Measurement of Atmospheric Ozone by Cavity Ring-down

ARTICLE pubs.acs.org/est

Measurement of Atmospheric Ozone by Cavity Ring-down Spectroscopy R. A. Washenfelder,*,†,‡ N. L. Wagner,†,‡ W. P. Dube,†,‡ and S. S. Brown‡ †

Cooperative Institute for Research in Environmental Sciences, 216 UCB, University of Colorado, Boulder, Colorado 80309, United States ‡ Chemical Sciences Division, Earth System Research Laboratory, National Oceanic and Atmospheric Administration, 325 Broadway, Boulder, Colorado 80305, United States ABSTRACT: Ozone plays a key role in both the Earth’s radiative budget and photochemistry. Accurate, robust analytical techniques for measuring its atmospheric abundance are of critical importance. Cavity ring-down spectroscopy has been successfully used for sensitive and accurate measurements of many atmospheric species. However, this technique has not been used for atmospheric measurements of ozone, because the strongest ozone absorption bands occur in the ultraviolet spectral region, where Rayleigh and Mie scattering cause significant cavity losses and dielectric mirror reflectivities are limited. Here, we describe a compact instrument that measures O3 by chemical conversion to NO2 in excess NO, with subsequent detection by cavity ring-down spectroscopy. This method provides a simple, accurate, and high-precision measurement of atmospheric ozone. The instrument consists of two channels. The sum of NO2 and converted O3 (defined as Ox) is measured in the first channel, while NO2 alone is measured in the second channel. NO2 is directly detected in each channel by cavity ring-down spectroscopy with a laser diode light source at 404 nm. The limit of detection for O3 is 26 pptv (2 sigma precision) at 1 s time resolution. The accuracy of the measurement is (2.2%, with the largest uncertainty being the effective NO2 absorption cross-section. The linear dynamic range of the instrument has been verified from the detection limit to above 200 ppbv (r2 > 99.99%). The observed precision on signal (2 sigma) with 41 ppbv O3 is 130 pptv in 1 s. Comparison of this instrument to UV absorbance instruments for ambient O3 concentrations shows linear agreement (r2 = 99.1%) with slope of 1.012 ( 0.002.

1. INTRODUCTION Ozone (O3) is one of the most important trace gases in the Earth’s atmosphere. In the troposphere, it is formed by photochemical reactions involving nitrogen oxides (NOx) and volatile organic compounds.1,2 Tropospheric ozone has negative health consequences and is designated as a criteria pollutant under the U.S. Clean Air Act of 1970 and its subsequent amendments. Accurate measurements of ozone are important for understanding the processes that control its production and loss, and in assessing control strategies. A number of in situ and remote sensing techniques have been developed to measure O3, including ultraviolet (UV) absorbance,3 chemiluminescence,4 chemical titration techniques,5,6 DOAS,7 and LIDAR,8 with UV absorbance and chemiluminescence being the most widely used for ambient in situ sampling. UV absorbance is an absolute method that typically uses a mercury lamp at 254 nm to exploit the strong absorption by O3, and is the basis for several commercially available instruments. Such instruments have sensitivity of ∼1 ppbv, response times of 5 - 10 s,9 and are susceptible to modest interference from other UV absorbers, such as aromatic hydrocarbons.10 Chemiluminescence instruments measure fluorescence from electronically excited NO2 produced by the reaction of O3 with excess NO,4 achieving sensitivity of ∼0.01 ppbv and time response of 2 s.4 These instruments require concentrated NO, a vacuum system, r 2011 American Chemical Society

and must be calibrated to determine the relationship between photon counts and O3 concentration. Cavity ring-down spectroscopy (CRDS) is a sensitive technique for direct absorption measurements and has been applied to atmospheric sampling.11 It may be considered absolute if the analyte’s absorption cross-section is well-known, interferences are small or well quantified, and sampling losses are minimal. CRDS has been used for spectroscopic studies of weak O3 absorption bands in the near-infrared spectral region,12-14 but atmospheric measurements of ambient O3 mixing ratios by CRDS have not, to our knowledge, been previously reported. The main challenge is that the strong absorption bands for O3 occur in the UV spectral region,15 where optical path lengths for CRDS measurements are reduced because dielectric mirrors have reduced reflectivity, and Rayleigh and Mie scattering are much stronger, leading to more significant cavity losses.11 Here, we describe a method to chemically convert O3 to NO2, with subsequent detection by CRDS at 404 nm. Chemical amplification instruments that measure peroxy radicals by their conversion to NO2 have previously observed an interference Received: October 1, 2010 Accepted: January 27, 2011 Revised: January 14, 2011 Published: March 02, 2011 2938

dx.doi.org/10.1021/es103340u | Environ. Sci. Technol. 2011, 45, 2938–2944

Environmental Science & Technology

ARTICLE

Figure 1. Schematic of the cavity ring-down spectrometer for measurement of O3 by conversion to NO2. The optical components consist of a CW diode laser, turning mirrors, beamsplitters, high reflectivity mirrors, lenses, optical fibers, and photomultiplier tubes. Ox, NO2, and NOx are measured in three channels by chemical conversion to NO2. A blank volume is included in the sampling line of the NO2 channel to match the residence times of the other two channels.

produced by the reaction of O3 with excess NO.16,17 In the instrument described here, the sum of converted ambient O3 and ambient NO2 is measured in one channel to provide a direct measurement of Ox (NO2 þ O3). Subtraction of the NO2 measured in a second, independent channel provides a direct measurement of O3 alone. This method is complementary to existing O3 measurement techniques and offers several advantages. First, as noted above, CRDS is an absolute method and provides analytical accuracy comparable to direct UV absorption spectroscopy. The conversion of O3 to NO2 does not compromise the accuracy of the method because it is simple, quantitative, and free from interference, as demonstrated below. Second, CRDS offers sensitivity, precision, and time response that are significantly better than UV absorbance instruments and approach that of chemiluminescence instruments. Finally, the technique has reduced requirements for size, power, weight, vacuum, and excess reagents compared to chemiluminescence. The CRDS O3 measurement expands upon an existing instrumental technique18 for NO2 and NOx, allowing all three species to be determined simultaneously in a compact instrument with high sensitivity and time response. The advantages described above make this O3 instrument a reasonable alternative for a wide range of applications, from routine monitoring to aircraft measurements.

2. EXPERIMENTAL SECTION 2.1. NO2 Detection Using Cavity Ring-down Spectroscopy. The use of CRDS for NO2 and NOx detection at 404 nm has been described previously.18 The O3 detection described here was demonstrated using two CRDS instruments in our laboratory. The first instrument was designed for deployment on the NOAA WP-3D aircraft, and duplicates the Fuchs et al.18 design to include an additional optical channel for the measurement of Ox, as shown in Figure 1 and summarized here. Although this

instrument is mounted on an optical bench, rather than a cage system, there is no fundamental difference in the optical setup or its operation. All data shown in Figures 2-6 were acquired using the aircraft instrument. The second instrument is the optical cage system described in Fuchs et al.,18 which was modified for measurement of NO2 and Ox, instead of NO2 and NOx. The cage system has higher precision due to its improved data acquisition system (∼2 kHz vs 1.2 kHz), and we report precision measurements acquired with this instrument in the Results and Discussion. Light is provided by a continuous wave (cw) diode laser with temperature-controlled diode (Power Technology Inc., FabryPerot diode model IQμ series) and an optical output of approximately 80 mW. The laser output is amplitude modulated with a square-wave signal at a repetition of 2 kHz and a duty cycle of 50%. An optical isolator, consisting of a polarizer and 1/4 wave plate, has been added to the instrument to reduce optical feedback to the laser. Using two beamsplitters and two turning mirrors, the laser is coupled into three optical cavities without active matching between the laser and cavity modes. When the laser is modulated off, the light intensity in the cavity exponentially decreases. The number density of the absorber [NO2] can be calculated by observing time constants with (τ) and without (τ0) the absorber present in the cavity, according to RL 1 1 ½NO2 ¼ ð1Þ cσNO2 τ τ0 where RL is ratio of the total cavity length to the length over which the absorber is present, c is the speed of light, and σNO2 is the absorption cross-section. Previous measurements of RL have shown that it is smaller than the geometric ratio,19 and we have used a value of 1.15, consistent with prior measurements. Further information about the principle of CRDS can be found elsewhere.11,20,21 2939

dx.doi.org/10.1021/es103340u |Environ. Sci. Technol. 2011, 45, 2938–2944


ARTICLE

Figure 3. Conversion efficiency of 200 ppbv O3 to NO2 as a function of NO concentration. The top axis shows the calculated NO number density. The bottom axis shows the pseudo first-order reaction rate coefficient, for k = 3.0 10-12 exp(-1500K/T). The NO þ O3 reaction data was fit to the relationship conversion efficiency = 1 - exp(k[NO]tresidence), yielding a residence time of 0.78 ( 0.02 s (black line).

Figure 2. (a) Measured line shape for CRDS laser. Center wavelength is 404.0 ( 0.1 nm with fwhm of 0.5 ( 0.15 nm. (b) Gray line shows Vandaele et al.24 NO2 cross-section. Black line shows the NO2 crosssection convolved with a Gaussian function with fwhm of 0.5 nm. The solid black hatch mark shows the center wavelength of the CRDS laser, corresponding to an NO2 cross-section of (6.1 ( 0.2) 10-19 cm2. (c) An example plot showing the method used to determine the effective NO2 cross-section by measuring the cavity extinction with known NO2 concentrations. The slope of the line in (c) is equivalent to the effective NO2 cross-section, found to be (6.27 ( 0.02) 10-19 cm2.

Each cavity consists of a pair of 2.54 cm diameter, 1 m radius of curvature mirrors (Advanced Thin Films), with measured reflectivity 99.9965%, giving a ring-down time constant of approximately 40 μs at their nominal center wavelength of 404 nm at ambient pressure (840 hPa in Boulder, CO; 1600 m ASL). The mirrors are mounted at either end of a 93 cm cell, with the center 78 cm constructed of 6.2 mm inner diameter Teflon through which the ambient air sample flows. Small purge flows (∼25 sccm of zero air for each mirror) are introduced directly at the mirrors to maintain their cleanliness. The mirror mounts and associated purge volumes were connected to the sample cell by flexible stainless steel bellows, as shown in Figure 1. The sample flow from the common inlet is divided into three separate cells.

Figure 4. Correlation plot comparing O3 measurements acquired by the CRDS to a UV absorbance instrument (Thermo Electron Scientific 49i). Each measurement was acquired and averaged for 60 s. Fitted slope is 0.988 ( 0.001 with intercept -0.2 ( 0.2, with r2 = 0.99998.

Sample air enters and exits each sample cell through PFA Teflon fittings. The sampling flow rate is maintained at 2.0 standard liters per min (slpm) by a mass flow controller for each channel. An oil-free scroll pump is used to pull sample air through the cells. Pressure and temperature were measured independently for each cell. The light transmitted through the end mirror of each cavity is collected with a fiber optic and projected through a bandpass filter (Thorlabs 405 nm center; 10 nm fwhm) onto a photomultiplier tube module (Hamamatsu HC 120). Ring-down transients at a repetition rate of 2 kHz were acquired with a PCI-based transient digitizer (National Instruments), coadded, and fit to a single exponential. The number of coadded transients can be varied, resulting in a time resolution of 0.2-1 Hz. Although the laser modulation frequency was 2 kHz, the actual acquisition rate for this instrument was reduced to approximately 1.2 kHz for the aircraft instrument, due to limitations in data transfer rates and computational requirements for fitting traces. Co-added traces were fitted following the method described in Fuchs et al.18 2940



Figure 5. (a) An Allan deviation plot for zero air measurements acquired by the CRDS, showing the relationship between integration time and 1σ precision. Ox (blue), NO2 (red), and O3 (Ox - NO2) (black) are shown. The dashed line indicates the relationship expected for statistically random noise. (b) Precision on signal for the CRDS O3 measurement (black) and two UV absorbance instruments (brown and blue).

The ring-down time constant in the absence of the absorber (τ0) was measured at regular intervals by periodically introducing a flow of zero air in excess of the total sample flow to the inlet, to account for potential drift due to changes in the alignment of the optical cavities. The overflow is sufficient that the cells are completely filled with zero air. Values of τ0 between zeros are determined by linear interpolation. Any small change in optical extinction due to pressure difference during the zero air addition is explicitly corrected using the known Rayleigh scattering of zero air,22 and was never greater than the equivalent of 6 pptv of NO2. Measurements in ambient or humidified air require an additional correction for water vapor, because the Rayleigh scattering crosssection of water vapor is smaller than that of dry air.18 However this correction affects both the Ox and NO2 channels similarly, and hence does not affect the O3 measurement. NO2 has strong absorptions throughout the visible spectral region, with a maximum near 420 nm.23,24 We have chosen to construct an instrument at 404 nm to minimize interference by other absorbers and because of the ready availability of lasers at this wavelength. Two dicarbonyl compounds, glyoxal and methylglyoxal, have absorptions in this spectral region,25,26 but the cross-sections in each case are approximately 10 times smaller than that of NO2 when convolved with the measured CRDS laser cross-section. These interferences are small and would cancel in the subtraction between the two measurement channels used for

ARTICLE

Figure 6. (a) Simultaneous measurements of ambient laboratory air by CRDS and a UV absorption instrument. Plot shows CRDS Ox (black), CRDS NO2 (red), and CRDS O3 (yellow). UV absorbance data is shown as the blue trace. (b) Correlation plot for 46 h of ambient data, with data averaged over one-minute time intervals. Error bars represent the standard deviation for each one minute average. Slope is 1.012 ( 0.002 with an intercept of -1.88 ( 0.04 ppbv and r2 = 0.991.

the O3 measurement. The instrument sensitivity to direct absorption by O3 is important because this would introduce an error into the direct measurement of NO2. At 404 nm, the absorption cross-section for O327 is more than 104 times smaller than that of NO2.23,24 This was confirmed by Fuchs et al.,18 who measured an O3 cross-section of (1.49 ( 0.01) 10-23 cm2 at 404 nm. The error introduced by 100 ppbv O3 at 298 K and 840 hPa is equivalent to 2.4 pptv of NO2, and represents a small interference that we have neglected here. 2.2. Calibration of the CRDS NO2 Measurement. Quantitative measurements require an accurate absorption cross-section for NO2 at the center wavelength and line width of the diode laser. The manufacturer specifies a center wavelength within a ( 5 nm range around 405 nm and gives only a nominal specification for the laser line width. Because the NO2 absorption cross-section is highly structured in this spectral region, greater precision is required. We employed two independent methods to determine the effective NO2 cross-section corresponding to the laser wavelength. First, using a grating spectrometer (Acton InSpectrum, 1200 g/mm grating), we measure the laser center wavelength and full width at half-maximum (fwhm). Figure 2a shows the measured spectrum of the laser in the aircraft instrument, which has a center wavelength of 404.0 ( 0.1 nm and a fwhm of 0.5 ( 0.15 nm. This laser overlaps approximately 6000 longitudinal cavity modes (and likely a much larger number of transverse modes), consistent with the observed passive coupling of the laser into the 2941



ARTICLE

cavity. Convolving the measured fwhm with an NO2 crosssection recorded at 294 K and less than 3 hPa,24 gives an effective NO2 cross-section of (6.1 ( 0.2) 10-19 cm2, shown in Figure 2b. The laser diode is temperature stabilized and its center wavelength appears stable within the uncertainty of the wavelength measurement. A shift in the laser center wavelength of (0.1 nm would change the effective NO2 cross-section by -0.4-0.7%. The 3% uncertainty reported here is determined from the larger absolute uncertainty in the reported NO2 crosssection.24 An additional potential source of error is the pressuredependence of the NO2 cross-section.23 Second, the effective NO2 cross-section was directly calibrated using known NO2 concentrations. This method is based on optical absorption spectroscopy and does not require calibrated gas mixtures, which have been shown to be problematic for NO2 in some cases (e.g., refs 28 and 29). Using an adjustable O3 source and a commercial UV absorption O3 instrument (Thermo Electron Scientific [TES] 49i), we generated known concentrations of NO2 in the range of 20-200 ppbv by conversion of O3 to NO2 in excess NO (see discussion in Section 2.3). This calibration method is appropriate for the CRDS instrument, which is sensitive to optical extinction by NO2, but insensitive to the large excess of NO. Figure 2c shows a plot of optical extinction versus NO2 concentration, whose slope is equal to σNO2/RL and whose intercept is due to NO2 present in the reagent NO, as described further below. The average effective NO2 cross-section from four separate determinations is (6.27 ( 0.02) 10-19 cm2 for RL = 1.15. The two methods for determining the effective NO2 crosssection are consistent within their combined uncertainties, with an absolute difference of approximately 3%. For the retrievals reported here, we have used an effective NO2 cross-section of (6.27 ( 0.02) 10-19 cm2 for the laser in the aircraft instrument. We determined the effective NO2 cross-section under pressure and temperature conditions of 820-835 hPa and 23.5 - 24.5 °C. Subsequent measurements were conducted under similar conditions. We were not able to detect changes in the effective NO2 cross-section over this small range of pressure and temperature; however, from 500 hPa to 1000 hPa, the effective NO2 cross-section increases by approximately 6% (Wagner, in preparation). Variations of the effective NO2 cross-section with pressure and temperature can be accounted for by controlling the cell pressure and temperature or by measuring the NO2 cross-section as a function of these parameters. 2.3. Chemical Conversion of O3 to NO2 in Excess NO. Two CRDS channels are required for the O3 measurement. Channel 1 measures Ox, which includes NO2 and O3 that has been converted to NO2 by titration with NO. Channel 2 measures ambient NO2. The difference between the measured Ox and NO2 represents the O3 concentration. O3 was quantitatively converted to NO2 by reaction with excess NO, according to the reaction NO þ O3 f NO2 þ O2

ð2Þ

with k = 3.0 10-12 exp(-1500 K/T) or 2 10-14 cm3 molecule-1 s-1 at 298 K.30 A standard cylinder provided 498 ppmv NO in N2 (Scott-Marin), and the addition was regulated by a needle valve and measured using a mass flow meter. The actual concentration of NO added to the reactor is not critical since it is an excess reagent. The reaction volume for the O3

converter in the current instrument consists of a 34 cm length of 0.95 cm inner diameter Teflon tubing. For a sample flow of 2.0 slpm at 298 K, the plug flow residence time in the reaction volume is calculated to be 0.60 s. For large excess NO concentrations, we expect that (2) will behave as a pseudo first order reaction. This was confirmed experimentally by varying the flow rate of NO addition, as shown in Figure 3. The measurements in Figure 3 were fitted to the relationship 1 - exp(-k[NO]tresidence), and the resulting value for the NO þ O3 reaction time was 0.78 ( 0.02 s. For a sample flow of 200 ppbv O3, complete flow conversion to NO2 in the reaction volume was achieved by an addition of 70 sccm NO at 498 ppmv, corresponding to an NO mixing ratio of approximately 17 ppmv or a number density of 3.6 1014 molecules cm-3 at 840 hPa in Boulder, CO (1600 m ASL). The O3 conversion efficiency for these conditions is determined to be 1-exp(-k[NO]tresidence) = 99.6%, and the subsequent CRDS O3 measurements reported here have been corrected by 0.4%. A disadvantage of O3 conversion to NO2 by reaction with excess NO is the unavoidable presence of a background due to NO2 contamination in the NO addition flow. Addition of NO through an FeSO4 converter31 reduces the NO2 contamination to 0.5-2 ppbv. This is part of the background signal, as the NO addition is continuous even when the inlet is under zero air overflow. The NO2 background achieves a constant, steady value within a few minutes, and subsequently drifts slowly relative to the frequency of zero measurements, which are recorded every 5-10 min.

3. RESULTS AND DISCUSSION 3.1. Accuracy and Comparison to TES Ozone Instruments. The absolute accuracy of the O3 measurement is determined by the uncertainty of each term in eq 1. This encompasses uncertainty in τ, τ0, σNO2/RL, pressure, temperature, dilution by purge flow and NO addition, O3 conversion to NO2, interference by other absorbers, and correction for differences in Rayleigh scattering due to pressure and water vapor. The largest uncertainty is σNO2/RL, determined using the TES ozone instrument with a relative error of approximately (2%. The other factors are estimated to be accurate within (0.1% (τ and τ0), (0.5% (pressure), (0.2% (temperature), (0.4% (flow dilution) (0.5% (O3 conversion to NO2), (0.2% (interference by glyoxal and methylglyoxal), and (0.1% (Rayleigh scattering correction for pressure and water vapor). Added in quadrature, this gives a total uncertainty of (2.2% for ambient measurements acquired at 1 s. To evaluate its accuracy, the CRDS O3 instrument was compared to two UV absorbance instruments, both manufactured by Thermo Electron Scientific [TES] (model 49i). The accuracy of these instruments is better than 2%,32 and the manufacturer indicates a 1 ppbv limit of detection. One of the two instruments is maintained in our laboratories as a standard instrument for sampling synthetic air, and is not used to sample ambient air. Ozone was generated synthetically using the Hg lamp in one of the TES instruments over a range 25-250 ppbv. It is important to note that this comparison is equivalent to the calibration procedure described above for the 404 nm absorption cross-section, and therefore should yield the same result to within the stated uncertainty of the measurement. The only difference is that the O3 to NO2 reactor is part of the Ox-NO2 instrument in this comparison, whereas it is part of the calibration source during NO2 cross-section measurements. Due to the slow 2942


Environmental Science & Technology response time of the TES instruments, each concentration was sampled for approximately 5 min, and the measurements were averaged for 1 min. We found a slope of 0.995 ( 0.002 between the two TES O3 monitors. Figure 4 shows the correlation plot between the CRDS and TES O3 monitor, with slope 0.988 ( 0.001 and intercept -0.2 ( 0.2 ppbv/ppbv, with r2 = 99.998%. The slope is consistent with an absolute accuracy for the CRDS measurement of (2.2%. 3.2. Limit of Detection and Precision. Figure 5a shows the relationship between the 1σ precision and integration time for a period of zero air measurements for the aircraft instrument. At 1 s integration time, the NO2 and Ox channels have 1σ precision of 23 and 17 pptv respectively. As expected, O3 (Ox-NO2) has a precision that is equivalent to a quadrature addition of Ox and NO2, with 1σ of 28 pptv in 1 s. The Allan deviation plot shows that the optimum O3 sensitivity occurs at an integration time of ∼10 s, with a value of 7 pptv. The O3 sensitivity of the optical cage system is more than a factor of 2 better than the aircraft instrument, with a 1σ precision of 13 pptv in 1 s or 5 pptv in 10 s. To evaluate the CRDS precision on signal, the two TES monitors were used in a similar configuration as described in Section 3.1 to generate a constant concentration of O3. Figure 5b shows an example of 15 min of data acquired with constant O3 concentration of 41.0 ppbv at 820 hPa and 24.2 °C (corresponding to an O3 number density 8.2 1011 cm-3). The change observed in the TES O3 measurements during the CRDS zeros are attributed to a pressure change in the shared inlet. The 1σ precision of the CRDS measurement on signal is 112 pptv in 1 s, and is a linear function of the O3 concentration. For a 41.5 ppbv concentration of O3, the 1σ precision of the cage system instrument is 65 pptv, also approximately factor of 2 improvement over the aircraft instrument. 3.3. Ambient Measurements of O3. Ambient CRDS O3 measurements were validated against the TES instrument during a 46 h period of laboratory air sampling with O3 mixing ratios in the range 7-29 ppbv. CRDS zeros were recorded for 20 s every 4 min. Figure 6a shows Ox, NO2, and O3 measured by CRDS together with the TES O3 measurement for a two hour period of measurements when ozone concentrations varied between 9.8 and 25.4 ppbv. While O3 and NO2 are anticorrelated in the ambient measurements, Ox itself is much less variable, as a result of the titration of O3 by NOx emissions in urban air. The correlation between the CRDS and TES instruments are shown in Figure 6b, as one minute averages. The error bars represent the standard deviation for each average. The slope is 1.012 ( 0.002 with an intercept of -1.88 ( 0.04 and r2 of 0.991. We attribute the error in the intercept to the TES instruments, which are known to suffer from zero offsets of this order. 3.4. Discussion. The CRDS method described in this work is complementary to existing O3 measurement techniques. It offers accuracy comparable to commercial UV absorption instruments, but with substantially improved sensitivity and time response that approaches that offered by research grade chemiluminescence instruments.4 The instruments described here demonstrate accuracy of 2.2%, 2σ sensitivity of 26 pptv at zero, and 2σ precision of 130 pptv on signal, all at 1 s time resolution. Furthermore, the approach is simple and robust. It uses a standard, commercially available light source and detector, and standard dielectric coated mirrors for the optical cavity. The sensitivity is such that O3 may be measured over a wide range in concentration, free from significant spectral or other interferences. Although it requires a toxic, excess reagent (NO) for the

ARTICLE

conversion of O3 to NO2, the required concentrations are dilute relative to those used in chemiluminescence applications. Furthermore, all measurements are made under flow at ambient operating pressure, such that the pumping requirements are more comparable to commercial, UV absorption monitors than to research grade chemiluminescence instruments. The combination of simultaneous NO2 and O3 measurements offers additional advantages in convenience and accuracy over existing instruments. Commercial UV absorbance instruments do not offer NO2 measurements. Thus, this instrument represents a simple, robust, and potentially low cost method for these two related measurements.

’ AUTHOR INFORMATION Corresponding Author

*E-mail: [email protected].

’ ACKNOWLEDGMENT We thank Eric Williams for loan of the TES ozone monitor. This work was funded in part by the NOAA Atmospheric Chemistry and Climate Program. ’ REFERENCES (1) Crutzen, P. J. Role of NO and NO2 in the chemistry of the troposphere and stratosphere. Annu. Rev. Earth Planet. Sci. 1979, 7, 443– 472. (2) Haagen-Smit, A. J. Chemistry and physiology of Los Angeles smog. Ind. Eng. Chem. 1952, 44 (6), 1342–1346. (3) Proffitt, M. H.; Mclaughlin, R. J. Fast-response dual-beam UVabsorption ozone photometer suitable for use on stratospheric balloons. Rev. Sci. Instrum. 1983, 54 (12), 1719–1728. (4) Ridley, B. A.; Grahek, F. E.; Walega, J. G. A small, high-sensitivity, medium-response ozone detector suitable for measurements from light aircraft. J. Atmos. Oceanic Technol. 1992, 9 (2), 142–148. (5) Komhyr, W. D. Electrochemical concentration cells for gas analysis. Ann. Geophys. (C. N. R. S.) 1969, 25 (1), 203–210. (6) Brewer, A. W.; Milford, J. R. The Oxford-Kew ozone sonde. Proc. R. Soc. London, Ser. A 1960, 256 (1287), 470–495. (7) Axelsson, H.; Edner, H.; Galle, B.; Ragnarson, P.; Rudin, M. Differential optical-absorption spectroscopy (DOAS) measurements of ozone in the 280-290 nm wavelength region. Appl. Spectrosc. 1990, 44 (10), 1654–1658. (8) Godin, S.; Megie, G.; Pelon, J. Systematic LIDAR measurements of the stratospheric ozone vertical distribution. Geophys. Res. Lett. 1989, 16 (6), 547–550. (9) Clemitshaw, K. C. A review of instrumentation and measurement techniques for ground-based and airborne field studies of gas-phase tropospheric chemistry. Crit. Rev. Environ. Sci. Technol. 2004, 34 (1), 1–108. (10) Kleindienst, T. E.; Hudgens, E. E.; Smith, D. F.; Mcelroy, F. F.; Bufalini, J. J. Comparison of chemiluminescence and ultraviolet ozone monitor responses in the presence of humidity and photochemical pollutants. J. Air Waste Manage. Assoc. 1993, 43 (2), 213–222. (11) Brown, S. S. Absorption spectroscopy in high-finesse cavities for atmospheric studies. Chem. Rev. 2003, 103 (12), 5219–5238. (12) Wachsmuth, U.; Abel, B., Linewidths and line intensity measurements in the weak 3A2(000) r X1A1(000) band of ozone by pulsed cavity ringdown spectroscopy. J. Geophys. Res., [Atmos.] 2003, 108, (D15), doi:10.1029/2002JD003126. (13) Enami, S.; Ueda, J.; Nakano, Y.; Hashimoto, S.; Kawasaki, M., Temperature-dependent absorption cross sections of ozone in the WulfChappuis band at 759-768 nm. J. Geophys. Res., [Atmos.] 2004, 109, (D5), doi: 10.1029/2003JD004097. 2943


Environmental Science & Technology (14) Campargue, A.; Kassi, S.; Romanini, D.; Barbe, A.; De BackerBarilly, M. R.; Tyuterev, V. G. CW-cavity ring down spectroscopy of the ozone molecule in the 6625-6830 cm-1 region. J. Mol. Spectrosc. 2006, 240 (1), 1–13. (15) Molina, L. T.; Molina, M. J. Absolute absorption cross sections of ozone in the 185-nm to 350-nm wavelength range. J. Geophys. Res., [Atmos.] 1986, 91 (D13), 14501–14508. (16) Cantrell, C. A.; Stedman, D. H.; Wendel, G. J. Measurement of atmospheric peroxy-radicals by chemical amplification. Anal. Chem. 1984, 56 (8), 1496–1502. (17) Liu, Y. D.; Morales-Cueto, R.; Hargrove, J.; Medina, D.; Zhang, J. S. Measurements of peroxy radicals using chemical amplification-cavity ringdown spectroscopy. Environ. Sci. Technol. 2009, 43 (20), 7791–7796. (18) Fuchs, H.; Dube, W. P.; Lerner, B. M.; Wagner, N. L.; Williams, E. J.; Brown, S. S. A sensitive and versatile detector for atmospheric NO2 and NOx based on blue diode laser cavity ring-down spectroscopy. Environ. Sci. Technol. 2009, 43 (20), 7831–7836. (19) Fuchs, H.; Dube, W. P.; Cicioira, S. J.; Brown, S. S. Determination of inlet transmission and conversion efficiencies for in situ measurements of the nocturnal nitrogen oxides, NO3, N2O5 and NO2, via pulsed cavity ring-down spectroscopy. Anal. Chem. 2008, 80 (15), 6010– 6017. (20) Berden, G.; Peeters, R.; Meijer, G. Cavity ring-down spectroscopy: Experimental schemes and applications. Int. Rev. Phys. Chem. 2000, 19 (4), 565–607. (21) Berden, G.; Engeln, R., Cavity Ring-Down Spectroscopy: Techniques and Applications; Wiley-Blackwell: New York, 2009; p 344. (22) Bodhaine, B. A.; Wood, N. B.; Dutton, E. G.; Slusser, J. R. On Rayleigh optical depth calculations. J. Atmos. Oceanic Technol. 1999, 16 (11), 1854–1861. (23) Vandaele, A. C.; Hermans, C.; Fally, S.; Carleer, M.; Colin, R.; Merienne, M. F.; Jenouvrier, A.; Coquart, B., High-resolution Fourier transform measurement of the NO2 visible and near-infrared absorption cross sections: Temperature and pressure effects. J. Geophys. Res., [Atmos.] 2002, 107, (D18), doi: 10.1029/2001JD000971. (24) Vandaele, A. C.; Hermans, C.; Simon, P. C.; Carleer, M.; Colin, R.; Fally, S.; Merienne, M. F.; Jenouvrier, A.; Coquart, B. Measurements of the NO2 absorption cross-section from 42000 cm-1 to 10000 cm-1 (238-1000 nm) at 220 and 294 K. J. Quant. Spectrosc. Radiat. Transfer 1998, 59 (3-5), 171–184. (25) Meller, R.; Raber, W.; Crowley, J. N.; Jenkin, M. E.; Moortgat, G. K. The UV-visible absorption spectrum of methylglyoxal. J. Photochem. Photobiol. A 1991, 62 (2), 163–171. (26) Volkamer, R.; Spietz, P.; Burrows, J.; Platt, U. High-resolution absorption cross-section of glyoxal in the UV-vis and IR spectral ranges. J. Photochem. Photobiol. A 2005, 172 (1), 35–46. (27) Brion, J.; Chakir, A.; Charbonnier, J.; Daumont, D.; Parisse, C.; Malicet, J. Absorption spectra measurements for the ozone molecule in the 350-830 nm region. J. Atmos. Chem. 1998, 30 (2), 291–299. (28) Fried, A.; Sams, R.; Dorko, W.; Elkins, J. W.; Cai, Z. T. Determination of nitrogen dioxide in air compressed gas mixtures by quantitative tunable diode-laser absorption spectrometry and chemiluminescence detection. Anal. Chem. 1988, 60 (5), 394–403. (29) Osthoff, H. D.; Brown, S. S.; Ryerson, T. B.; Fortin, T. J.; Lerner, B. M.; Williams, E. J.; Pettersson, A.; Baynard, T.; Dube, W. P.; Ciciora, S. J.; Ravishankara, A. R., Measurement of atmospheric NO2 by pulsed cavity ring-down spectroscopy. J. Geophys. Res., [Atmos.] 2006, 111, (D12), doi: 10.1029/2005JD006942. (30) Sander, S. P.; Friedl, R. R.; Ravishankara, A. R.; Golden, D. M.; Kolb, C. E.; Kurylo, M. J.; Molina, M. J.; Moortgat, G. K.; Keller-Rudek, H.; Finlayson-Pitts, B. J.; Wine, P. H.; Huie, R. E.; Orkin, V. L. Chemical Kinetics and Photochemical Data for Use in Atmospheric Studies; NASA Jet Propulsion Laboratory: Pasadena, Calif., 2006. (31) Fehsenfeld, F. C.; Dickerson, R. R.; Hubler, G.; Luke, W. T.; Nunnermacker, L. J.; Williams, E. J.; Roberts, J. M.; Calvert, J. G.; Curran, C. M.; Delany, A. C.; Eubank, C. S.; Fahey, D. W.; Fried, A.; Gandrud, B. W.; Langford, A. O.; Murphy, P. C.; Norton, R. B.; Pickering, K. E.; Ridley, B. A. A ground-based intercomparison of NO,

ARTICLE

NOx, and NOy measurement techniques. J. Geophys. Res., [Atmos.] 1987, 92 (D12), 14710–14722. (32) Williams, E. J.; Fehsenfeld, F. C.; Jobson, B. T.; Kuster, W. C.; Goldan, P. D.; Stutz, J.; McCleanny, W. A. Comparison of ultraviolet absorbance, chemiluminescence, and DOAS instruments for ambient ozone monitoring. Environ. Sci. Technol. 2006, 40 (18), 5755–5762.

2944


Measurement of Atmospheric Ozone by Cavity Ring-down

Recommend Documents