Differential optical absorption spectroscopy (DOAS) uses the unique absorption of specific electromagnetic energy wave lengths by chemicals in the ultra violet (UV), visible (V), and near infrared (NIR) spectrum to identify and quantify individual chemicals.

The underlying operating principal in DOAS is rooted in a modification of the Beer-Lambert Law which relates the intensity of light transmitted to the intensity of light received after traveling a set distance. As stated on the University of Heidelburg's Satellite Group (2007) webpage:

The Beer-Lambert law can not directly be applied to atmospheric measurements because of several reasons:

• Besides the absorption of the trace gases also light extinction due to scattering on molecules and aerosols occurs.
• In the atmosphere the absorptions of several species always add up to the total absorption. Thus in most of the cases it is not directly possible to measure only one specific species.
• Usually the initial intensity I0 can not be measured at all or with sufficient accuracy.

These limitations can be solved by applying the method of Differential Optical Absorption Spectroscopy (DOAS). The DOAS technique relies on the measurement of absorption spectra instead of the light intensity at a single wavelength only. Thus it is possible to separate the absorption structures of several atmospheric species from each other as well as from the extinction due to scattering on molecules and aerosols. The key principle of DOAS is the separation of the absorption into a part which represents broad spectral features and in another part representing narrow spectral features.

The typical system includes a light emitter that collimates the light from a source lamp, a transmitting telescope, a telescope that collects the attenuated light beam, a spectrometer, a single or multichannel receiving/detector unit, and a control and processing computer (see Exhibit 1).

Exhibit 1. Monostatic Configuration of UV-DOAS System

Source

The open path UV-DOAS generally uses either a tungsten halogen or xenon arc lamp as a light source, although deuterium lamps can also be used. The light is collimated before being transmitted through a telescope to a receiving unit.

Optics

Unless operated in the passive mode, the system requires a sending and receiving telescope. Incoming light can be focused either directly into the spectrometer or onto a fiber optics cable where it is transmitted to the spectrometer. There are some commercial systems that use one telescope for both transmitting and receiving and are restricted to a monostatic configuration.

Spectrometer

In a typical setup, light from the telescope enters the spectrometer where a mirror collimates it and sends it to a diffraction grating. The diffracted light falls on a second mirror that focuses it on a detector for counting. The spectrometer may also utilize a multi-channel optical analyzer. The information gathered is then sent to a computer for evaluation and analysis. The user of this system should compare system spectrometers before deciding which vendor to choose as their different designs can fit some applications better than others.

Output

The background spectra for UV-DOAS is dominated by N₂ and O₂ background scattering (Rayleigh and Mie). This scattering produces broad spectral bands. Target analytes produce molecular absorption patterns that have narrow band widths. By using a mathematical subtraction algorithm, the narrow band widths can be extracted from the overall spectrum and fitted to reference spectra using least squares regression to match the spectrum with target analytes.

Deployment

UV-DOAS systems can be deployed in a monostatic, bistatic, or passive configuration. Depending on the commercial system used and the target analytes, the measured one-way path length can be up to 10 kilometers. Exhibit 1 shows a monostatic configuration where the collimated light is beamed through the area to be measured to a retroreflector that bounces it back through the measurement area to the system's receiving optics. In this configuration the path length is, in effect, doubled. In a bistatic configuration (Exhibit 2), the area to be measured is located between the sending and receiving portions of the system. In a passive mode the system uses light from the sun as the source. The passive system is often used for high-altitude measurement of atmospheric gases from platforms mounted on balloons.

Exhibit 2. Bistatic Configuration of UV-DOAS System

Although vendor literature makes claims that a wide range of organic and inorganic chemicals can be detected by UV-DOAS, the technology has found its most widespread use in detecting inorganic gases and vapors (SO_x, NO_x, NH₃, H₃S, HF), monoaromatics (BTEX), and aldehydes (HCHO). While dependent upon the system chosen, path length, deployment configuration, and meteorological conditions, among other things, detection limits found in the literature along with vendor claims indicate that UV-DOAS generally can detect the above species at the ppb to sub ppb level. Exhibit 3 presents detection limits published on the Royal Society of Chemistry webpage (http://www.rsc.org/). Oxygen is a strong interferent with the monoaromatics, and UV-DOAS cannot be used for detecting CO. However, the DOAS method is valid for CO when used in a passive mode and the light source includes near infrared wavelengths.

Exhibit 3 Detection Limits for Air Pollutants Using UV-DOAS1
Pollutants	Detection Limit µg/m3	Measurement Range µg/m3
SO2	1	0-1000
NO2	1	0-1000
O3	3	0-1000
HNO2	1	0-500
HCHO (Formaldehyde)	2	0-500
CH3CHO (Acetaldehyde)	2	0-500
Benzene	3	0-500
Toluene	3	0-500
p-m-Xylene	10	0-500
o-Xylene	3	0-500

¹ Modified from an article by Miroslav Radojevic on the Royal Society of Chemistry webpage

Multi-Axis Differential Optical Absorption Spectroscopy (MAX-DOAS)

MAX-DOAS involves taking simultaneous measurements using either passive or active techniques from a single location along various horizontal/vertical paths (axis). The measurements allow for the construction of a vertical profile of the chemicals of concern. The instrument generally consists of a spectrometer, detector, and two or more telescopes that are connected to the detector via fiber optic cable. Two examples of the profile can be found at Bales et al. (2006) and University of Bremen (2006). If a charged-coupled device (CCD) is employed as a detector, up to 10 measurements can be taken at the same time. In this configuration, each fiber optic signal is focused on one CCD row. The University of Bremen website contains a number of articles on MAX-DOAS in specific and DOAS in general.

DOAS Tomography

In land-based tomography, telescopes are used to direct light to pre-positioned retroreflectors at different heights and/or distances. Concentrations of chemicals along these paths are determined, and the data are converted into 2- or 3-D tomograms. At least two different source points are needed to construct a 3-D tomogram. The method was demonstrated by Pundt et al. (2005) to display concentrations of exhaust gases along a heavily used highway and has applications for urban air quality measurements and to calibrate/verify contaminant air transport models. Wang et al. (2005) reports using a MAX-DOAS to collect data from aircraft overflights that were converted into 2 or 3-D tomograms. This technique uses ground reflectance or scattered sunlight as the light source.

To improve the resolution of active open path DOAS tomography, researchers at the University of Heidelberg have developed a multi-beam instrument consisting of a Newtonian telescope that is capable of sending and receiving up to six different light paths with one lamp source (Mettendorf et al. 2003). The emitting light comes from small mirrors positioned near the lamp. The return light signal is coupled into different fiber optic cables that direct the light to a spectrograph connected with a CCD detector (Mettendorf et al. 2005 and Pundt and Mettendorf 2005).

Measurement and Monitoring Initiative Project Support

To support the identified need for remote sensing of airborne contaminants, EPA's Technology Innovation Program funded a UV-DOAS project (http://clu-in.org/programs/21m2/projects/6.pdf) The project, which was carried out in the Chicago area, tested the ability of the instrument to monitor fence-line concentrations of aromatic compounds during a remedial action at a landfill. The UV-DOAS data, averaged across a 200-m distance, were compared to data taken at points along the same path using canisters for collection and GC/MS for analysis. The UV-DOAS data were consistently higher than those obtained from the canisters indicating that there was at least one source that the canisters were not sampling.

Air Quality Management

The most prevalent use of UV-DOAS is in monitoring ambient air for smog-related chemicals (e.g., NO_x, SO_x, O₃, formaldehyde). Since it has a range up to ten or more kilometers it can be used to calibrate and verify Eulerian models used in air quality management. The average concentration of a variety of chemicals over a given path provides a much better approximation of a smog condition than a number of point sources. Continuous measurements provide a 24-hour look at atmospheric conditions that would be very expensive to replicate for the same suite of chemicals over the same distance with point source equipment. For specific case studies see Mathew, et al. (2001), Jimínez, et al. (2000a), and Reisinger, et al. (1996). Full citations of these with accompanying abstracts are in the 21M² literature database.

In 1999, the Maine Department of Environmental Protection (MDEP) installed a UV-DOAS system in the city of Portland to measure ozone, sulfur dioxide, nitrogen dioxide, benzene, toluene, total xylenes, formaldehyde, and phenol. The system was setup on the roof of the University of Southern Maine Library and emitted light across Interstate 295 to a receiver in a commuter parking lot. The real time measurements were provided on a web page for the public to view and for MDEP evaluation. The monitoring program ended Janaury 3 2007. Allegheny County in the Pittsburgh area is currently using UV-DOAS for air quality monitoring (ACHD 2006).

A more recent development is the use of DOAS to develop tomographic depictions of pollutant concentrations. Two 50 m towers were constructed 80 m from a major highway. Four retroreflectors were placed on each tower with 10 m spacing. One tower is 800m from the UV-DOAS instruments and the other is 650 m. Two DOAS systems took measurements of light reflected from the towers giving a total of 16 pathways (eight paths parallel to the highway and eight crossing it). The data collected were converted into 2- and 3-D concentration images that show above background concentrations of various pollutants above and down wind of the highway. As would be expected the pollutant concentrations were diluted the further above or away from the highway the measurements were taken (Pundt et al. 2005).

A monitoring system using three multi-beam DOAS instruments was installed in the city of Heidelberg, Germany, in 2005. The system uses 10 to 20 retroreflector arrays spaced around the city and up to 5 km away to construct 2-D tomograms of trace gas concentrations (Poehler et al. 2005).

Fenceline Monitoring at a Chemical Plant

Emissions of air pollutants such as benzene, sulfur dioxide, ammonia and particulate can be related to regulatory violations. In some instances, ambient concentrations of pollutants can raise public health concerns. To better understand the nature of these emissions and the risks to public health, the Air Enforcement Division (AED) operates a mobile remote sensing laboratory at facility fence lines, residential areas, and other off-site locations. The data helps inform the Agency during traditional regulatory investigations and can provide data for purposes of administrative or judicial orders under Clean Air Act section 303, Emergency powers.

AED's mobile laboratory includes an open-path ultra violet differential optical absorption spectrometer (UV DOAS), a real-time particulate matter monitor and sample collection system, and other measurement analyzers as required. Wind direction and wind speed data is also collected for purposes of data evaluation. The instrumentation provides highly time-resolved pollution data over relatively large areas, allowing for a variety of analyses for determining the source of emissions and their importance in terms of public health.

For example, UV DOAS has been used to measure benzene, a known human carcinogen, and other aromatic hydrocarbons in residential areas downwind of refineries, petrochemical plants, and other facilities suspected of violating benzene control requirements. Sulfur dioxide has also been measured downwind of combustion sources, including a coal-fired power plant in Ohio where residents and public health agencies were concerned about exposure to the plants' emissions. In a joint effort with the State of Missouri, Department of Natural Resources and public health agencies, the mobile laboratory was used to measure ammonia and particulate matter in the vicinity of a large-scale swine confined animal feeding operation (e.g., PSF/Continental Grain). In that case, the data helped to inform the need for ammonia controls and the potential impacts to public health.

Ambient measurements of source-related air pollutants has been a useful tool for identifying problem areas and determining the need for injunctive relief. AED plans to continue this program and to support EPA Regional offices and states in these special investigations.

Under a consent decree, a UV-DOAS system is being used for fence-line monitoring at the Westlake Petrochemicals facility in Westlake, Louisiana. The data generated from this system, which is measuring primarily monoaromatics are made available to the public on an Internet webpage. A special condition of the decree requires the company to provide data to any nearby resident requesting it by the next business day following the request.

Governments in Australia routinely use UV-DOAS to perform fenceline monitoring near facilities suspected of emitting excessive amounts of pollutants. For example, a DOAS system was installed and is operated at a grammar school located approximately 500 m from an oil refinery. The instrument, which has a 430 m open path, is monitored continuously. The results of the sampling indicated that benzene has exceeded action levels on several occasions, but in general, the concentrations were below levels of concern. The two exceedances occurred during light winds blowing from the direction of the refinery (EPA Victoria 2005, EPA South Australia 2003).

Schaefer et al. (2002) reports on using a mono-static DOAS system equipped with three retroreflectors to measure fugitive emissions from tanker loading operations at a river harbor. The effluent concentrations in the plume and meterological measurements were used to determine emission source strengths.

In cooperation with the Olin Corporation, EPA conducted a UV-DOAS survey at a chlor-alkali facility that used the mercury cell process to manufacture chlorine gas and sodium hydroxide. The system was deployed in a bi-static configuration on the roof of the plant's cell building. Measurements for mercury were taken across the roof vent. During the study period, approximately 472 gm/day of elemental mercury were being emitted from the vent (USEPA 2002).

Open Path Vehicle Emissions Monitoring

UV-DOAS equipment has been successfully deployed to detect vehicles that emit excessive amounts of pollutants through their mufflers. The system consists of a sending and receiving unit located on one side of a road and a retroreflector on the other. The ideal location is at a stop sign or stop light where vehicle engines will exert a high torque while accelerating from a stop thereby emitting the most pollutants. The equipment measures changes in gases as cars pass; and at the same time, a camera records the vehicle license plate. Those vehicles associated with elevated pollutant emissions are issued citations asking the owner to have the vehicle repaired and then inspected at an appropriate inspection station.

Upper Atmosphere Monitoring of Trace Gases and Pollutants

UV-DOAS instruments that operate in a passive open path mode and are housed in balloons or aircraft, have been used to investigate concentrations of pollutants in the atmosphere at various heights. For example, the THESEO2000 project measured O₃, NO₂, BrO, OClO, and H₂O from a balloon-mounted UV-DOAS. Measurements were taken at various altitudes up to the float height of approximately 30 km using direct sun spectroscopy. The direct measurement data from these balloon experiments were used to test the predictions of 3-D photochemical transport models such as SLIMCAT (Chipperfield, 1999) and Reprobus (Lefevre, et al., 1994). For more information on this project, visit http://www.nilu.no/projects/theseo2000/.

The Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY), a passive UV/VIS/NIR instrument, was launched on ENVISAT on March 1, 2002. It measures wavelengths between 240 nm and 2,380 nm. By using the DOAS method, it is possible to retrieve the atmospheric absorptions of O₃, NO₂, BrO, OClO, SO₂, HCHO, O₂, O₄, CO, CO₂, CH₄, and N₂O on a global basis. Typical spatial resolution in nadir mode is 30 by 60 km, which can be improved to 30 by 15 km (Frankenberg et al. 2003).

Monitoring Ammonia Emissions from Animal Facilities

The U.S. EPA is concerned about ammonia emissions from confined animal (hog) feeding operations. Estimates show that each hog can produce up to 8 pounds of ammonia per year. At a large-scale feeding facility this could amount to over 1,000 pounds per day. Secrest (2001) reports on the use of open path FTIR and UV-DOAS for measuring ammonia concentrations downwind from two feeding facilities. The UV-DOAS had a detection limit of 4 ppb (100 m path) and a linear range up to 993 ppb. The actual path used for the monitoring was 150 m. The distance to the first facility was 800 m, and the distance to the second facility was 400 m. At 800 meters distance, concentrations of over 700 ppb were recorded. The continuous monitoring showed wide variations in concentrations during a 24-hour period. These variations were probably due to changes in temperature and wind speed. The author concludes that having continuous contaminant concentration data coupled with meteorological data provide a much better basis for estimating potential exposure and subsequent health effects than that which can be obtained with canister style point source monitoring.

Mount et al. (2002) are studying the release of ammonia gas at the Washington State University research dairy farm using a UV-DOAS system. They have found concentrations of ammonia in the tens of ppm in the barn and concrete yard area, 100s of ppb to low ppm over the slurry lagoon area, and low ppm after application of slurry onto pasture land. The DOAS system has a detection limit of around 1 ppb for ammonia.

Uses Suggested in Vendor Literature

Vendors that manufacture UV-DOAS equipment can be located through journal announcements and on the Internet. The following list contains applications in which the vendors claim their products have been successfully deployed:

1. Urban air pollution monitoring
2. Industrial fence-line monitoring
3. Airport monitoring
4. Traffic monitoring
5. In-plant storage/process area monitoring
6. Stratospheric pollutant monitoring
7. Tropospheric pollutant monitoring

Monitoring Aromatic Hydrocarbons at a Wastewater Treatment Plant

Jimínez et al. (2000b) report on a demonstration project using UV-DOAS to monitor the emissions of benzene, toluene, ethylbenzene, xylenes, 1,2,4-trimethylbenzene, and phenol at a wastewater treatment plant in Lausanne, Switzerland. The plant runs an activated sludge process with fine bubble-diffused aeration for treating city wastewater. It also has a facility for incinerating the produced sludge. The demonstration project used an innovative method for obtaining oxygen spectra to achieve better detection limits and resolution for the monoaromatic data. A monostatic, coaxial DOAS system was used with a focal path from the roof of one building across the roof of the sludge-handling building to a retroreflector. This resulted in a maximum path length of 220 m (which is short) for obtaining the best achievable detection limits. The highest concentrations of BTEX were observed when there was a vertical (up) wind that generally occurred around noon. This circumstance indicated that vertical mixing was not occurring to any degree without wind turbulence, and might also indicate that more than one path would be needed for full characterization. The authors conclude that the system was successful in monitoring the monoaromatic emissions at low ppb levels and could be used to monitor other facilities like this one.

The EPA ETV program has tested two UV-DOAS instruments, and verification reports are available on the ETV website at (http://www.epa.gov/etv):

Opsis Inc. - AR 500 Ultraviolet DOAS Open-Path Monitor
Verification Statement (PDF, 42.7K) | Verification Report (PDF, 1.14M)

Spectrex - SafEye 420 Ultraviolet Open-Path Monitor
Verification Statement (PDF, 44 K) | Verification Report (PDF, 554K)