Field chemical sensing with LEDs

Reliable concentration assessment of the major atmospheric oxidants (hydroxyl free radicals, nitrate radicals, and ozone, for example), and their precursors (nitrous acid, nitrogen dioxide, and formaldehyde) is essential for understanding and predicting chemical processes that affect regional air quality and global climate change. Monitoring these short-lived atmospheric constituents in real-time in situ is challenging because of their high reactivity, which results in lifetimes of around 1-100 seconds (this is short by comparison with greenhouse gases, for example). Furthermore, these short-lived oxidant species have ultra-low concentrations that measure in parts per billion by volume (ppbv) to parts per quadrillion by volume (ppqv).

Since the last decade, atmospheric environmental monitoring has benefited from the development of novel sensors with high sensitivity and specificity. These breakthroughs were made possible by significant breakthroughs in photonic technology, including greater availability of light sources at desired wavelengths, as well as high performance photodetectors that provide fast response and shot-noise-limited sensitivity. In our work, we monitored atmospheric nitrous acid (HONO) and nitrogen dioxide (NO₂) using LED-based incoherent broadband cavity-enhanced absorption spectroscopy (LED-IBBCEAS). The IBBCEAS technique¹ relies on the use of a high-finesse optical cavity to provide the necessary ultra-high sensitivity, in conjunction with a broadband light source, such as an LED or xenon arc lamp, to monitor trace gas species in the visible and UV spectral regions. LEDs, widely used today in lighting and video displays, are advantageous in optical sensing because they allow access to spectral regions that involve strong fundamental electronic transition (see Figure 1).

Figure 1. (top) Strong and structured broadband absorptions of atmospheric molecules in the UV and visible spectral regions allow for high sensitivity detection of key atmospheric species. (bottom) Characterization of some LEDs (a-c) used in IBBCEAS testing.^2–4

The LED-IBBCEAS technique offers several advantages. The ∼1m long optical cavity has an effective absorption path length of ∼1–10km, which allows for enhanced detection sensitivity while keeping the setup very compact for high spatial resolution measurement. The method allows for simultaneous quantitative assessment of multiple atmospheric species using broadband light sources. In addition, it eliminates the need for laser beam mode matching and the application of dither (applied movement) to the optical cavity, as required for cavity ring-down spectroscopy. IBBCEAS does not require the optical alignment used in off-axis integrated cavity output spectroscopy, making our apparatus more stable and robust. Furthermore, the technique is cost-effective, since it uses low-cost LEDs.

Figure 2. (left) The prototype LED-IBBCEAS instrument deployed for open-path outdoor measurement testing. An LED emission picture is shown in the inset. (right) Simultaneous concentration measurement of atmospheric nitrous acid (HONO) and nitrogen dioxide (NO₂) ². Upper panel shows measured (gray) and fit (red) absorption spectra of 3.1±0.3parts per billion by volume (ppbv) HONO and 22.2±0.5ppbv NO₂ in ambient air. The absorption coefficients of NO₂(blue)⁷ and HONO (purple),⁸ and the spectral baseline (green) from the fit are attached as reference. Lower panel shows the residual spectrum.

We demonstrated a prototype LED-IBBCEAS instrument (see Figure 2) operating at ∼365nm for simultaneous in situ open-path measurement of HONO and NO₂ in ambient air.² The instrument's sensitivity and specificity showed its potential for field observation. However, for applications in urban and coastal environments with heavy aerosol particle levels, Mie scattering by aerosols may strongly degrade the effective optical path length, and hence the detection sensitivity. We deployed the LED-IBBCEAS apparatus with a closed cavity (see Figure 3) at a suburban site of Tung Chung in Hong Kong, where our objective was simultaneous monitoring of environmental HONO and NO₂, with detection limits of 300pptv for HONO and 1ppbv for NO₂ during a 120 second acquisition period.⁵ We measured daytime and nighttime concentrations and compared these with data from instruments commercially available for routine measurements. For example, we tested the HONO concentration measurement against a long path absorption photometer (QUMA, Model LOPAP-03.)⁶ We used an NOy analyzer (TEI, Model 42CY) equipped with a blue light converter to validate the NO₂ measurement from LED-IBBCEAS. Quantitative assessments from these techniques showed similar data during day and night. Figure 3 shows time series measurements of environmental HONO and NO₂ concentrations for the period 14:20 on May 5, 2012, through 14:20 on May 6.

Figure 3. (left) The LED-IBBCEAS instrument deployed in a field inter-comparison campaign in Hong Kong. (right) Time series measurements of environmental HONO and NO₂concentrations from 14:20 on May 5, 2012, to 14:20 on May 6.

Testing LED-IBBCEAS in a real atmospheric environment demonstrated the technique's feasibility for measuring environmental HONO and NO₂ concentrations in a manner that is free of chemical and spectral interference. During the in-field campaign, we observed occasional inconsistencies in the quantitative assessment of HONO, resulting from instability of the instrument due to variations in temperature, pressure, humidity, vibration and wind. It is crucial, therefore, to provide real-time spectral baseline correction and cavity mirror reflectivity calibration in our instrument to enable successful field observation. We continue to work to address these issues to make the LED-IBBCEAS technique viable for real-world sensing.