference cell are shown in Figs. 4(a) and 4(b) respectively. It can be seen that FWHM of reference cell was 0.2 pm more than that of OH detection cell. While the lineshape difference between two cells is not a constant, and as shown in Fig. 4(c) the difference coefficient becomes large with the wavelength deviating from the resonance wavelength. So if the fluorescence in OH detection cell corrected only by the reference cell will bring a large error, and correction procedures for fluorescence measurement of OH radical need to consider the dependence of the difference coefficient on wavelength. As shown in Fig. 4(d), for the case when the fluorescence in reference cell deviates less than 90%, the uncertainty of fluorescence measurement in OH detection cell is reduced to only 4% when the difference coefficients are considered during correction. This ensures that OH radical fluorescence measurement is not affected by laser wavelength due to temperature in the case of small wavelength deviations. Besides, the factors such as inner pressure, pumping speed and stability of pyrolysis source are also important sources of uncertainty. After the optimization of signal-to-noise ratio (SNR) of fluorescence measurement in reference cell, the uncertainty of fluorescence measurement in reference cell was less than 2%.
Fig. 4 (a) the result of scan and its normalization in OH detection cell; (b) The result of scan and its normalization in reference cell; (c) the lineshape difference between the two cells. (d) The fluorescence in OH detection cell corrected by difference coefficient and fluorescence in reference cell.
2.4. Measurement and control unit
The timing of the AIOFM-LIF system is controlled by a digital delay generator (DG645, Stanford Research Systems, USA), which is triggered externally by the synchronous output of laser source. In a cycle, the MCP is usually turned on after the end of the laser pulse with delay of 80ns, and lasts for 550 ns during the fluorescence detection to avoid the possible overloading of MCP and the decrease of sensitivity in long-term field measurements. Then, the photon counter is turned on following the MCP function. As the signal detected by the photon counter consists of the fluorescence and solar scattering noise, the MCP was turned on again at 2.5 µs after the laser pulse for around 3 µs in order to collect the solar scattering signal separately from the laser scattering signal.
A software, which includes laser control, pressure control, photon acquisition, laser power acquisition, and other functions, is programmed in LabView for self-operation and measurement. The whole measurement process is divided into three stages: (a) scanning and determining laser wavelength, (b) on-resonance measurement, and (c) off-resonance measurement. The resonance excitation wavelength of OH radical is first examined and fixed by two laser scans as aforementioned during (a) stage. Then OH radical is measured on OH resonance line during (b) stage. Finally, the background signal is measured on off-resonance lines during (c) stage. These three stages were run in every cycle for 10 min. The OH radical concentration can be calculated in real time based on the software.
3. OH radical standard source
As aforemention, LIF is not an absolute measurement technology, so the accuracy of calibration determines the accuracy of OH measurement. The response of AIOFM-LIF system is calibrated by water photolysis-ozone actinometry. OH radical standard source consists of a bubbler, a flow tube, and a light absorption cell as shown in Fig. 5. The bubbler is used to get the appropriate humidified air by bubbling water vapor into the zero air. In the flow tube, the humidified air flows under laminar conditions at the flow rate of 20 L/min. The 184.9 nm radiation emitted by a mercury lamp is collimated through the lens and intersects perpendicularly with gas flow in the flow tube forming photolysis regions. A light absorption cell is added to adjusted to 185nm light intensity at the light inlet, whereas narrowband filter and phototube are added at the light outlet to measure 185nm light intensity. The generation of OH and HO2 radicals and O3 is represented in R1–R4 chemical reactions.
As the luminous flux in photolysis region is difficult to accurately measure, ozone is used to characterize the light flux. The concentration of ozone produced in the photolysis region is linearly correlated with the185 nm light intensity. The relationship between light intensity and ozone concentration can be established by adjusting Mercury lamp intensity. According to reactions R1-R4, OH concentration can be deduced by chemical radiometry according to the Eq. (1). Mercury lamp intensity is adjusted to establish.
where factor 1/2 is the ratio of quantum yields of OH and O3, σH2O and σO2 are the corresponding absorption cross-sections of H2O and O2 at 185 nm, [O2], [H2O] and [O3] represent the concentrations of oxygen, water vapor and ozone respectively, and P represents the flow profile of ozone in flow tube. Among them, [H2O], [O3], and σO2 are the main sources of errors affecting the accuracy of OH radical standard source, accurate measurements of them are requisite to improve the standard source.3.1. Ozone measurement
Ozone concentration in the flow tube is obtained by measuring the concentration of NO2 using a home-made Cavity Ring Down Spectrometer (CRDS) [42]. Sufficiently high concentration of NO is fully mixed with O3 in the sampled gas to immediately produce NO2, then NO2 can be detected directly by CRDS. The detection limit of CRDS can be optimized to 16 ppt by extending the measurement time to 30 s. The ozone concentration of the exhaust gas from the flow tube was 0.90 ± 0.02 ppb. Under laminar flow conditions, ozone concentration has a distinct gradient distribution in the flow tube. The flow profile of ozone factor P, defined as the ratio of ozone concentration in center of the flow tube to that of the exhaust gas from the flow tube, were also measured and determined as 1.81 ± 0.05 for this flow tube.
3.2. Water vapor measurement
Water vapor concentration was measured by high precision Humidity & Temperature Probes (HC2-SH, Rotronic, Switzerland) before the mixed gas enters flow tube. Considering the uncertainty of measuring water vapor with this tool, we calibrate the Probes with 911-0016 ammonia (NH3, H2O) analyzer (Los Gatos Research, USA). The measurement error of water vapor concentration by ammonia (NH3, H2O) analyzer is at 100 ppm. Generally, the range of water vapor concentration in the actual field calibration is from 8000 ppm to 16000 ppm, so the maximum error of water vapor measurement is estimated to be less than 3%.
3.3. Oxygen absorption section
According to Eq. (1), σH2O and σO2 at 185 nm are necessary to obtain [OH]. A value of (7.1 ± 0.2) × 10−20 cm2 is acquired for σH2O [43]. For σO2, the deviation of the measurement from different studies is large up to 60% [44,45]. Oxygen absorption cross-section is a function of oxygen column concentration and mercury lamp performance, and the emission spectrum of the 184.9 nm band from mercury pen-ray lamps varies from lamp to lamp and even becomes complicated by overlapping with several features in the Schumann-Runge band of the O2 spectrum. The oxygen absorption cross section has also measured again in this work for the OH radical standard source based on Lambert's law as shown in Fig. 5. Under the typical atmospheric oxygen concentration at 21% (corresponding oxygen column concentration 9.59 × 1018 cm−2), the accurate value of σO2 = (1.25 ± 0.08) × 10−20 cm2 was obtained for the OH radical standard source.
3.4. Overall error of OH radical standard source
OH radical concentration can be determined according to Eq. (1) and can be controlled within a certain range by adjusting water vapor concentration in the flow tube. Under the condition of 0.8% water vapor concentration, the concentration of OH radical is 1.48 × 109 cm−3. As shown in Table 2, the overall uncertainty of the OH radical standard source is about 10%.
Table 2. Overall uncertainty in OH calibration
4. Sensitivity of the AIOFM-LIF system
The fluorescence counting S (cps) measured in OH detection cell is proportional to laser power P (mw) and OH radical (cm−3) as shown in Eq. (2):
where C represents the detection sensitivity of LIF system (cps (molecules cm−3)−1 mw−1), which is related to the specific parameters of the instrument and the spectral parameters of OH radicals. Accurately detecting the sensitivity of AIOFM-LIF system is key to precisely measure atmospheric OH radical concentration. The accuracy of sensitivity is also affected by water vapor quenching and laser power. For this reason, the influence of relevant factors on the sensitivity of the system has been studied separately.4.1. Quenching coefficient for water on excited OH radical
The fluorescence lifetime of OH radicals greatly influences the detection sensitivity of the system. The quenching effect of O2, N2, and H2O on OH fluorescence is evident in the atmosphere, and particularly the concentration of H2O in ambient air varies greatly, which is different from those of N2 and O2 in space-time distribution [38]. In principle, the fluorescence lifetime of OH radicals can be determined as Eq. (3). Given their corresponding quenching coefficients as summarized in Table 3, it seems that at 298 K and 1% water vapor concentration, the quenching effect of water on OH fluorescence is eight times as much as that of N2 and 20 times as much as that of O2. Fluorescence lifetime is reduced by approximately 11% with water vapor increasing from 0% to 1% (see Fig. 6(b)).
Table 3. Quenching rate coefficients (k = A T1/2 – B T3/2 + C) [46,47]
Fig. 6 (a) The black dot line represents the sensitivity varies with water vapor concentration; the red dot line represents the results after correcting the effect of water vapor quenching, the unit is cps (molecules cm−3)−1 mw−1; (b) the coefficient of Water Vapor Quenching. (c)OH radical fluorescence intensity curve corresponding to the different OH radical concentration
As shown in Fig. 6(a), the sensitivity of AIOFM-LIF system decreases by 15% with water vapor concentration increasing from 0.5% to1.6%, but after being corrected by the influence of water vapor quenching, the decrease is less than 7%. The OH radical fluorescence intensity corresponding to the OH radical concentration in the range of 3.8 × 108 cm−3 - 2.1 × 109 cm−3 is plotted in Fig. 6(c), and an excellent linearity and insignificant intercept can be obtained by Least-squares analysis, which indicates that the effect of water vapor on sensitivity can be almost deducted.
4.2. Effect of laser energy on sensitivity
The influence of laser energy on sensitivity has also been investigated for the AIOFM-LIF system. It shows that the sensitivity appears to decrease with laser energy, as shown in Fig. 7. A least squares fitting to the data shows about 1% decrease in sensitivity with 1 mw decreasing of laser power. Fortunately, this decline is linear, and sensitivity can be corrected using laser energy with a correction error about 6%.
4.3. Detection limit
Under the laser power (P) of 20 mw, the typical sensitivity C of the system is at 9.9 × 10−8 cps (molecules cm−3)−1 mw−1. Wherein, the uncertainty of standard OH radical is 10% (at 0.8% water vapor), that of OH radical fluorescence count due to wavelength correction is 4%, that of sensitivity corrected by laser energy is 6%, and that of sensitivity corrected by water vapor is 7%. Considering all of these factors, total error of the sensitivity for AIOFM-LIF system is valued at 14.2%.
The detection limit of the system is calculated according to Eq. (4):
where S/N is signal-to-noise ratio at the detection limit (S/N = 2), m and n are the numbers of fluorescence and background measurement (m = n = 1), respectively, t is the integration time of 60 s. Under the typical condition, the background signal rate SBG (cps) is about 4 cps, comprising laser scattering noise, MCP dark noise (1 cps), and solar scattering noise (3 cps at around noon). According to the evaluation, the detection limit of AIOFM-LIF system for OH concentration of are (1.84 ± 0.26) × 105 cm−3 and (3.69 ± 0.52) × 105 cm−3 under night and noon conditions, respectively.5. Field observation
Following the series characterization, AIOFM-LIF system is integrated into a mobile observing platform for field observation. OH detection cell, measurement unit, control unit, and gas supply unit are integrated into a 1.1m × 1.5m × 1.5m box. Considering that A high temperature will lead to a big white noise of MCP and poor detection limit, the temperature insider the box was maintained at approximately 291 K. During “A comprehensive STudy of the Ozone foRmation Mechanism in Shenzhen” (STORM) campaign (30/09/2018-11/11/2018), AIOFM-LIF system was deployed in Peking University Shenzhen Graduate School, Shenzhen, China. The box is set on the rooftop of C building as shown in Fig. 8. Laser source including reference cell was situated inside the laboratory of the fourth floor of the building. A piece of 8 m-long optical fiber, 8 m-long vacuum tubing, and some 10 m-long electric cables were used to assemble the instruments in the laboratory and AIOFM-LIF system on the rooftop.
During the campaign, AIOFM-LIF system has been calibrated for 22 times in ambient environment to ensure data validity. The calibrations were conducted usually at night, and the maximum of the calibration deviations was less than 7%. This deviation was brought to an evaluation of the performance for AIOFM-LIF system, and the ultimate error of this system is at 16%.
It seems that AIOFM-LIF system performed very well under high humidity condition in the Southern China. Valid OH radical data of 31 days were obtained except for the rainy days. It is believed that the most important source of OH radical results from the photolysis of ozone at wavelengths shorter than 340 nm. This source yields an oxygen atom (O1D) and subsequently reacts with ambient water vapor to form two OH radicals. So, ozone photolysis frequency j(O1D) was measured by a spectroradiometer. The time series of OH radical and j(O1D) for this campaign are shown in Fig. 9. The highest concentration of OH radical even reached 1.6 × 107 cm−3. Diurnal variation of the 31-day data was averaged, and the mean diurnal cycles of OH together with j(O1D) averaged over 15min time intervals are depicted in Fig. 10. The average peak concentration of OH radical is 6.6 × 106 cm−3, which appeared at about 12:00 in daytime. A good agreement on diurnal trends of OH and j(O1D) has been observed.
Fig. 9 Time series of OH and j(O1D) during the STORM campaign from September 30 to October 31: experimental values of OH denoted by red symbols, corresponding j(O1D) is given by blue lines
The consistent trend can be seen more clearly if we look the detail of the time series of OH radical and j(O1D) in one day, for example, on October 30 as shown in Fig. 11(a). A comparison of the measured OH-j(O1D) dependencies for the “STORM” campaign is shown in Fig. 11(b). A linear fit to the OH-j(O1D) relationship yields a slope of 2.44 × 1011 cm−3 s and an offset of 6.9 × 105 cm−3 with the correlation (R2) up to 0.77. The slope for this campaign is significantly lower than the previous observations in China (PRD, Beijing, Wangdu) [30–33], while it is comparable to the observations at the Meteorological Observatory Hohenpeissenberg [20]. Generally, it is presumed that high NOx and high VOCs may be the main factors that lead to a smaller slope.
Fig. 11 (a) The time series of OH radical and JO1D on October 30; (b) The correlation between the concentration of OH radicals and J(O1D) rate during “STORM” campaign.
During the STORM campaign, the NOx mixing ratio varied from 0.3 to 125 ppb. Data were selected for j(O1D) >1 × 10−5 s−1 to investigate NOx dependence, and each OH measurement is normalized to its corresponding j(O1D) measurement by being multiplied by an average j(O1D) value (1.9 × 10−5 s−1). These normalized OHJ-norm data were averaged over equal log (NOx, in ppb) intervals of 0.1. The resulting data are shown in Fig. 12.
It seems that the correlation between the OHJ-norm data with NOx is more complicated than anticipated, which maybe results from the different oxidation mechanism. Under low NOx environment, it is believed that OHJ-norm declined due to decreasing OH production rate from HO2 + NO reaction. Under high NOx concentration, OHJ-norm also decreased, which can be explained by increasing importance of OH loss reaction with NO2. This nonlinear dependency was also predicted by box-model calculations and has been observed in other work [19,21,48]. Different behaviors from previous model predicted that the OHJ-norm can reach a maximum almost at approximately 8 ppb NOx, which is in good agreement with the measured results here. High concentrations of CO may also contribute the decrease of OH concentration, which generally results in the shift of OH concentration maximum to high NOx values. However, more work are need to learn this effect.
6. Conclusion
A mobile AIOFM-LIF system has been developed for OH radical field observation. By a series of optimization and the setup of OH radical standard source with high accuracy, the accuracy of OH radical measurement for the system has been improved greatly. The dependence of the system detection sensitivity has been quantified at various ambient water vapor and laser power. The ultimate system detection sensitivity is 9.9 × 10−8 cps (molecules cm−3)−1 mw−1, and the detection limits are (1.84 ± 0.26) × 105 cm−3 and (3.69 ± 0.52) × 105 cm−3 at night and noon, respectively. The ultimate error of this system is approximately 16%. For the field observation, self-operation and self-measurement for this system are achieved by laser wavelength controlling, sensitive detection and precise correction of the fluorescence, accurate deduce of the solar scattering noise and so on.
AIOFM-LIF system has been deployed in Shenzhen, China, during the STORM campaign for the first time. The time series of OH radical for 31 days has been obtained, and the peak of OH radical daily average concentration at 6.6 × 106 cm−3 appears at around 12:00. An analysis between measured OH and j(O1D) showed a high correlation (R2 = 0.77), and OH radical concentration is strongly influenced by solar irradiance. The nonlinear dependence of OH concentration on NOx has been observed during this campaign. The first deployment of AIOFM-LIF system has demonstrated its suitability for measuring OH radicals with high sensitivity and accuracy in a polluted environment, and its potential application in more field observation is anticipated.
Funding
National Key R&D Program of China (2017YFC0209401, 2017YFC0209403); National Natural Science Foundation of China (91644107, 61805257, and 61575206); Strategic Priority Research Program of the Chinese Academy of Sciences (XDB05040200).
References
1. L. Jaeglé, D. J. Jacob, W. H. Brune, I. Faloona, D. Tan, B. G. Heikes, Y. Kondo, G. W. Sachse, B. Anderson, G. L. Gregory, H. B. Singh, R. Pueschel, G. Ferry, D. R. Blake, and R. E. Shetter, “Photochemistry of HOx in the upper troposphere at northern midlatitudes,” J. Geophys. Res. Atmos. 105(D3), 3877–3892 (2000). [CrossRef]
2. D. E. Heard and M. J. Pilling, “Measurement of OH and HO2 in the troposphere,” Chem. Rev. 103(12), 5163–5198 (2003). [CrossRef] [PubMed]
3. A. Hofzumahaus, F. Rohrer, K. Lu, B. Bohn, T. Brauers, C. C. Chang, H. Fuchs, F. Holland, K. Kita, Y. Kondo, X. Li, S. Lou, M. Shao, L. Zeng, A. Wahner, and Y. Zhang, “Amplified Trace Gas Removal in the Troposphere,” Science 324(5935), 1702–1704 (2009). [CrossRef] [PubMed]
4. D. Kubistin, H. Harder, M. Martinez, M. Rudolf, R. Sander, H. Bozem, G. Eerdekens, H. Fischer, C. Gurk, T. Klüpfel, R. Königstedt, U. Parchatka, C. L. Schiller, A. Stickler, D. Taraborrelli, J. Williams, and J. Lelieveld, “Hydroxyl radicals in the tropical troposphere over the Suriname rainforest: comparison of measurements with the box model MECCA,” Atmos. Chem. Phys. 10(19), 9705–9728 (2010). [CrossRef]
5. L. Xue, R. Gu, T. Wang, X. Wang, S. Saunders, D. Blake, P. K. K. Louie, C. W. Y. Luk, I. Simpson, Z. Xu, Z. Wang, Y. Gao, S. Lee, A. Mellouki, and W. Wang, “Oxidative capacity and radical chemistry in the polluted atmosphere of Hong Kong and Pearl River Delta region: analysis of a severe photochemical smog episode,” Atmos. Chem. Phys. 16(15), 9891–9903 (2016). [CrossRef]
6. J. Mao, X. Ren, L. Zhang, D. M. Van Duin, R. C. Cohen, J. H. Park, A. H. Goldstein, F. Paulot, M. R. Beaver, J. D. Crounse, P. O. Wennberg, J. P. DiGangi, S. B. Henry, F. N. Keutsch, C. Park, G. W. Schade, G. M. Wolfe, J. A. Thornton, and W. H. Brune, “Insights into hydroxyl measurements and atmospheric oxidation in a California forest,” Atmos. Chem. Phys. 12(17), 8009–8020 (2012). [CrossRef]
7. D. Stone, L. K. Whalley, and D. E. Heard, “Tropospheric OH and HO2 radicals: field measurements and model comparisons,” Chem. Soc. Rev. 41(19), 6348–6404 (2012). [CrossRef] [PubMed]
8. K. Hens, A. Novelli, M. Martinez, J. Auld, R. Axinte, B. Bohn, H. Fischer, P. Keronen, D. Kubistin, A. C. Nölscher, R. Oswald, P. Paasonen, T. Petäjä, E. Regelin, R. Sander, V. Sinha, M. Sipilä, D. Taraborrelli, C. Tatum Ernest, J. Williams, J. Lelieveld, and H. Harder, “Observation and modelling of HOx radicals in a boreal forest,” Atmos. Chem. Phys. 14(16), 8723–8747 (2014). [CrossRef]
9. P. S. Stevens, J. H. Mather, and W. H. Brune, “Measurement of Tropospheric OH and HO2 by laser-induced fluorescence at low pressure,” J. Geophys. Res. Atmos. 99(D2), 3543–3557 (1994). [CrossRef]
10. F. Holland, M. Hessling, and A. Hofzumahaus, “In situ measurement of tropospheric OH radicals by laser-induced fluorescence - a description of the KFA instrument,” J. Atmos. Sci. 52(19), 3393–3401 (1995). [CrossRef]
11. M. R. Heal, D. E. Heard, M. J. Pilling, and B. J. Whitaker, “On the development and validation of FAGE for local measurement of tropospheric OH and HO2,” J. Atmos. Sci. 52(19), 3428–3441 (1995). [CrossRef]
12. T. M. Hard, R. J. Obrien, C. Y. Chan, and A. A. Mehrabzadeh, “Tropospheric Free Radical Determination by FAGE,” Environ. Sci. Technol. 18, 768–777 (1984). [CrossRef]
13. Y. Kanaya, Y. Sadanaga, J. Hirokawa, Y. Kajii, and H. Akimoto, “Development of a ground-based LIF instrument for measuring HOx radicals: Instrumentation and calibrations,” J. Atmos. Chem. 38(1), 73–110 (2001). [CrossRef]
14. P. O. Wennberg, R. C. Cohen, N. L. Hazen, L. B. Lapson, N. T. Allen, T. F. Hanisco, J. F. Oliver, N. W. Lanham, J. N. Demusz, and J. G. Anderson, “Aircraft-borne, laser-induced fluorescence instrument for the in situ detection of hydroxyl and hydroperoxyl radicals,” Rev. Sci. Instrum. 65(6), 1858–1876 (1994). [CrossRef]
15. H. P. Dorn, U. Brandenburger, T. Brauers, M. Hausmann, and D. H. Ehhalt, “In-situ detection of tropospheric OH radicals by folded long-path laser absorption. Results from the POPCORN field campaign in August 1994,” Geophys. Res. Lett. 23(18), 2537–2540 (1996). [CrossRef]
16. H. Fuchs, H. P. Dorn, M. Bachner, B. Bohn, T. Brauers, S. Gomm, A. Hofzumahaus, F. Holland, S. Nehr, F. Rohrer, R. Tillmann, and A. Wahner, “Comparison of OH concentration measurements by DOAS and LIF during SAPHIR chamber experiments at high OH reactivity and low NO concentration,” Atmos. Meas. Tech. 5(7), 1611–1626 (2012). [CrossRef]
17. R. L. Mauldin, C. A. Cantrell, M. Zondlo, E. Kosciuch, F. L. Eisele, G. Chen, D. Davis, R. Weber, J. Crawford, D. Blake, A. Bandy, and D. Thornton, “Highlights of OH, H2SO4, and methane sulfonic acid measurements made aboard the NASA P-3B during transport and chemical evolution over the Pacific,” J. Geophys. Res. Atmos. 108(D20), 8796 (2003). [CrossRef]
18. W. Zhao, B. Fang, X. Lin, Y. Gai, W. Zhang, W. Chen, Z. Chen, H. Zhang, and W. Chen, “superconducting-magnet-based faraday rotation spectrometer for real time in situ measurement of OH radicals at 106 molecule/cm3 level in an atmospheric simulation chamber,” Anal. Chem. 90(6), 3958–3964 (2018). [CrossRef] [PubMed]
19. D. H. Ehhalt, “Photooxidation of trace gases in the troposphere,” Phys. Chem. Chem. Phys. 1(24), 5401–5408 (1999). [CrossRef]
20. F. Rohrer and H. Berresheim, “Strong correlation between levels of tropospheric hydroxyl radicals and solar ultraviolet radiation,” Nature 442(7099), 184–187 (2006). [CrossRef] [PubMed]
21. F. Holland, A. Hofzumahaus, R. Schafer, A. Kraus, and H. W. Patz, “Measurements of OH and HO2 radical concentrations and photolysis frequencies during BERLIOZ,” J. Geophys. Res. Atmos. 108(D4), 8246 (2003). [CrossRef]
22. Y. Kanaya, A. Hofzumahaus, H.-P. Dorn, T. Brauers, H. Fuchs, F. Holland, F. Rohrer, B. Bohn, R. Tillmann, R. Wegener, A. Wahner, Y. Kajii, K. Miyamoto, S. Nishida, K. Watanabe, A. Yoshino, D. Kubistin, M. Martinez, M. Rudolf, H. Harder, H. Berresheim, T. Elste, C. Plass-Dülmer, G. Stange, J. Kleffmann, Y. Elshorbany, and U. Schurath, “Comparisons of observed and modeled OH and HO2 concentrations during the ambient measurement period of the HOx Comp field campaign,” Atmos. Chem. Phys. 12(5), 2567–2585 (2012). [CrossRef]
23. F. L. Eisele, G. H. Mount, D. Tanner, A. Jefferson, R. Shetter, J. W. Harder, and E. J. Williams, “Understanding the production and interconversion of the hydroxyl radical during the Tropospheric OH Photochemistry Experiment,” J. Geophys. Res. Atmos. 102(D5), 6457–6465 (1997). [CrossRef]
24. J. Mao, X. Ren, S. Chen, W. H. Brune, Z. Chen, M. Martinez, H. Harder, B. Lefer, B. Rappenglück, J. Flynn, and M. Leuchner, “Atmospheric oxidation capacity in the summer of Houston 2006: comparison with summer measurements in other metropolitan studies,” Atmos. Environ. 44(33), 4107–4115 (2010). [CrossRef]
25. D. Tan, I. Faloona, J. B. Simpas, W. Brune, J. Olson, J. Crawford, M. Avery, G. Sachse, S. Vay, S. Sandholm, H. W. Guan, T. Vaughn, J. Mastromarino, B. Heikes, J. Snow, J. Podolske, and H. Singh, “OH and HO2 in the tropical Pacific: results from PEM-Tropics B,” J. Geophys. Res. Atmos. 106(D23), 32667–32681 (2001). [CrossRef]
26. X. Ren, J. R. Olson, J. H. Crawford, W. H. Brune, J. Mao, R. B. Long, Z. Chen, G. Chen, M. A. Avery, G. W. Sachse, J. D. Barrick, G. S. Diskin, L. G. Huey, A. Fried, R. C. Cohen, B. Heikes, P. O. Wennberg, H. B. Singh, D. R. Blake, and R. E. Shetter, “HOx chemistry during INTEX-A 2004: Observation, model calculation, and comparison with previous studies,” J. Geophys. Res. Atmos. 113(D5), 1 (2008). [CrossRef]
27. J. Lelieveld, T. M. Butler, J. N. Crowley, T. J. Dillon, H. Fischer, L. Ganzeveld, H. Harder, M. G. Lawrence, M. Martinez, D. Taraborrelli, and J. Williams, “Atmospheric oxidation capacity sustained by a tropical forest,” Nature 452(7188), 737–740 (2008). [CrossRef] [PubMed]
28. T. A. M. Pugh, A. R. MacKenzie, C. N. Hewitt, B. Langford, P. M. Edwards, K. L. Furneaux, D. E. Heard, J. R. Hopkins, C. E. Jones, A. Karunaharan, J. Lee, G. Mills, P. Misztal, S. Moller, P. S. Monks, and L. K. Whalley, “Simulating atmospheric composition over a South-East Asian tropical rainforest: performance of a chemistry box model,” Atmos. Chem. Phys. 10(1), 279–298 (2010). [CrossRef]
29. L. K. Whalley, P. M. Edwards, K. L. Furneaux, A. Goddard, T. Ingham, M. J. Evans, D. Stone, J. R. Hopkins, C. E. Jones, A. Karunaharan, J. D. Lee, A. C. Lewis, P. S. Monks, S. J. Moller, and D. E. Heard, “Quantifying the magnitude of a missing hydroxyl radical source in a tropical rainforest,” Atmos. Chem. Phys. 11(14), 7223–7233 (2011). [CrossRef]
30. K. D. Lu, F. Rohrer, F. Holland, H. Fuchs, B. Bohn, T. Brauers, C. C. Chang, R. Häseler, M. Hu, K. Kita, Y. Kondo, X. Li, S. R. Lou, S. Nehr, M. Shao, L. M. Zeng, A. Wahner, Y. H. Zhang, and A. Hofzumahaus, “Observation and modelling of OH and HO2 concentrations in the Pearl River Delta 2006: a missing OH source in a VOC rich atmosphere,” Atmos. Chem. Phys. 12(3), 1541–1569 (2012). [CrossRef]
31. K. D. Lu, F. Rohrer, F. Holland, H. Fuchs, T. Brauers, A. Oebel, R. Dlugi, M. Hu, X. Li, S. R. Lou, M. Shao, T. Zhu, A. Wahner, Y. H. Zhang, and A. Hofzumahaus, “Nighttime observation and chemistry of HOx in the Pearl River Delta and Beijing in summer 2006,” Atmos. Chem. Phys. 14(10), 4979–4999 (2014). [CrossRef]
32. Z. Tan, H. Fuchs, K. Lu, A. Hofzumahaus, B. Bohn, S. Broch, H. Dong, S. Gomm, R. Häseler, L. He, F. Holland, X. Li, Y. Liu, S. Lu, F. Rohrer, M. Shao, B. Wang, M. Wang, Y. Wu, L. Zeng, Y. Zhang, A. Wahner, and Y. Zhang, “Radical chemistry at a rural site (Wangdu) in the North China Plain: observation and model calculations of OH, HO2 and RO2 radicals,” Atmos. Chem. Phys. 17(1), 663–690 (2017). [CrossRef]
33. F. Rohrer, K. Lu, A. Hofzumahaus, B. Bohn, T. Brauers, C.-C. Chang, H. Fuchs, R. Haeseler, F. Holland, M. Hu, K. Kita, Y. Kondo, X. Li, S. Lou, A. Oebel, M. Shao, L. Zeng, T. Zhu, Y. Zhang, and A. Wahner, “Maximum efficiency in the hydroxyl-radical-based self-cleansing of the troposphere,” Nat. Geosci. 7(8), 559–563 (2014). [CrossRef]
34. K. D. Lu, A. Hofzumahaus, F. Holland, B. Bohn, T. Brauers, H. Fuchs, M. Hu, R. Häseler, K. Kita, Y. Kondo, X. Li, S. R. Lou, A. Oebel, M. Shao, L. M. Zeng, A. Wahner, T. Zhu, Y. H. Zhang, and F. Rohrer, “Missing OH source in a suburban environment near Beijing: observed and modelled OH and HO2 concentrations in summer 2006,” Atmos. Chem. Phys. 13(2), 1057–1080 (2013). [CrossRef]
35. H. Chen, R. Hu, P. Xie, X. Xing, L. Ling, Z. Li, F. Wang, Y. Wang, J. Liu, and W. Liu, “A hydroxyl radical detection system using gas expansion and fast gating laser-induced fluorescence techniques,” J. Environ. Sci. (China) 65, 190–200 (2018). [CrossRef] [PubMed]
36. M. Martinez, H. Harder, D. Kubistin, M. Rudolf, H. Bozem, G. Eerdekens, H. Fischer, T. Klüpfel, C. Gurk, R. Königstedt, U. Parchatka, C. L. Schiller, A. Stickler, J. Williams, and J. Lelieveld, “Hydroxyl radicals in the tropical troposphere over the Suriname rainforest: airborne measurements,” Atmos. Chem. Phys. 10(8), 3759–3773 (2010). [CrossRef]
37. S. C. Smith, J. D. Lee, W. J. Bloss, G. P. Johnson, T. Ingham, and D. E. Heard, “Concentrations of OH and HO2 radicals during NAMBLEX: measurements and steady state analysis,” Atmos. Chem. Phys. 6(5), 1435–1453 (2006). [CrossRef]
38. I. C. Faloona, D. Tan, R. L. Lesher, N. L. Hazen, C. L. Frame, J. B. Simpas, H. Harder, M. Martinez, P. Di Carlo, X. R. Ren, and W. H. Brune, “A laser-induced fluorescence instrument for detecting tropospheric OH and HO2: Characteristics and calibration,” J. Atmos. Chem. 47(2), 139–167 (2004). [CrossRef]
39. S. Dusanter, D. Vimal, P. S. Stevens, R. Volkamer, and L. T. Molina, “Measurements of OH and HO2 concentrations during the MCMA-2006 field campaign - Part 1: Deployment of the Indiana University laser-induced fluorescence instrument,” Atmos. Chem. Phys. 9(5), 1665–1685 (2009). [CrossRef]
40. Y. Kanaya and H. Akimoto, “Gating a channel photomultiplier with a fast high-voltage switch: reduction of afterpulse rates in a laser-induced fluorescence instrument for measurement of atmospheric OH radical concentrations,” Appl. Opt. 45(6), 1254–1259 (2006). [CrossRef] [PubMed]
41. D. Amedro, K. Miyazaki, A. Parker, C. Schoemaecker, and C. Fittschen, “Atmospheric and kinetic studies of OH and HO2 by the FAGE technique,” J. Environ. Sci. (China) 24(1), 78–86 (2012). [CrossRef] [PubMed]
42. R. Hu, D. Wang, and P. Xie, “Diode laser cavity ring-down spectroscopy for atmospheric NO2 measurement,” Acta Opt. Sin. 36, 320–327 (2016).
43. W. J. Bloss, J. D. Lee, C. Bloss, D. E. Heard, M. J. Pilling, K. Wirtz, M. Martin-Reviejo, and M. Siese, “Validation of the calibration of a laser-induced fluorescence instrument for the measurement of OH radicals in the atmosphere,” Atmos. Chem. Phys. 4(2), 571–583 (2004). [CrossRef]
44. E. J. Lanzendorf, T. F. Hanisco, N. M. Donahue, and P. O. Wennberg, “Comment on: “The measurement of tropospheric OH radicals by laser‐induced fluorescence spectroscopy during the POPCORN Field Campaign” by Hofzumahaus et al. and “Intercomparison of tropospheric OH radical measurements by multiple folded long‐path laser absorption and laser induced fluorescence” by Brauers et al,” Geophys. Res. Lett. 24(23), 3037–3038 (1997). [CrossRef]
45. D. J. Creasey, D. E. Heard, and J. D. Lee, “Absorption cross-section measurements of water vapour and oxygen at 185 nm. Implications for the calibration of field instruments to measure OH, HO2 and RO2 radicals,” Geophys. Res. Lett. 27(11), 1651–1654 (2000). [CrossRef]
46. R. A. Copeland and D. R. Crosley, “Temperature dependent electronic quenching of OH( A2 Σ +, v ’=0) between 230 and 310 K,” J. Chem. Phys. 84(6), 3099–3105 (1986).
47. A. E. Bailey, D. E. Heard, D. A. Henderson, and P. H. Paul, “Collisional quenching of OH (,by H2O between 211 and 294 K and the development of a unified model for quenching,” Chem. Phys. Lett. 302(1-2), 132–138 (1999). [CrossRef]
48. D. Poppe, M. Wallasch, and J. Zimmermann, “The dependence of the concentration of oh on its precursors under moderately polluted conditions: a model study,” J. Atmos. Chem. 16(1), 61–78 (1993). [CrossRef]
