Anal Bioanal Chem (2004) 379 : 51–55 DOI 10.1007/s00216-004-2515-3

S P E C I A L I S S U E PA P E R

J. A. Morales · J. Treacy · S. Coffey

Urban ozone measurements using differential optical absorption spectroscopy

Received: 30 September 2003 / Revised: 8 January 2004 / Accepted: 12 January 2004 / Published online: 13 February 2004 © Springer-Verlag 2004

Abstract In order to improve the air quality in Europe the European Commission has issued a number of directives with regard to acceptable levels of a range of gaseous pollutants, which includes ozone. Therefore, monitoring of this compound is necessary to comply with EU legislation, to provide improved pollution warnings for those who are sensitive to air pollutants as well as providing valuable data for environmental planning. Open-path spectroscopic techniques, such as differential optical absorption spectroscopy (DOAS), are ideal for monitoring pollutants because of the advantages they offer over classical methods and point-source analysers. A DOAS system has been installed in Dublin city centre to monitor a range of criteria pollutants including ozone. Observations of urban background ozone concentrations are presented. The measurements are compared with those obtained using a UV point-source analyser and are presented in the context of the current EU directive. The influence of trans-boundary pollution from mainland Europe leading to ozone episodes is also discussed. Observations of high ozone during this measurement campaign coincided with the influx of photochemically polluted air masses which originated over continental Europe. For the analysed time interval, the data suggest that the ground ozone level in Dublin might be significantly influenced by long-range transport from the United Kingdom and continental Europe. Keywords Differential optical absorption spectroscopy (DOAS) · Open-path spectroscopy · Ozone · Pollution · Trans-boundary pollution

Awarded a Poster Prize on the occasion of the Colloquium Spectroscopicum Internationale XXXIII, Granada, 7–12 September 2003 J. A. Morales (✉) · J. Treacy · S. Coffey Facility for Optical Characterisation and Spectroscopy (FOCAS), Dublin Institute of Technology, Kevin Street, 8 Dublin, Ireland Tel.: +353-1-402-2819, Fax: +353-1-402-4999, e-mail: [email protected]

Introduction It is now well established that high levels of particular trace gases (air pollutants) at ground level cause significant adverse health effects, damage vegetation and deteriorate buildings [1]. To improve air quality in Europe the European Commission has issued a number of directives with regard to acceptable levels of a range of gaseous pollutant species such as CO, SO2, NO, NO2, O3 and C6H6 [2]. Ground level ozone is the most complex, difficult to control and pervasive of the six principal pollutants according to the US EPA [3]. Ozone is the primary constituent of smog, however it is not emitted directly into the air by specific sources but created by photochemical reactions involving sunlight, nitrogen oxides (NOx) and volatile organic compounds (VOC). Ozone has important adverse health and environmental effects. It is a powerful respiratory irritant which has been linked to asthma and other health conditions, including a reduction in the body’s resistance to infection [4, 5, 6, 7, 8]. While adults with impaired respiratory systems and children are particularly susceptible, long-term exposure to even moderately high levels of ozone can lead to large reductions in lung function, inflammation of the lung lining and more frequent and severe respiratory discomfort even in healthy people [9, 10, 11, 12]. High ozone levels also cause respiratory problems in animals, lead to a reduction in agricultural crop yields and cause damage to forest ecosystems [3, 13]. For these reasons it is therefore necessary to monitor ground level ozone to provide improved pollution warnings for populations sensitive to air pollutants and to make available valuable data for environmental planning. Wet chemical methods for measuring air pollutants, having poor temporal resolution and poor selectivity, have been largely abandoned in favour of instrumental methods of analysis based on infrared spectroscopy (CO), UV fluorescence spectroscopy (SO2), chemiluminescence (NO and NO2), gas chromatography (C6H6) and UV absorption spectroscopy (O3). An urban monitoring network generally

52 Fig. 1 DOAS arrangement for measuring urban background pollution levels over Dublin city during the present campaign (24 hours a day from 03/26/03 to 05/02/03)

consists of a number of air quality monitoring stations (AQMSs) linked to a central control station where diagnostic checks, calibrations and data retrieval can be carried out. These AQMSs, while providing an accurate representation of pollution levels at fixed locations, are inevitably subject to emissions from local sources and hence a comprehensive monitoring network is required for measurements to be representative of large urban areas. In addition, a different monitoring technique is necessary for the analysis of each individual pollutant. More recently, open-path spectroscopic techniques such as differential optical absorption spectroscopy (DOAS) and Fourier-transform infrared spectroscopy (FTIR) have been developed, in which the concentration of several atmospheric pollutants can be accurately measured over relatively large areas with a single system [14, 15, 16]. This arrangement offers a number of distinct advantages over point-source analysers: (i) average concentrations are obtained over a large area rather than at a single point, hence minimising interferences from local emission sources; (ii) non-contact with the air eliminates the potential for loss of sample on surfaces, which may occur when ambient air is drawn into a conventional analyser; (iii) less maintenance and less frequent calibration is required; and (iv) a number of compounds can be detected with a single instrument thus reducing the overall cost. The DOAS principle is based on measurements of light absorption by the analyte in question using the so-called differential absorption [17, 18, 19, 20, 21]. This quantity can be defined as the part of the total absorption of any molecule “rapidly” varying with wavelength so the light variation due to atmospheric changes is eliminated.

Fig. 2 Pictures of the DOAS system from the roof-top of the Faculty of Science building at the Dublin Institute of Technology. Top: light source, receiver and optical path. Bottom: receiver

53 Fig. 3 Scheme of the DOAS system used in the present work

In the present work urban ozone studies in Dublin were carried out using a commercial DOAS spectrometer. The data obtained are compared with those acquired with a UV point-source analyser and observations are discussed in terms of the current EU directive on ozone and the influence of trans-boundary pollution from mainland Europe.

Experimental The DOAS system used to measure urban background ozone concentrations is a multi-path OPSIS AR300 (Opsis AB, Furulund, Sweden) consisting of four 150-W Xe arc lamps strategically placed at different locations in Dublin city centre and a receiver located on the roof of the Science Faculty building at DIT Kevin Street (Fig. 1). Pictures of the open-path system are shown in Fig. 2. The light arriving from the lamp is collected by a parabolic receiving mirror and focused onto the end of a 10-m fibre-optic cable. This cable transmits the light to the analyser (spectrometer and data acquisition system) situated in a room underneath the receiver. The spectrometer is a 0.5-m Czerny–Turner instrument operating at a recording time in the order of 10 ms to avoid the destruction of spectral information due to light variations caused by turbulence in the air. This is accomplished using a rotating disc (300 revolutions/min) with 20 radially placed slits. The scans are accumulated and stored in a register with 1,000 channels corresponding to 40 nm of spectrum. The integration time for ozone was 30 s. System control and data analysis was carried out using the Enviman Software [22]. A scheme with the main components of the DOAS is shown in Fig. 3. For comparison purposes a point-source UV ozone analyser was used (API Model 450 Ozone Analyser, Advanced Pollution Instrumentation Inc., El Sobrante, CA, USA). A Syntech Spectras gas chromatograph (Model GC855 series 600 BTX, Syntech B.V., Groningen, The Netherlands) was used to measure ambient concentrations of benzene and toluene. Both of these analysers were located on the top floor of the Science Faculty building. In addition, a small meteorological station was installed on the roof-top to measure ambient temperature, wind speed, wind direction and sunlight intensity.

Results and discussion Data from the DOAS instrument were compared with those obtained from the UV point-source analyser. While DOAS measurements were made using four monitoring paths in order to minimise local emission sources and variations, the differences in ozone concentrations were insignificant between the four paths. This is not surprising since ozone is a secondary pollutant and is expected to be reasonably well mixed in the air mass within the boundary layer. A comparison between daytime ozone concentrations obtained using the point-source analyser and the DOAS instrument is shown in Fig. 4. It can be seen that, apart from small differences in the morning values, both techniques achieved similar results. Small differences may be caused by the fact that DOAS measurements are characterising ozone over a larger urban area, described by the four paths used, while the UV instrument gives ozone concentration at the local point of the roof.

Fig. 4 Ozone concentrations obtained during daytime using a UV point-source analyser and the DOAS instrument (see text for details)

54

Fig. 5 Maximum 1-h average ozone concentrations for each day of the measurement period (03/26/03–05/02/03) Fig. 7 Ozone concentration and wind direction for the period 04/19/03–04/23/03

Fig. 6 Maximum 8-h average ozone concentrations for each day of the measurement period (03/26/03–05/02/03)

Ozone measurements obtained with the DOAS system were studied in the context of current EU legislation [2]. The present Council Directive establishes a public information threshold of 180 µg m–3 (1-h average) and an 8-h average of 110 µg m–3 for the four daily periods to be measured over the following intervals: 00:00 to 08:00, 08:00 to 16:00, 16:00 to 24:00 and 12:00 to 20:00. The results obtained from the measurement period of 38 days between 26 March 2003 and 6 May 2003 are shown in Figs. 5 and 6. The maximum 1-h average observed for each of the 38 days is shown in Fig. 5. It can be seen that during this period the limit of 180 µg m–3 was not exceeded. The maximum 8-h average for each of the 38 days is shown in Fig. 6. In this case the EU threshold of 110 µg m–3 was exceeded on a number of occasions. The period (19 May 2003–23 May 2003) during which these excesses occurred was studied in more detail. The correlation between ozone concentration and wind direction (Fig. 7) clearly shows high ozone concentrations being associated with easterly winds. When westerly winds predominate the ozone concentration is relatively low due to the influx of clean Atlantic air. This behaviour has also

Fig. 8 Ozone concentration and toluene to benzene ratio for the period 04/19/03–04/23/03

been observed during other measurement campaigns [23]. It can be concluded that the high ozone levels observed are mainly as a result of transport of ozone precursors and ozone-laden air masses from continental Europe and the United Kingdom. Further support for these conclusions comes from measurements of the toluene to benzene ratio. The source of these aromatics in urban air is predominately from motor vehicle exhausts. The ratio observed in newly formed urban air in Europe is approximately 4, similar to that found in the fuel. The dominant loss process for aromatic hydrocarbons in ambient air is reaction with hydroxyl radicals. Given that toluene is approximately six times more reactive than benzene towards hydroxyl radicals, k (OH+toluene)=7.5×10–12 cm3 molecule–1 s–1 and k (OH+benzene)= 1.2×10–12 cm3 molecule–1 s–1 [24], it is expected that the toluene to benzene ratio will decrease as the air mass

55 Acknowledgements J.A. Morales gratefully acknowledges FOCAS and DIT for the award of an Arnold Graves postdoctoral scholarship. FOCAS is funded under the National Development Plan 2000–2006 with assistance from the European Regional Development Fund. The authors also acknowledge the Journal of Analytical and Bioanalytical Chemistry, Springer and the organising and scientific committee of the Colloquium Spectroscopicum Internationale XXXIII for the award of a poster prize for the present work.

References

Fig. 9 Meteorological chart for 04/23/03 (typical of the period 04/19/03–04/23/03) for western Europe and the northeast Atlantic

ages. This is clearly illustrated in Fig. 8 where high ozone levels are associated with low toluene/benzene values, suggesting the presence of photochemically polluted air over Dublin during this period. The meteorological situation over Europe during this time period is shown in Fig. 9. The presence of an anticyclone over northern Europe resulted in a broad east to south-easterly air flow. Such meteorological conditions in summer have also been shown in previous studies to be conducive to high ground level ozone concentrations in Dublin [25].

Conclusions Urban background ozone concentrations have been measured in Dublin using a DOAS instrument and a conventional point-source analyser. Small differences in the ozone measurements are more likely to be due to location of the instruments rather than measurement method. The DOAS technique does however offer a number of important advantages over point-source analysers, such as minimal interference from local emission sources, non-contact with the air, low maintenance, less frequent calibration and multi-analyte capability. While the 1-h average ozone concentration did not exceed the EU threshold, the 8-h average did on more than one occasion. A study of the meteorological conditions and the toluene to benzene ratio during the period of these excesses clearly demonstrates that these high ozone levels were as a result of trans-boundary pollution from air masses which originated over mainland Europe. For the analysed time interval, the data suggest that the ground ozone level in Dublin might be significantly influenced by long-range transport from the United Kingdom and continental Europe.

1. Harrison RM (1996) Air pollution: sources, concentrations and measurement. In: Harrison RM (ed) Pollution: causes, effects and control. Royal Society of Chemistry Information Services, Cambridge, pp 144–168 2. Council Directive 1996/62/EC (1996) Commission of the European Communities, Brussels 3. United States Environmental Protection Agency home page (2003) http://www.epa.gov. Updated 6 May 2003. Cited 8 May 2003 4. Walters S, Ayres S (1996) The health effects of air pollution. In: Harrison RM (ed) Pollution: causes, effects and control. Royal Society of Chemistry Information Services, Cambridge, pp 253–255 5. Hazucha MJ (1987) J Appl Physiol 62:1671–1680 6. McDonnell WF, Kehri HR, Abdul-Saleem S et al (1991) Arch Environ Health 46:145–150 7. Department of Health (1991) Advisory group on the medical aspect of air pollution episodes. First report: ozone. HMSO, London 8. Thurtson GJ, Ito K, Kinney PL, Lippmann M (1992) J Expo Anal Environ Epidemiol 2:429–450 9. Lippmann M, Lioy P, Leikauf G et al (1983) Adv Mod Environ Toxicol 5 423–446 10. Lioy PL, Vollmuth TA, Lippmann M (1985) J Air Pollut Control Assoc 35:1068–1071 11. Specktor DM, Lippmann M, Lioy PJ et al (1988) Am Rev Respir Dis 137:313–320 12. Krunekreef B, Kinney PL, Ware JH, Dockery D et al (1991) Environ Health Perspect 90:189–193 13. Clean Air Trust home page. http://www.cleanairtrust.org. Updated April 2003. Cited 8 May 2003 14. Killinger DK, Mooradian A (eds) (1983) Optical and laser remote sensing techniques. Springer, Berlin Heidelberg New York 15. Measures KM (1984) Laser remote sensing. Wiley-Interscience, New York 16. Svanberg S (1990) Optical environmental monitoring. In: Inguscio M, Demtroder W (eds) Applied laser spectroscopy. Plenum, New York, pp 417–434 17. Edner H, Ragnarson P, Spanare S, Svanberg S (1993) Appl Optics 32:327–332 18. Platt U, Perner D, Patz HW (1979) J Geophys Res 84:6329– 6335 19. Platt U, Perner D (1980) J Geophys Res 85:7453–7458 20. Platt U, Perner D (1983) Measurement of atmospheric trace gases by long path differential UV/visible absorption spectroscopy. In: Killinger DK, Mooradian A (eds) Optical and laser remote sensing techniques. Springer, Berlin Heidelberg New York, pp 97–105 21. Platt U (1994) Differential optical absorption spectroscopy (DOAS). In: Sigrist MW (ed) Air monitoring by spectroscopic techniques. Wiley, New York, pp 30–36 22. Enviman Software (2001) OPSIS AB, Furulund 23. Treacy J, Morales JA, Garland WE, Sweeney J. Ozone transport studies in Ireland using differential optical absorption spectroscopy (to be published) 24. Atkinson R (1989) J Phys Chem Ref Data, Monograph No. 1, 1 25. Leavey M, Sweeney J (1990) Int J Climatol 10:191–201