Open Access
Add to BibSonomyAbstract
A continuous-wave, rapidly swept cavity-ringdown spectroscopic technique has been developed for localized atmospheric sensing of trace gases at remote sites. It uses one or more passive open-path optical sensor units, coupled by optical fiber over distances of >1 km to a single transmitter/receiver console incorporating a photodetector and a swept-frequency diode laser tuned to molecule-specific near-infrared wavelengths. Ways to avoid interference from stimulated Brillouin scattering in long optical fibers have been devised. This rugged open-path system, deployable in agricultural, industrial, and natural atmospheric environments, is used to monitor ammonia in air. A noise-limited minimum detectable mixing ratio of ~11 ppbv is attained for ammonia in nitrogen at atmospheric pressure.
© 2014 Optical Society of America
1. Introduction
Cavity-ringdown spectroscopy (CRDS), using near-infrared continuous-wave (cw) tunable diode lasers and fiber-optical components, enables highly sensitive quantitative detection of optical absorption by gas-phase molecules at trace levels in air [1–5]. For a sample located inside an optical cavity, the concentration of absorbing molecules is measured via the decay rate (i.e., “ringdown”) of light from the optical cavity, with a characteristic ringdown time τ.
Our distinctive rapidly swept cw-CRDS approach, reviewed in [5], measures optical buildup and subsequent ringdown decay while rapidly varying either the length of the optical cavity or the frequency of the laser light. This results in relatively uncomplicated, compact, rugged cw-CRDS instruments, with the optical transmitter and receiver collocated in a single console that can be widely separated from one (or several) fiber-coupled ringdown-cavity unit(s). A competitive noise-equivalent absorption (NEA) detection sensitivity of 5 × 10−10 cm–1 Hz–1/2 can be realized in our swept-cavity variant of the cw-CRDS approach [6]. Furthermore, a N × 1 fiber-optical switch or combiner can be used to incorporate several pre-tuned laser sources for quasi-simultaneous multi-wavelength sensing of multiple absorbing species [7], as in our swept-cavity cw-CRDS experiments [8] that demonstrated quantitative trace-level spectroscopic sensing of greenhouse-gas mixtures such as carbon dioxide (CO2), water (H2O) vapor, and methane (CH4) at ambient concentrations in air, with multiple species simultaneously monitored within the millisecond period of a ringdown-cavity sweep cycle.
In this paper, we report for the first time several innovative technical advances and scientific applications of these rapidly swept cw-CRDS methods, as follows:
- • Remote multi-site sensing is facilitated by one (or more) fiber-coupled passive CRDS sensor unit(s) that can be situated at long distances (e.g., tens of km) from a single transmitter/receiver console, including development of ways to avoid interference from backward-propagating Brillouin scattering when the fiber length exceeds ~10 m.
- • A pre-aligned passive optical sensor unit has been developed for fiber-coupled SF cw-CRDS open-path agricultural sensing measurements in a remote ~0.5-m optical cavity.
- • Rapidly swept fiber-coupled cw-CRDS is applied to open-path trace-level sensing of ammonia gas (NH3) in air at characteristic absorption wavelengths and without conveying the NH3(g) analyte via a gas-flow system, in readiness for remote agricultural trace-level sensing of NH3 in outdoor air (e.g., from livestock wastes, soils, and vegetation).
This body of previously unreported laser-spectroscopic research has made significant progress in atmospheric sensing methodology and has potential for key environmental applications.
Within the last decade, there have been major technological advances in performance of narrowband tunable coherent optical sources [11], such as quantum cascade lasers (QCLs) [12,13], operating at relatively long infrared (IR) wavelengths (e.g., longer than ~3 µm). These are especially useful for high-resolution, ultra-sensitive spectroscopy in the mid-IR region, where most molecules have strong fundamental rovibrational absorption bands. For instance, the mid-IR fundamental absorption of NH3(g) in the vicinity of 10 µm is at least 100 times stronger than its absorption in near-IR overtone/combination bands around 1.55 µm [14]. For some spectroscopic sensing applications, therefore, QCLs have gained favor over their near-IR counterparts, such as communications-band tunable diode lasers (TDLs).
Nevertheless, such near-IR TDLs (e.g., operating at communication-band wavelengths shorter than 1.65 µm) are exclusively employed in our rapidly swept cw-CRDS experiments [5–10], in order to take advantage of convenient, efficient fiber-optical and other photonic components that are essential in fiber-coupled spectroscopic sensing systems, such as the instruments described in this paper. The prime advantage of using mid-IR lasers such as QCLs to access strong fundamental molecular absorption bands is therefore overridden in the present context by the advantage of long-distance transmission over a distributed network of single-mode optical fibers, combined with implementation of less elaborate, lower-cost near-IR TDLs. Moreover, it will be evident (e.g., from our cw-CRDS measurements of NH3 in air, reported in Sec. 3 below) that our rapidly swept cw-CRDS instrumentation has high detection sensitivity, in terms of low noise-equivalent absorption (NEA) and short electronic duty cycle; this compensates for much of the disadvantage of reduced spectroscopic line strengths in near-IR overtone/combination bands (relative to those in mid-IR fundamental bands).
2. Development of fiber-coupled swept-frequency (SF) cw-CRDS techniques
2.1 Rapidly swept fiber-coupled cw-CRDS instrumentation
The essential layout of fiber-coupled swept-frequency (SF) cw-CRDS instrumentation, on which this investigation depends, is depicted schematically in Fig. 1.
Fig. 1 Layout of an optical-fiber-coupled rapidly swept cw-CRDS instrument, based on a swept-frequency (SF) tunable laser. The gas-phase sample is located between the reflectors of the passive high-finesse ringdown cavity, to and from which coherent near-IR radiation (bold lines) is conveyed via a three-port optical circulator (OC), single-mode optical fiber, and coupling optics. Backward-propagating light reflected from the cavity is diverted via OC to a photodetector (PD1), where it is monitored in an optical-heterodyne mode of detection. Backward- and forward-propagating ringdown-signal waveforms are shown as red and blue insets, respectively. Electric field amplitudes of light at various points in the system are denoted by EL, EI, EF, and EB.
This instrument can be expanded to multiplex SF cw-CRDS operation by including an optical switch to direct the light to and from a distributed fiber-coupled sensor network comprising multiple optical cavities, each of which serves as a passive CRDS sensor unit. It can also be used to multiplex several lasers (each tuned to a molecule-specific near-IR wavelength) for quasi-simultaneous detection of multiple species [8]. The detection configuration in Fig. 1 is “single-ended,” with both optical transmitter and receiver collocated in a single console that can be widely separated from the passive ringdown-cavity unit (or units in the case of multi-site measurements) and amenable to open-path sensing applications at remote locations. A second photodetector (PD2 – not shown in Fig. 1) can also be used to monitor forward-propagating cw-CRDS light transmitted by the cavity. In contrast to our previously reported rapidly swept cw-CRDS instrument designs [5–10], that in Fig. 1 does not include a piezoelectric translator (PZT) to allow the cavity length to be varied, as is intrinsic to swept-cavity (SC) cw-CRDS. The optical cavity’s two reflectors and its coupling to fiber optics are rigidly fixed, after pre-alignment to optimize signals. Operation of the ringdown cavity is therefore completely passive, with no electrical connections (e.g., for cavity mirror control) between the single transmitter/receiver console and the fiber-coupled sensor unit(s). This approach facilitates remote open-path trace-gas sensing, which depends on having one or more rugged, compact, and cost-effective CRDS sensor units arranged in a distributed fiber-optical network. In some of the laboratory-based rapidly swept cw-CRDS experiments reported here (Secs 2.2 and 3.2), a different PZT-controlled ringdown cavity enclosed in an inert, evacuable gas cell [6] has been used to characterize our SF and SC cw-CRDS approaches prior to open-path studies (Sec. 3.3). A PZT-controlled cavity can also be used for finer resolution in SF cw-CRDS experiments by successively stepping the cavity length [10].
In SF cw-CRDS measurements, the wavelength or frequency of monochromatic tunable coherent light is swept rapidly through a cavity resonance, for which the optical bandwidth is Δνcavity = (2πτ)–1. Characteristic optical transient signals then build up with a rapidly rising leading edge, either in transmission or in reflection. SF cw-CRDS optical waveforms observed in transmission display an exponentially decaying tail (from which the characteristic ringdown time τ can be extracted) [5,9,10]. As in SC cw-CRDS [5–7,15], optical-heterodyne detected (OHD) waveforms can be observed in reflection by SF cw-CRDS [5,9,10]. These comprise a decaying envelope (with exponential decay time 2τ) of full-wave oscillations with steadily decreasing period, which can readily be demodulated electronically for efficient ringdown-time measurements [5–7,9,10,15]. Similar transient signals have been attributed to cavity-beating decay [16,17]. Comparable full-wave oscillations (but with constant period, owing to the constant offset frequency-shift interval) have been reported in frequency-switched OHD cw-CRDS measurements [18].
The mechanism of the oscillatory waveforms observed in transmission and reflection via rapidly swept cw-CRDS experiment is well understood and has been numerically simulated by ourselves ([5,7,19]) and others (e.g., [16,20–26]). These simulations show how, for a given sweep speed and a particular cavity length and finesse, the coupling efficiency of light into the cavity decreases as the cavity finesse F and/or the sweep rate increase. The times required to attain optical build-up and to record subsequent ringdown decay are traded off against the optical build-up efficiency and the precision of ringdown-decay measurements. In the case of SF cw-CRDS (on which this paper focuses), with a time-dependent laser optical frequency ν(t), the observed waveforms can be simulated in terms of the sweep rate υSF = dν(t) / dt, the laser bandwidth Δνlaser, and the free spectral range FSR and finesse F of the cavity [5].
The rapidly swept cw-CRDS approaches on which we focus are based on a passive cavity in which build-up of successive ringdown waveforms is effectively self-triggered and detected when the laser frequency comes into resonance during the sweep cycle. This contrasts with more elaborate cavity-locked CRDS techniques [26–29] which underlie various commercially available cw-CRDS gas analyzer systems [30]. Our rapidly swept cw-CRDS approaches are more flexible in the context of fiber-coupled and multi-site sensing applications and are more amenable to open-cavity gas sensing at remote locations. The open-cavity approach is advantageous in the case of “sticky” molecules such as NH3(g), where analyte gas tends to be held up in the flow systems needed to access the enclosed cavity of a typical cavity-locked CRDS instrument. For instance, recent papers on NH3 breath analysis using such commercial CRDS instruments [31–33] have reported precautions needed to deal with adsorption and desorption, sample humidity, and memory effects that can influence instrumental response times and quantitative accuracy.
The performance of a cw-CRDS system depends critically on the short-term frequency stability of its laser source. Rapid laser-frequency fluctuations during the build-up time result in CRDS signal fluctuations, reduced signal-to-noise ratio, and degraded detection sensitivity. In our rapidly swept CRDS systems (e.g., as in Fig. 1), the build-up time is typically a fraction of the ringdown time τ, which corresponds to a few μs. Laser sources with short-term frequency stability better than ~100 kHz are therefore ideal. Compact, rugged, low-cost distributed-feedback (DFB) lasers are preferred, but we find [8] that more elaborate external-cavity diode lasers (ECDLs) provide superior short-term frequency stability, yielding ten-times better CRDS detection sensitivity than DFB lasers. Two communications-band ECDLs have been used in this work: New Focus 6262 laser head with 6200 controller (tunable at ~1.51–1.59 μm; ~5-mW output power); Photonetics Tunics-Plus (~1.50–1.64 μm; ~3 mW).
Other key optical components of fiber-coupled SF cw-CRDS instruments used here (e.g., as in Fig. 1) are as in our previous reports [6–10]. For measurements of NH3(g), these include: a pair of high-reflectivity (R) concave cavity mirrors (with typical reflectivity R > 0.9999 at ~1.55 μm with 1-m radius of curvature and 1° wedge angle); single-mode optical fiber, with FC/APC connectors; a three-port fiber-optical circulator OC; an additional 30-dB optical isolator to further protect the ECDL from residual optical reflections; a fast photodetector PD1 (New Focus model 1617, 800-MHz bandwidth), used in reflection for the OHD mode of operation; a low-noise sensitive photodetector PD2 (New Focus model 2053, 10-MHz bandwidth), used in the transmission mode not shown in Fig. 1. For multi-site SF cw-CRDS experiments, additional 1 × N fiber-optical switches can be included to provide quasi-simultaneous access to multiple passive CRDS sensor units on a distributed fiber-optical network (not shown in Fig. 1). Cavity-ringdown waveforms are recorded by a fast digitizer (National Instruments, model PCI-5122, 14 bit, 100-MHz bandwidth). Additional analog-to-digital data collection channels can record various experimental conditions such as laser scan voltage and optical transmission of a reference gas cell. Customized LabVIEW software was developed for instrument control, recording of spectra and data analysis. Typically, a single SF cw-CRDS data point (corresponding to an elapsed time of ~1 s) is derived from 1024 decay waveforms, merged in groups of 32 for τ-fitting and then averaged over 32 such fits.
2.2 Remote sensing by fiber-coupled swept-frequency cw-CRDS
Our fiber-coupled SF cw-CRDS measurements use large lengths (e.g., ≥1 km) of single-mode optical fiber so that passive ringdown-cavity sensor unit(s) can be located far away from the single transmitter/receiver console. It can be advantageous to use the OHD mode of operation in reflection (via photodetector PD1, as in Fig. 1), with efficient demodulation of the resulting full-wave optical-signal oscillations [5–7,9,10,15]. Figure 2 shows results from such an OHD instrument. However, using fiber-optical lengths exceeding ~10 m and an incident laser power as low as ~0.5 mW, we find that stimulated Brillouin scattering (SBS) [34] in the long optical fiber becomes a serious problem. As shown in Fig. 2(d), the backward-propagating OHD light tends to be obscured by accompanying light due to SBS, which is inevitably generated in the optical fiber and co-propagates back with the OHD light to PD1. Figure 2 shows results from an OHD instrument as in Fig. 1. These OHD SF cw-CRDS measurements were made in single sweeps of laser frequency (at wavelength λlaser ≈1.57 µm) from a New Focus 6263 ECDL, using an empty linear ringdown cavity with fixed length d = 0.531 m and laser-frequency sweep rate υSF ≈0.13 MHz µs–1. Figures 2(a) and 2(b) show a decaying envelope of full-wave oscillations recorded with a short optical-fiber length L = 10 m; the OHD CRDS waveforms are consistent with a ringdown time τ = 18 µs, a cavity finesse F ≈3.2 × 104, and a mirror reflectivity R ≈0.9999. Figure 2(c) is a plot of the logarithmic demodulation of the signal envelope shown in Fig. 2(a), efficiently performed by analog electronics [5–7,9,10,15]; its linear slope is proportional to a decay time of 2τ. Figure 2(d) was recorded under the same conditions as for Fig. 2(b), but with L increased 100-fold, from 10 m to 1 km; the CRDS waveforms are now swamped by SBS interferences.
Fig. 2 Fiber-coupled SF cw-CRDS waveforms measured in reflection, using a ringdown cavity with FSR = 282 MHz and υSF ≈0.13 MHz µs–1. In Figs. 2(a)–2(c), the fiber length L = 10 m, whereas L = 1 km in Fig. 2(d), where the CRDS waveforms are obscured by SBS interferences.
Figure 2(d) shows that, if the length L of the single-mode optical fiber shown in Fig. 1 exceeds ~10 m, typical laser powers of a few mW will generate unacceptably high feedback from SBS in that fiber. To avoid this problem, we use a pair of single-mode optical fibers – one to carry the SF ECDL radiation from the transmitter/receiver console to the ringdown cavity unit and the other to return the SF cw-CRDS signal for detection without SBS interference, e.g., as in the OHD cw-CRDS layout shown in Fig. 3.
Fig. 3 Layout of a SF cw-CRDS instrument, fiber-coupled over long distances, which uses a second optical fiber and optical circulator (OC2) to return backward-propagating light from the remote cavity to photodetector PD1 in the transmitter/receiver console, where it is monitored in OHD mode. This layout separates OHD cw-CRDS signal light (the waveform of which is inset in red) from backward-propagating SBS light (depicted by grey arrows) which is diverted by OC1 to a beam dump, thereby avoiding SBS interferences that arise in long optical fibers.
The twin-fiber instrument depicted in Fig. 3 retains the original OHD approach of Fig. 1. A second optical circulator (OC2) returns backward-propagating OHD cw-CRDS signal light from the ringdown cavity to photodetector PD1 within the transmitter/receiver console. The OHD light of interest (with electric field amplitude EI + EB, color-coded red) is thereby separated from backward-propagating SBS light (depicted by grey arrows), which is diverted by OC1 to a beam dump. We have verified experimentally that the twin-fiber OHD cw-CRDS approach of Fig. 3 is indeed free from SBS that is generated by laser light in long optical fibers (e.g., with L > 10 m). We note that the short length (less than a few meters) of optical fiber between OC2 and the ringdown cavity does not generate significant SBS in the OHD cw-CRDS signal light that is returned via OC2 and a second long fiber to the transmitter/receiver console. This is as in our earlier OHD measurements [5,9,10] where shorter optical-fiber lengths (< 10 m) were used without generating any noticeable SBS interferences. For OHD measurements, the laser bandwidth Δνlaser needs to be small enough for full-wave OHD-beat oscillations in the SF cw-CRDS signal to be resolved. This is generally true for ECDL-type lasers, but (as discussed in Sec. 2.1) not for less stable DFB-type lasers.
In this paper, we have adopted an alternative approach based on detection of the forward-propagating transmitted cw-CRDS signal, to tolerate greater laser-frequency instability and to take advantage of higher sensitivity of the low-noise photodetector PD2 (rather than the faster time-response of photodetector PD1 that is needed in the OHD mode). Figure 4 depicts this alternative SBS-free twin-fiber-coupled approach; corresponding results are shown in Fig. 5.
Fig. 4 Alternative layout of a long-fiber-coupled SF cw-CRDS instrument for remote sensing. A second optical fiber and extra coupling optics are used to return forward-propagating cw-CRDS light transmitted by the remote cavity to photodetector PD2 in the transmitter/receiver console; this avoids SBS interferences that are generated by laser light in long (e.g., >10 m) optical fibers. A corresponding signal waveform is inset in blue. The backward-propagating light reflected from the cavity, accompanied by SBS light (labelled in grey), is diverted via OC to a beam dump.
Fig. 5 SF cw-CRDS waveforms measured by using a twin-fiber-coupled forward-propagating instrument as in Fig. 4, with a pair of long (L = 1 km) optical fibers; an exponential-decay fitting curve is superimposed on plot (a). These waveforms are seen to be free of SBS interference.
The instrumental layout shown in Fig. 4 uses a second set of coupling optics to collect cw-CRDS light transmitted by the cavity (with electric field amplitude EF, color-coded blue) and returns it to the photodetector PD2 located within the transmitter/receiver console. The backward-propagating OHD light (with electric field amplitude EI + EB, in red) returns, with the unwanted interfering SBS light, along the original fiber to the transmitter/receiver console, where it is diverted via optical circulator (OC1) to a beam dump. In this forward-propagating variant of a twin-fiber SF cw-CRDS instrument, the transmitted signal is not affected by SBS interferences because back-scattered SBS does not propagate in the forward direction and so does not enter the ringdown cavity.
The results shown in Fig. 5 were obtained with an instrumental layout as in Fig. 4, using a pair of long (L = 1 km) optical fibers and other experimental conditions (such as υSF ≈0.13 MHz µs–1 and sweep ranges) similar to those of Fig. 2. The forward-propagating waveforms are cleanly resolved in Figs. 5(a) and 5(b), in contrast to the OHD waveforms in Fig. 2(d) that are obscured by SBS interference. As in Fig. 2, the waveforms in Fig. 5 were recorded in single frequency sweeps at λ ≈1.57 µm; the optical power was 1.5 mW (three times that in Fig. 2). An exponential fit to the single ringdown decay in Fig. 5(a) yields τ = 18 µs; the typical standard deviation for τ averaged over 103 consecutive ringdown events is ± 0.02 µs. As already explained, rapid fluctuations in the laser frequency during the build-up time cause ringdown decays to deviate from an ideal exponential profile. This contribution dominates the final uncertainty and detection sensitivity of our CRDS measurements.
We have also successfully used the twin-fiber forward-propagating CRDS method of Fig. 4 over much longer distances L, using a pair of ~20-km spools of single-mode optical fiber, offering fresh fiber-coupled sensing prospects over distances of tens of km.
2.3 Instrumentation for fiber-coupled measurement of CH4(g), CO2(g), and H2O(g) emissions
Methane is an important greenhouse gas, contributing significantly to radiative forcing and hence anthropogenic global warming. It has attracted numerous recent near-IR spectroscopic studies (e.g., [8,29,35–39]). In the last 200 years, globally averaged concentrations of CH4(g) have more than doubled to 1799 ± 2 ppbv in 2010 [40]; much of this increase can be ascribed to anthropogenic sources of CH4(g) and a reduced tropospheric removal rate. In the context of the CSIRO Flagship Livestock Methane Research Cluster (LMRC) [41], we are motivated to develop efficient fiber-coupled cw-CRDS instrumentation for open-range sensing of CH4(g) emissions from ruminant livestock grazing pastures and rangelands. The LMRC aims “to develop accurate and practical methods to measure and reduce livestock methane emissions in the northern Australia beef herd” in view of the significant (>5%) contribution to Australia’s greenhouse-gas emissions from beef cattle in open-range settings. We have therefore examined ways to incorporate fiber-coupled cw-CRDS sensor units with localized, passive, open-path ringdown cavities into field-based eddy-covariance instruments [42] able to monitor fluxes of CH4(g), CO2(g), and H2O(g) together with relevant meteorological and other spectroscopic data. Recent multi-instrument investigations by the LMRC group [41] at the CSIRO Chiswick field station near Armidale, NSW (~500 km north of Sydney) have focused on emission of CH4(g) from herds of beef cattle under controlled conditions.
In our previous multi-species cw-CRDS study [8] of greenhouse-gas molecules (CO2, H2O, CH4) in outdoor air, tailored multi-wavelength radiation from a set of pre-tuned TDLs was used to achieve the following minimum detectable mixing ratios: ~2 ppmv CO2, ~100 ppmv H2O, and ~65 ppbv CH4. Partial pressures of CO2, H2O, and CH4 in ambient outdoor air were measured [5,8] to be: 0.285 Torr (380 ppmv), 10.5 Torr (1.4 × 104 ppmv), and 1.7 mTorr (2.3 ppmv), respectively; note that the partial pressure of H2O previously reported [5,8] has now been corrected from 10.5 mTorr to 10.5 Torr. Our previous cw-CRDS studies of CH4(g) [5,8] were performed on a relatively weak characteristic spectral feature at 1635.414 nm [39]. However, more sensitive spectral features of CH4(g) are beyond the 1640-nm operating limit of ECDLs that are available to us at present. In future SF cw-CRDS measurements of CH4 in air, we propose to use a commercially available ECDL with high stability at longer wavelengths around 1650 nm, in order to access the ~3-times-stronger 1650.956-nm or 1653.725-nm absorption features of CH4(g) [39] and attain a projected minimum detectable mixing ratio of ~5 ppbv (superior to our current CRDS limit of ~65 ppbv for CH4 in air [5,8]).
As part of the LMRC collaboration [41], we have developed specially designed open-path ringdown-cavity sensor units for remote SF cw-CRDS measurements localized within a ~0.5-m optical cavity; these units can be mounted near standard eddy-covariance meteorological and CH4(g) sensors. The SF cw-CRDS sensor units are rugged, pre-aligned, and completely passive, with no electrical cables or moving parts; their only connections are by twin optical fibers (e.g., as in Fig. 4) and by a gas line to allow the cavity-mirror surfaces to be flushed. During field tests at CSIRO Chiswick, three such fiber-coupled, passive, open-path SF cw-CRDS sensors were mounted (e.g., on eddy-covariance masts) at long distances (≥75 m) from the single transmitter/receiver console located in an instrumentation shed. However, reliable SF cw-CRDS measurements of CH4(g) could not be made at that time, mainly because the frequency stability of the DFB-type lasers employed for those measurements was inadequate. As explained above, further open-path environmental sensing of CH4(g) by fiber-coupled SF cw-CRDS awaits acquisition of a commercially available ECDL that is sufficiently stable at longer wavelengths around 1650 nm. Meanwhile, our research focus has shifted to sensing of NH3(g), as discussed in Sec. 3 where further design and performance details are presented.
3. Rapidly swept fiber-coupled cw-CRDS sensing of NH3 in air
3.1 Context of CRDS-based sensing of NH3 in air
The presence of NH3(g) in our natural, industrial, and agricultural, atmospheric environments is widely recognized [43,44] and of international concern, particularly in agriculture [45] which contributes “about two thirds of global NH3 emissions” [44]. Although NH3(g) itself is not a direct cause of global warming (given its relatively short lifetime in the atmosphere), it is regarded as an “indirect greenhouse gas” owing to its subsequent interchanges with soil and vegetation, including conversion to the significant direct greenhouse gas, nitrous oxide (N2O) [45], of which the 100-year global warming potential is ~300 and ~9 times that of CO2 and CH4, respectively. Efforts in the Australian agricultural context therefore aim to understand and control total atmospheric emissions of nitrogen (N), generally as NH3(g) and N2O(g). For instance, in managing feedlots for livestock such as beef cattle, current issues include the contribution of NH3(g) from sources such as manure to acidification and eutrophication of sensitive ecosystems, degradation of air quality (e.g., by formation of secondary particulate matter via atmospheric reactions of NH3 with sulfate or nitrate) and its impacts on animal and human health, and the effect of directly and indirectly evolved N2O(g) on global greenhouse-gas concentrations. More than 70% of the crude-protein dietary intake of beef cattle is recycled to the atmosphere as NH3(g) and N2O(g) [46], but estimates from biological models differ considerably relative to emissions measured using various instrumental methods.
We are therefore preparing our fiber-coupled rapidly swept cw-CRDS technique to complement conventional methods, such as open-path Fourier-transform infrared (FTIR) spectrometry, open-path near-IR laser absorption, eddy covariance, and chemiluminescence gas analysis, in measurements of NH3(g) emissions from managed herds of beef cattle at an agricultural field station. As already explained, our cw-CRDS approach has advantages over established measurement methods, including its amenability to use of multiple passive sensor units that are fiber-optically linked remotely to a single optical transmitter/receiver console.
CRDS is just one of many ways to measure concentrations of NH3 in air, including eleven instrumental methods, the agricultural field-based performance of which has been evaluated in detail [47]. Many of these approaches depend on gas-transfer systems, in which the NH3(g) analyte is conveyed from the open-air sensing point(s) to an enclosed-instrument detection system (e.g., via lengths of heated PTFE tubing). Quantitative detection of “sticky” molecules (e.g., NH3) is subject to problems such as adsorption and desorption, sample humidity, and slow instrument response times. Such gas-transfer problems apply to various forms of cavity-enhanced spectroscopy, and CRDS in particular, used for “enclosed-cavity” measurements of NH3(g) concentration [31–33,48]. Of the open-path IR-spectroscopic methods that avoid gas-transfer problems, NH3(g) concentration measurements by FTIR [47,49] and long-path IR laser absorption [47,49–51] spectrometry need to be averaged over a relatively long column of air in the optical path between the instrument and its retro-reflector. More localized measurements of NH3(g) concentrations can be made by open-path forms of cavity-enhanced IR spectroscopy, facilitated by the relative compactness of optical-cavity dimensions. The few examples of such an approach to localized open-path detection of NH3 in air by IR-laser spectroscopy include cavity-enhanced absorption [52] (enclosed in an environmental chamber rather than truly in the open air) and ~9-µm QCL field-based absorption spectroscopy using multi-pass Herriott cells (rather than a resonant optical cavity) [53,54]. As is demonstrated in Sec. 3.2, our fiber-coupled rapidly swept cw-CRDS approach can attain a noise-limited minimum detectable mixing ratio of ~11 ppbv of NH3 in N2(g) – highly competitive with sub-ppmv detection levels previously obtained [52–54] in other localized open-path measurements of NH3 in air. The amenability of fiber-coupled rapidly swept cw-CRDS to open-cavity measurements of NH3 in environmental air at remote locations therefore has distinct novelty.
As explained above, our fiber-coupled, open-path rapidly swept cw-CRDS method offers advantages relative to conventional measurement approaches that entail transfer of NH3(g) over lengths of tubing between the sensing points and the instrumental console. It is therefore proposed in future field trials to add our open-path fiber-coupled cw-CRDS technique to the range of instruments that are currently deployed to measure environmental concentrations and fluxes of NH3(g) and to compare the cw-CRDS readings to those of the established devices.
3.2 Fiber-coupled cw-CRDS measurements of NH3 in N2(g) at 1 atm enclosed in a gas cell
The survey spectra in Fig. 6, covering the near-IR wavelength range of ~1520–1550 nm, establish suitable characteristic near-IR laser-absorption wavelengths for reliable detection of NH3(g). To enhance trace-level sensing of NH3 in air, these need to correspond to relatively high absorption line strengths and to be as free as possible from spectroscopic interferences due to other atmospheric species such as CO2(g) and H2O(g).
Fig. 6 Near-IR swept-cavity cw-CRDS absorption scan (a) for ~40 ppmv of NH3 in N2(g) at 1 atm, measured by a tunable ECDL, compared (b) with a stick spectrum from published data [55]. Asterisks mark two prominent absorption features of NH3(g) at 1522.5-nm and 1531.7-nm.
Figure 6(a) comprises a swept-cavity cw-CRDS absorption spectrum, for ~40 ppmv of NH3(g) in N2(g) at a total pressure of 1 atm in a PZT-controlled ringdown-cavity gas cell composed of stainless steel and PTFE [6]. The spectrum in Fig. 6(a) was recorded by slowly scanning the output wavelength of the Photonetics Tunics-Plus tunable ECDL, while the length of the ringdown cavity was rapidly dithered at ~1 kHz (over a small fraction of its FSR). Figure 6(b) shows a comparable stick spectrum based on high-resolution near-IR FTIR spectra for pure NH3(g) [55]. Comparison with spectra [39] of CO2(g) and H2O vapour indicates that the prominent 1522.5-nm and 1531.7-nm absorption features of NH3(g) are reasonably free of interference when measured in air. However, trace-level sensing of NH3 in air requires allowance for the wings of adjacent absorption features of H2O(g) and for weak underlying absorption lines of CO2(g) [39]. This is verified in (one of many) previous near-IR TDL-spectroscopic studies of NH3 in air [56], as is the choice of 1522.5-nm and 1531.7-nm NH3(g) features in Fig. 6. The latter wavelength region has been used routinely in our subsequent near-IR cw-CRDS studies of NH3 in air; relevant IR spectra are shown in Fig. 7.
Fig. 7 Higher-resolution cw-CRDS spectra (a, b, e) of NH3(g) accompanied by reference spectra, all recorded at ~1531.7 nm by slowly scanning the laser frequency. Trace (a) is SC cw-CRDS spectrum, whereas traces (b) and (e) are SF cw-CRDS spectra; all three are recorded for a mixture of 9 ± 0.5 ppmv of NH3 in N2(g) at a total pressure of 1 atm. The reference spectra (for a mixture of NH3 and CH4, each at 100 Torr in a sealed quartz cell) are recorded (c) in simple transmission mode, simultaneously with (a), and (d) in WMS mode, simultaneously with (b). Trace (e), plotted at ~1 data point/s, is further discussed in the text.
These higher-resolution IR spectra of NH3(g) at ~1531.7 nm have been recorded by using a New Focus 6263 tunable ECDL and the same PZT-controlled ringdown-cavity gas cell, with a pair of 1-km fiber spools added between the ringdown-cavity unit and the optical transmitter/receiver console to establish a rapidly swept cw-CRDS system configuration similar to that depicted in Fig. 4. Examples of various measurements are shown in Fig. 7. Traces (a) and (b) are SC and SF cw-CRDS spectra, respectively, for ~9 ppmv of NH3 in N2(g) at a total pressure of 1 atm; trace (a) was recorded under SC conditions (cavity-sweep rate of ~0.13 MHz µs–1 and range of ~0.25 FSR) that are comparable to typical SF conditions as specified in Secs. 2.1 and 2.2. The accompanying traces (c) and (d) comprise reference spectra for a mixture of NH3(g) and CH4(g), each at 100 Torr in a sealed quartz cell, recorded respectively in simple transmission and wavelength-modulation spectroscopy (WMS) modes.
This WMS facility has been combined (for the first time, as far as we know) with the SF cw-CRDS instrument to provide a reliable, convenient way to set and hold the tunable laser at a desired peak wavelength for long-term measurement of gas absorption (e.g., by NH3 or CH4). Figures 7(a) and 7(c) were recorded simultaneously in a slow scan of the ECDL frequency. Likewise, to simultaneously record SF cw-CRDS and WMS plots, as in Figs. 7(b) and 7(d), the ECDL center frequency ν0laser was scanned slowly while the time-dependent ECDL frequency ν(t) was rapidly dithered with υSF ≈0.13 MHz µs–1, an amplitude of ~2½ ringdown-cavity FSRs (~0.7 GHz), and a sinusoidal modulation frequency of ~100 Hz. The inset in Fig. 7(b) magnifies small steps that are intrinsic to our mode of SF cw-CRDS detection [9,10]; of the 2 or 3 discrete cavity resonance frequencies (separated by the cavity FSR) in the rapidly dithered frequency range of ν(t), that closest to the slowly scanned ECDL center frequency ν0laser is sampled. The recorded spectroscopic data points therefore change in steps from one cavity resonance frequency to the next; the step size (FSR = 282 MHz) of these cavity resonance frequencies is small relative to the spectral width of the absorption feature at atmospheric pressure. The precise cavity resonance frequencies depend on the cavity length d and the refractive index of the intra-cavity gaseous medium, so that the measured frequency of the absorption peak is ± 0.5 FSR with respect to the nearest cavity resonance frequency. This uncertainty in spectral sampling frequencies therefore results in an estimated uncertainty of ± 2% in the absorption coefficient value at the peak of the spectral profile (and hence in the derived NH3 concentration). Under the conditions of Doppler and pressure broadening that prevail in Fig. 7, the absorption feature of NH3(g) at ~1531.7 nm is expected [56] to appear as a doublet with peaks separated by ~12 GHz (~0.4 cm–1) in the 35-GHz scans (a)–(d), for which the abscissa origins are set to match the more prominent peaks of the CRDS plots (a) and (b). Figure 7(e) presents SF cw-CRDS signals at ~1531.5–1531.7 nm in a sequence where the scan of ECDL center frequency ν0laser is at first off, then on (data points 0 to ~540), then off again; abscissa data-point units correspond to an elapsed time of ~1 s.
Relative to lower-pressure reference spectra (c) and (d), the N2-broadened peaks at ~1531.7 nm in Figs. 7(a) and 7(b) are shifted to lower frequency by ~0.7 GHz. This indicates an effective pressure-shift coefficient of approximately –0.8 GHz atm–1 for NH3 in N2(g) (comparable to other results [57] at ~1516 nm); simulated Voigt-profiles indicate that the components of the ~1531.7-nm feature have different pressure-broadening and pressure-shift coefficients. SC and SF cw-CRDS scans in Figs. 7(a) and 7(b) were recorded on successive days, during which loss of NH3(g) from the statically filled cell [6] (e.g., by adsorption inside the cell) is very small. Using available line strengths [56], our Voigt-profile simulations of the spectroscopic features in Fig. 7 establish the relationship between CRDS signal amplitudes and mixing ratios of NH3 in N2(g) or air: a peak absorption coefficient of 1.0 × 10−6 cm–1 corresponds to ~3.5 ppmv of NH3 in N2 at 1 atm. In Fig. 7, the estimated mixing ratio of NH3 in N2(g) is 9 ± 0.5 ppmv, consistent with our multi-step mixture dilution procedures.
The left-hand portion of Fig. 7(e) shows a ~4-minute low-absorption measurement with λlaser fixed at ~1531.5 nm. The central portion of Fig. 7(e) shows the spectrum recorded for ~9 ppmv of NH3 in N2(g) during ~9 minutes as λlaser is scanned over ~20 GHz from ~1531.5 nm towards the more prominent peak (at ~1531.7 nm) in Figs. 7(a) and 7(b). The scan of λlaser is then stopped and the right-hand portion (data points >540) shows a ~7½-minute peak-absorption measurement. The two fixed-λlaser sections of Fig. 7(e) show magnified insets of noise levels from which standard deviations of the ringdown times τ and the NEA sensitivity can be derived. The average absorption coefficient at the ~1531.7-nm peak (for a mixing ratio of ~9 ppmv of NH3 in 1 atm of N2 gas) is ~2.5 × 10−6 cm–1; its standard deviation is ± 6 × 10−9 cm–1, which is about double that of the baseline in the left-hand part of Fig. 7(e). Significantly, that SF cw-CRDS baseline noise indicates that we have attained a NEA detection sensitivity of 3 × 10−9 cm–1 Hz–1/2 (for the CRDS instrument itself in this wavelength region) and a noise-limited minimum detectable mixing ratio of ~11 ppbv for NH3 in N2(g).
The performance of our fiber-coupled, open-cavity, near-IR cw-CRDS instrument is therefore competitive with that of an early enclosed-cavity, QCL-based cw-CRDS system used to study NH3 in N2(g) at ~8.5 µm [58]. This is consistent with the ~100-fold absorption cross-section advantage [14] of QCL-based mid-IR fundamental spectroscopy of NH3(g) over near-IR overtone and combination bands around 1.55 µm as in our fiber-coupled, open-path cw-CRDS applications. More recently, sub-ppbv detection limits for NH3 in air are advertised for high-performance near-IR “wavelength-scanned” (WS) CRDS gas analyzer systems that are commercially available [30,59] (e.g., as used in [31–33,47]). We believe that our SF cw-CRDS detection sensitivity could be enhanced substantially by using a laser with better short-term frequency stability, by incorporating higher-reflectivity mirrors for longer ringdown times τ, and by detecting stronger spectral absorption features. Meanwhile, our fiber-coupled open-cavity SF cw-CRDS approach has distinct advantages over higher-sensitivity enclosed-cavity CRDS instruments in applications that entail detection of “sticky” molecules (such as NH3, where lengths of heated PTFE tubing are needed to transfer sampled gas to the instrument) and for monitoring of trace gases in one or more remote locations.
3.3 Fiber-coupled open-path SF cw-CRDS sensing of NH3 in air
The cw-CRDS spectra of NH3 in N2(g) in Figs. 6(a), 7(a), 7(b), and 7(e) were recorded in a PZT-controlled, evacuable, inert ringdown-cavity gas cell [6] that was statically filled with sample mixtures prepared from high-purity cylinder gases. This is necessary to calibrate sensitivity and verify instrumental performance, as a precursor to field-based measurements of NH3(g) (e.g., from livestock wastes or fertilized soil and vegetation) in open-air settings.
Three manifestations of an open-path SF cw-CRDS ringdown-cavity sensor unit for such environmental gas-sensing applications are illustrated in Fig. 8. Its design, as shown in Fig. 8(a), was developed in the course of our LMRC collaboration [41] on field-based sensing of CH4(g) emissions from livestock, as depicted in Fig. 8(b) and outlined in Sec. 2.3.
Fig. 8 A passive, pre-aligned open-path ringdown-cavity sensor unit mounted for environmental sensing (a) in the laboratory, (b) on an eddy-covariance tower, and (c) in a campus garden. A pair of yellow-clad single-mode optical fibers connect the sensor unit, via free-space coupling optics, to and from the optical transmitter/receiver console that can be located far away (>1 km in experiments reported here). Gas lines to circulate filtered mirror-cleansing air are also visible.
The sensor units are applicable to trace-level remote sensing of NH3(g) in outdoor air, as illustrated in Fig. 8(c). They can be ruggedly pre-aligned and operated passively, with no electrical cables or control elements; their only connections are twin single-mode optical fibers (as in Fig. 4) and gas lines to pass filtered mirror-flush air from an electrically powered circulation system – either a single circulator at the centrally located transmitter/receiver console or, for very long optical fiber separations, with individual circulation systems (e.g., solar-powered) adjacent to each remote sensor unit. The open-path ringdown-cavity structure shown in Fig. 8 comprises a rigid four-rod frame joining two solid end-pieces to which the two highly reflective mirrors are attached (each inside a cylindrical chamber to circulate flush gas to keep the mirrors dust-free). The pressure inside these chambers is maintained slightly below that of ambient air, to prevent spillage of purge gas into the open path. There has been no sign of any contamination of the cavity mirrors after being in operation for weeks. The cavity mirrors are separated by a distance d = 430 mm (i.e., FSR = 350 MHz), but the combined depth of the chambers reduces the effective optical pathlength deff through absorbing gas to ~360 mm, as determined by measuring the physical dimensions with an estimated tolerance of ± 2 mm. The measured CRDS absorption coefficient must therefore be scaled up by a factor of 1.19 ( = 430/360) when estimating concentrations from CRDS measurements when cavity mirrors are protected by purge gas. After careful alignment (which remains stable over many days) and with mirror coatings optimized to detect NH3(g) at λlaser ≈1532 nm, typical ringdown times are τ ≈40 µs, indicating mirror reflectivity R ≈0.99996. Free-space coupling optics (each comprising two mirrors and two lenses, one set of which is on the right-hand side of the Fig. 8(a)) are symmetrically mounted near each end of the cavity structure; these connect the sensor unit to the twin optical fibers that convey light from/to the remotely located optical transmitter/receiver console (separated by large fiber lengths L >1 km in our work).
Prior to environmental sensing of NH3(g) emissions, laboratory-based experiments were performed, using the open-path ringdown-cavity sensor unit with a pair of 1-km fiber spools carrying light from/to the optical transmitter/receiver console. Figure 9(a) shows genuine open-path measurements, affected by gas-flow and apparent ± 20% concentration fluctuations, while Fig. 9(b) corresponds to an experiment in which a more stable ( ± 5%) concentration of NH3(g) in air is established inside a long (~0.3-m) open-ended tube, aligned horizontally on the ringdown-cavity axis and with a vertical side-arm containing a steady source of NH3(g) by spontaneous dissociation of ammonium bicarbonate at room temperature (~21 °C).
Fig. 9 SF cw-CRDS measurements continuously recorded at ~1531.7 nm, monitoring NH3(g) in air under two sets of experimental conditions, as explained in text below. Open-path measurements (a) of NH3(g) in air display rapid natural fluctuations which are substantially reduced when the analyte gas is confined (b) in an open-ended intra-cavity tube.
Several ways to generate trace-level concentrations of NH3(g) in laboratory air within the ringdown cavity were explored, including those of Figs. 9(a) and 9(b), as follows:
- A. NH3 cylinder gas released upwind of the ringdown-cavity sensor unit – see Fig. 9(a)
A metered flow of 1% NH3 in helium from a gas cylinder was released from a row of pinholes in a horizontal tube ~0.25 m away from the horizontal axis of the open-path ringdown-cavity sensor unit and swept transversely through the cavity by laboratory air from a domestic electric fan. This is intended to simulate conditions in the field with a localized source of NH3(g) upwind from the sensor. Representative results (recorded over a ~10 minute period) are shown in Fig. 9(a), with the ~1531.7-nm peak CRDS signal at low and high fan speeds (preceded and followed by a zero-flow baseline). These exhibit significant fluctuations, comprising approximately ± 20% of the amplitude. The measured average absorption coefficient is (1.4 ± 0.3) × 10−6 cm–1 at the lower fan speed (~2 m s–1), which corresponds to a mixing ratio of ~6 ± 1 ppmv of NH3 in 1 atm of air. At higher fan speed (~3 m s–1), the mixing ratio is correspondingly reduced by ~30%, since the NH3 mixes with a larger volume of air as it is swept across the ringdown cavity.
- B. Volatilization of NH3(g) from aqueous ammonia – not displayed in Fig. 9
The volatilization of NH3 in aqueous ammonia solution from NH4OH(aq) to NH3(g) is an environmentally significant equilibrium process that is favored by alkaline conditions (high pH). We measured the ~1531.7-nm CRDS signal for NH3(g) volatilized from concentrated aqueous ammonia solution (originally 28%) in a beaker placed near the open-path ringdown cavity. Successive overnight runs displayed a steady evaporative decline in the detected mixing ratio of NH3(g) in air, after averaging over pronounced short-term fluctuations in the CRDS signal level.
- C. Thermal decomposition of ammonium bicarbonate – not displayed in Fig. 9
Ammonium bicarbonate, NH4HCO3(s), decomposes thermally (e.g., above ~36 °C) to form equimolar proportions of NH3(g), CO2(g), and H2O(g). The plots of ~1531.7-nm CRDS signal for NH3(g) were obtained by heating 0.8 g (0.01 mol) of NH4HCO3(s) in a small ceramic evaporating basin held above a candle located below the center point of the open-path ringdown cavity. The detected mixing ratio of NH3(g) in air dropped to zero after ~5 minutes, by which time the solid had completely disappeared. We note that, in the absence of the evaporating basin, the CRDS signal was completely insensitive to particles or thermal refractive-index gradients from the candle plume. As in item B, the instantaneous mixing ratio of NH3(g) in air exhibits sporadic fluctuations.
- D. Room-temperature decomposition of ammonium bicarbonate – see Fig. 9(b)
At room temperature, NH4HCO3(s) decomposes sufficiently to establish a steady-state concentration of NH3(g), together with CO2(g) and H2O(g), by confinement inside the ~0.3-m open-ended intra-cavity tube. In this way, the ~1531.7-nm CRDS signal for NH3(g) was measured over a period of >24 hours, an 11-hour portion of which is shown in Fig. 9(b). Compared to open-path measurements A–C and E, short-term fluctuations in the mixing ratio of NH3(g) in air are much smaller. Spikes in Fig. 9(b) are tentatively attributed to occasional intrusions of large dust particles into the intra-cavity beam path. Elsewhere in the plot, fluctuations are reduced to ± 5% of the recorded SF cw-CRDS signal amplitude; the measured average absorption coefficient is (1.9 ± 0.1) × 10−6 cm–1 which corresponds to a mixing ratio of 8.0 ± 0.4 ppmv for NH3(g) in air.
- E. Urease-catalyzed reaction of urea – not displayed in Fig. 9
The widely used fertilizer urea, (NH2)2CO(s), reacts with water and is enzymatically catalyzed by urease to evolve 2 mol of NH3(g) and 1 mol of CO2(g) from 1 mol of urea. A slurry of ~10 g each of urea, garden lime, and water plus a few mg of urease in a beaker was placed below the center point of the open-path ringdown cavity, thereby generating NH3(g) in air, as a precursor of subsequent fertilized-garden experiments. The ~1531.7-nm CRDS signals for NH3(g) that were recorded again fluctuated markedly as in B and C.
All of the experiments performed used the SF cw-CRDS transmitter/receiver instrument set up with λlaser maintained at the ~1531.7-nm peak of NH3(g), as established in Fig. 7. The CRDS signal (as well as intensity transmitted by the reference cell, cavity-mode stability, etc.) were continuously recorded. In each case, the CRDS-detected fraction of the overall quantity of NH3(g) actually generated is dependent on various environmental factors such as wind velocity and confinement of analyte gas in the cavity.
Our measurements, including those presented in Figs. 9(a) and 9(b), persistently display rapid natural fluctuations in the apparent concentration of NH3(g) at trace levels in laboratory air. Similar fluctuations are found in other localized open-path optical sensing methods (e.g., WMS [42]). True open-path conditions (A) yield fluctuations amounting to ± 20% of the recorded SF cw-CRDS signal amplitudes (and are worse in cases B, C, and E). Such fluctuations have resisted all efforts to eliminate them and are attributed to actual rapid variations of the local concentration of NH3(g) in laboratory air, owing to incomplete mixing of NH3(g) when it is emitted from a point source into ambient air. The cw-CRDS signal fluctuations are substantially reduced in experiments, as in Figs. 9(a) and 9(b), where steadier NH3(g) concentrations were established (a) by using a fan to entrain NH3(g) in air that is swept through the ringdown cavity and (b) by confining the gas inside a long open-ended tube centered on the ringdown-cavity axis. These experiments discredit the notion that the fluctuations might arise from an instrumental artefact. We have further verified that our observed NH3(g) concentration fluctuations are real by making other open-path SF cw-CRDS measurements of ambient CO2(g) and H2O(g) in air, where the source of absorbing species is spatially extended so that mixing is complete and effectively constant concentrations are observed.
Finally, we report experiments using the open-path ringdown-cavity sensor unit mounted adjacent to a Macquarie University campus garden, as previously depicted in Fig. 7(c). Representative results are shown in Fig. 10. In these experiments, the sensor in the garden area is remotely separated by a pair of long optical fibers with L ≈1.15 km (each comprising a 150-m optical cable plus a 1-km fiber spool) from the optical transmitter/receiver console in a nearby building. The source of NH3(g) in outdoor air was an aqueous slurry of ~0.5 kg of urea and garden lime with ~20 mg of urease mixed with ~0.5 kg of aged cow manure on a ~0.2-m2 bed of straw in the garden, close (less than 1 m) to the open-path ringdown-cavity sensor unit. The open-path SF cw-CRDS signals in Fig. 10 were recorded continuously with the ECDL tuned to the ~1531.7-nm absorption peak of NH3(g): (a) a brief measurement over ~20 minutes (with an ensuing baseline obtained by temporarily covering the fertilized garden area to block emissions of NH3) and (b) a 4-hour portion of a much more extensive measurement.
Fig. 10 Open-path SF cw-CRDS measurements (plotted in blue) continuously recorded to sense NH3(g) emissions in outdoor air at ~1531.7-nm (a) over ~20 minutes with ensuing baseline and (b) during a 4-hour period. The red plot in Fig. 10(b) is a 300-point (~5-minute) running average (magnified ten-fold on the ordinate scale) of the rapid natural fluctuations in the instantaneous mixing ratio of NH3(g) in air. As shown in Fig. 7(c), the open-path ringdown-cavity sensor unit was mounted adjacent to a campus garden. The sensor is remotely separated by twin optical fibers (total length L ≈1.15 km) from the optical transmitter/receiver console in a nearby building. The source of NH3(g) was fertilizer on a ~0.2-m2 bed of straw in the garden (see text).
The blue plots in Fig. 10 exhibit pronounced fluctuations in observed signals for NH3(g) in outdoor air. As in cases B, C, and E above, we take these to correspond to actual short-term variations in the local mixing ratio of NH3(g) in air along the axis of the open-path ringdown cavity, incompletely mixed after emission from a relatively localized (~0.2 m2) source. A 300-point (~5-minute) running average of these rapid natural fluctuations, with 10 × magnification on the ordinate scale, is shown in the accompanying red plot in Fig. 10(b) which yields a 4-hour-averaged absorption coefficient of 0.075 × 10−6 cm–1 and a correspondingly averaged mixing ratio of 300 ppbv, with observed natural variations of ± 0.025 × 10−6 cm–1 and ± 100 ppbv, respectively, for NH3 generated in outdoor air. These observed natural variations far exceed our instrumental uncertainties of ~3 × 10−9 cm–1 and ~11 ppbv, respectively, derived for NH3 in N2(g) in the context of Fig. 7 above.
Our rugged, pre-aligned instrument can be operated continuously without adjustment for many days. The long single-mode optical fiber works reliably and is unaffected by ambient temperature cycles or other environmental factors (e.g., mechanical vibrations). The measurements were carried out in fine weather at a relatively sheltered site. Our system has not yet been operated under rainy conditions. We intend in future to make further weather-proofing refinements prior to proposed measurements in agricultural field environments, but that work is beyond the scope of this paper. More extensive measurements could be made by adding meteorological sensors (e.g., of wind velocity) under eddy-covariance conditions, typically with a 10-Hz data rate [42,60] which is readily attainable by our SF cw-CRDS instruments.
4. Conclusions
We have refined versatile highly sensitive rapidly swept cw-CRDS techniques [5–10,15,19] for remote, localized, open-path near-IR spectroscopic sensing of environmentally significant molecules at trace levels in air. As outlined in Sec. 1, this has entailed various innovative technical advances: fiber-coupled swept-frequency techniques, enabling CRDS sensor units to operate passively at remote sites; cavity-enhanced absorption spectroscopy by rapid dithering of the laser frequency ν(t) while slowly scanning its central value ν0laser; long-distance sensing at fiber-coupled multiple sites ≥1 km away from a single transmitter/receiver console; ways to avoid interference from stimulated Brillouin scattering when the fiber length exceeds ~10 m; a rugged pre-aligned passive optical sensor design for fiber-coupled open-path measurements localized within a ~0.5-m optical cavity (e.g., for agricultural field applications).
Our rapidly swept fiber-coupled cw-CRDS approach has been applied to localized open-path spectroscopic sensing of NH3(g) in air. Because the open-path structure of the cavity sensor probes the air sample directly without needing to divert the analyte-in-air sample to an enclosed detection system, it avoids problems from adsorption, etc. that complicate reliable detection of “sticky” molecules such as NH3. In laboratory-based experiments on NH3 in N2(g) (Figs. 6 and 7), we have identified suitable spectroscopic features for unambiguous detection of NH3(g) and established quantitative relationships between the CRDS-measured absorption coefficient (of NH3 in N2(g) or air) and the corresponding mixing ratio. We have demonstrated a noise-limited minimum detectable mixing ratio of ~11 ppbv for NH3 in N2(g) at atmospheric pressure. Using assorted ways (e.g., as in Fig. 9) to generate NH3(g) in air, we have explored appropriate localized open-path CRDS techniques for field-based sensing, including possible ways to address troublesome signal fluctuations due to incomplete mixing and natural instability in the local mixing ratio for NH3 in air. SF cw-CRDS measurements of NH3(g) emissions in outdoor air have been made using a rugged open-path ringdown-cavity sensor unit (Fig. 8) mounted adjacent to a fertilized area of a campus garden, remotely separated by a pair of optical fibers (L ≈1.15 km) from the optical transmitter/receiver console in a nearby building. From these experiments (Fig. 10), we have derived an average mixing ratio of ~300 ppbv for NH3 volatilization in outdoor air, in readiness for future agricultural sensing of NH3 in air (e.g., from livestock wastes, soils, and vegetation).
Acknowledgments
We acknowledge financial and collegial support from the CSIRO Flagship Livestock Methane Research Cluster (LMRC) [41] and from the National Basic Research Program of China (973 Program, Project No. 2013CB632803). Dr Yabai He is also associated with the National Measurement Institute, Sydney, Australia.
References and links
1. K. W. Busch and M. A. Busch, eds., Cavity-Ringdown Spectroscopy: an Ultratrace-Absorption Measurement Technique, Vol. 720 of ACS Symposium Series (Am. Chem. Soc., 1999).
2. G. Berden, R. Peeters, and G. Meijer, “Cavity ring-down spectroscopy: experimental schemes and applications,” Int. Rev. Phys. Chem. 19(4), 565–607 (2000). [CrossRef]
3. B. A. Paldus and A. A. Kachanov, “An historical overview of cavity-enhanced methods,” Can. J. Phys. 83(10), 975–999 (2005). [CrossRef]
4. G. Berden and R. Engeln, eds., Cavity Ring-Down Spectroscopy: Techniques and Applications (Wiley-Blackwell, 2009).
5. B. J. Orr and Y. He, “Rapidly swept continuous-wave cavity-ringdown spectroscopy,” Chem. Phys. Lett. 512(1-3), 1–20 (2011). [CrossRef]
6. Y. He and B. J. Orr, “Detection of trace gases by rapidly-swept continuous-wave cavity ringdown spectroscopy: pushing the limits of sensitivity,” Appl. Phys. B 85(2-3), 355–364 (2006). [CrossRef]
7. Y. He and B. J. Orr, “Rapidly swept, continuous-wave cavity ringdown spectroscopy with optical heterodyne detection: single- and multi-wavelength sensing of gases,” Appl. Phys. B 75(2-3), 267–280 (2002). [CrossRef]
8. Y. He, R. Kan, F. V. Englich, W. Liu, and B. J. Orr, “Simultaneous multi-laser, multi-species trace-level sensing of gas mixtures by rapidly swept continuous-wave cavity-ringdown spectroscopy,” Opt. Express 18(19), 20059–20071 (2010). [CrossRef] [PubMed]
9. Y. He and B. J. Orr, “Rapid measurement of cavity ringdown absorption spectra with a swept-frequency laser,” Appl. Phys. B 79(8), 941–945 (2004). [CrossRef]
10. Y. He and B. J. Orr, “Continuous-wave cavity ringdown absorption spectroscopy with a swept-frequency laser: rapid spectral sensing of gas-phase molecules,” Appl. Opt. 44(31), 6752–6761 (2005). [CrossRef] [PubMed]
11. I. T. Sorokina and K. L. Vodopyanov, eds., Solid-State Mid-Infrared Sources (Springer, 2003).
12. F. Capasso, R. Colombelli, C. Gmachl, T. L. Myers, M. S. Taubman, R. M. Williams, C. G. Bethea, K. Unterrainer, H. Y. Hwang, D. L. Sivco, A. Y. Cho, A. M. Sergent, H. C. Liu, and E. A. Whittaker, “Quantum cascade lasers: ultrahigh-speed operation, optical wireless communication, narrow linewidth, and far-infrared emission,” IEEE J. Quantum Electron. 38(6), 511–532 (2002). [CrossRef]
13. R. F. Curl, F. Capasso, C. Gmachl, A. A. Kosterev, B. McManus, R. Lewicki, M. Pusharsky, G. Wysocki, and F. K. Tittel, “Quantum cascade lasers in chemical physics,” Chem. Phys. Lett. 487(1-3), 1–18 (2010). [CrossRef]
14. M. S. Taubman, P. M. Aker, D. C. Scott, M. D. Wojcik, B. D. Cannon, J. T. Munley, T. L. Myers, V. T. Nguyen, C. A. Bonebrake, and J. F. Schultz, “Long wave infrared cavity enhanced sensors,” Pacific Northwest National Laboratory Report PNNL-15103 (October 2004), Fig. 2.1, p. 2.2. http://www.pnl.gov/main/publications/external/technical_reports/PNNL-15103.pdf
15. Y. He and B. J. Orr, “Optical heterodyne signal generation and detection in cavity ringdown spectroscopy based on a rapidly swept cavity,” Chem. Phys. Lett. 335(3-4), 215–220 (2001). [CrossRef]
16. J. Ye, “Ultrasensitive high resolution laser spectroscopy and its application to optical frequency standards,” Ph. D. thesis (University of Colorado, 1997); note Fig. 6.2, p. 121.
17. J. Ye, L.-S. Ma, and J. L. Hall, “Using FM methods with molecules in a high finesse cavity: a demonstrated path to 10−12 absorption sensitivity,” Chap. 15 in K. W. Busch and M. A. Busch, eds., Cavity-Ringdown Spectroscopy: an Ultratrace-Absorption Measurement Technique, Vol. 720 of ACS Symposium Series (Am. Chem. Soc., 1999), pp. 233–253; note Fig. 3, p. 241.
18. M. D. Levenson, B. A. Paldus, T. G. Spence, C. C. Harb, R. N. Zare, M. J. Lawrence, and R. L. Byer, “Frequency-switched heterodyne cavity ringdown spectroscopy,” Opt. Lett. 25(12), 920–922 (2000). [CrossRef] [PubMed]
19. Y. He and B. J. Orr, “Ringdown and cavity-enhanced absorption spectroscopy using a continuous-wave tunable diode laser and a rapidly swept optical cavity,” Chem. Phys. Lett. 319(1-2), 131–137 (2000). [CrossRef]
20. J. Poirson, F. Bretenaker, M. Vallet, and A. Le Floch, “Analytical and experimental study of ringing effects in a Fabry–Perot cavity. Application to the measurement of high finesses,” J. Opt. Soc. Am. B 14(11), 2811–2817 (1997). [CrossRef]
21. M. J. Lawrence, B. Wilke, M. E. Husman, E. K. Gustafson, and R. L. Byer, “Dynamic response of a Fabry–Perot interferometer,” J. Opt. Soc. Am. B 16(4), 523–532 (1999). [CrossRef]
22. J. W. Hahn, Y. S. Yoo, J. Y. Lee, J. W. Kim, and H.-W. Lee, “Cavity ringdown spectroscopy with a continuous-wave laser: calculation of coupling efficiency and a new spectrometer design,” Appl. Opt. 38(9), 1859–1866 (1999). [CrossRef] [PubMed]
23. J. Morville, D. Romanini, M. Chenevier, and A. Kachanov, “Effects of laser phase noise on the injection of a high-finesse cavity,” Appl. Opt. 41(33), 6980–6990 (2002). [CrossRef] [PubMed]
24. B. Bakowski, L. Corner, G. Hancock, R. Kotchie, R. Peverall, and G. A. D. Ritchie, “Cavity-enhanced absorption spectroscopy with a rapidly swept diode laser,” Appl. Phys. B 75(6-7), 745–750 (2002). [CrossRef]
25. M. Mazurenka, A. J. Orr-Ewing, R. R. Peverall, and G. A. D. Ritchie, “Cavity ring-down and cavity enhanced spectroscopy using diode lasers,” Ann. Rep. Prog. Chem. C: Phys. Chem. 101, 100–142 (2005). [CrossRef]
26. J. H. van Helden, R. Peverall, and G. A. D. Ritchie, “Cavity enhanced techniques using continuous wave lasers,” Chap. 2 in G. Berden and R. Engeln, eds., Cavity Ring-Down Spectroscopy: Techniques and Applications (Wiley-Blackwell, 2009), pp. 27–56.
27. T. G. Spence, C. C. Harb, B. A. Paldus, R. N. Zare, B. Willke, and R. L. Byer, “A laser-locked cavity ring-down spectrometer employing an analog detection scheme,” Rev. Sci. Instrum. 71(2), 347–353 (2000). [CrossRef]
28. B. G. Fidric, R. A. Provençal, S. M. Tan, E. R. Crosson, A. A. Kachanov, and B. A. Paldus, “Bananas, explosives and the future of cavity ring-down spectroscopy,” Opt. Photon. News 14(7), 24–29 (2003). [CrossRef]
29. E. R. Crosson, “A cavity ring-down analyzer for measuring atmospheric levels of methane, carbon dioxide, and water vapor,” Appl. Phys. B 92(3), 403–408 (2008). [CrossRef]
30. For example: Picarro, Inc., (Sunnyvale, CA 94085, USA); http://www.picarro.com/gas_analyzers
31. G. Neri, A. Lacquaniti, G. Rizzo, N. Donato, M. Latino, and M. Buemi, “Real-time monitoring of breath ammonia during haemodialysis: use of ion mobility spectrometry (IMS) and cavity ring-down spectroscopy (CRDS) techniques,” Nephrol. Dial. Transplant. 27(7), 2945–2952 (2012). [CrossRef] [PubMed]
32. F. M. Schmidt, O. Vaittinen, M. Metsälä, M. Lehto, C. Forsblom, P.-H. Groop, and L. Halonen, “Ammonia in breath and emitted from skin,” J Breath Res 7(1), 017109 (2013). [CrossRef] [PubMed]
33. O. Vaittinen, F. M. Schmidt, M. Metsälä, and L. Halonen, “Exhaled breath biomonitoring using laser spectroscopy,” Current Analyt. Chem. 9(3), 463–475 (2013). [CrossRef]
34. A. B. Ruffin, “Stimulated Brillouin scattering: an overview of measurements, system impairments, and applications,” in P. A. Williams and G. W. Day, eds., Technical Digest: Symposium on Optical Fiber Measurements,2004, NIST Special Publication 1024 (National Institute of Standards and Technology, 2004), pp. 23–28; http://www.nist.gov/manuscript-publication-search.cfm?pub_id=31763
35. B. L. Fawcett, A. M. Parkes, D. E. Shallcross, and A. J. Orr-Ewing, “Trace detection of methane using continuous wave cavity ring-down spectroscopy at 1.65 μm,” Phys. Chem. Chem. Phys. 4(24), 5960–5965 (2002). [CrossRef]
36. R. Kan, W. Liu, Y. Zhang, J. Liu, M. Wang, D. Chen, J. Chen, and Y. Cui, “Large scale gas leakage monitoring with tunable diode laser absorption spectroscopy,” Chin. Opt. Lett. 4, 116–118 (2006).
37. R. Kan, W. Liu, Y. Zhang, J. Liu, M. Wang, D. Chen, J. Chen, and Y. Cui, “A high sensitivity spectrometer with tunable diode laser for ambient methane monitoring,” Chin. Opt. Lett. 5, 54–57 (2007).
38. C. Wang, N. Srivastava, B. A. Jones, and R. B. Reese, “A novel multiple species ringdown spectrometer for in situ measurements of methane, carbon dioxide, and carbon isotope,” Appl. Phys. B 92(2), 259–270 (2008). [CrossRef]
39. L. S. Rothman, I. E. Gordon, Y. Babikov, A. Barbe, D. C. Benner, P. F. Bernath, M. Birk, L. Bizzocchi, V. Boudon, L. R. Brown, A. Campargue, K. Chance, E. A. Cohen, L .H. Coudert, V. M. Devi, B. J. Drouin, A. Fayt, J.-M. Flaud, R. R. Gamache, J. J. Harrison, J.-M. Hartmann, C. Hill, J. T. Hodges, D. Jacquemart, A. Jolly, J. Lamouroux, R. J. Le Roy, G. Li, D. A. Long, O. M. Lyulin, C. J. Mackie, S. T. Massie, S. Mikhailenko, H. S. P. Müller, O. V. Naumenko, A. V. Nikitin, J. Orphal, V. Perevalov, A. Perrin, E. R. Polovtseva, C. Richard, M. A. H. Smith, E. Starikova, K. Sung, S. Tashkun, J. Tennyson, G. C. Toon, Vl. G. Tyuterev, and G. Wagner, “The HITRAN2012 molecular spectroscopic database,” J. Quant. Spectrosc. Radiat. Transfer 130, 4–50 (2013); see also earlier editions of the HITRAN database (e.g., 2008 and 2004).
40. S. Kirschke, P. Bousquet, P. Ciais, M. Saunois, J. G. Canadell, E. J. Dlugokencky, P. Bergamaschi, D. Bergmann, D. R. Blake, L. Bruhwiler, P. Cameron-Smith, S. Castaldi, F. Chevallier, L. Feng, A. Fraser, M. Heimann, E. L. Hodson, S. Houweling, B. Josse, P. J. Fraser, P. B. Krummel, J.-F. Lamarque, R. L. Langenfelds, C. Le Quéré, V. Naik, S. O’Doherty, P. I. Palmer, I. Pison, D. Plummer, B. Poulter, R. G. Prinn, M. Rigby, B. Ringeval, M. Santini, M. Schmidt, D. T. Shindell, I. J. Simpson, R. Spahni, L. P. Steele, S. A. Strode, K. Sudo, S. Szopa, G. R. van der Werf, A. Voulgarakis, M. van Weele, R. F. Weiss, J. E. Williams, and G. Zeng, “Three decades of global methane sources and sinks,” Nat. Geosci. 6(10), 813–823 (2013). [CrossRef]
41. The CSIRO Flagship Livestock Methane Research Cluster (LMRC) is outlined at http://www.csiro.au/lmrc
42. D. McDermitt, G. Burba, L. Xu, T. Anderson, A. Komissarov, B. Riensche, J. Schedlbauer, G. Starr, D. Zona, W. Oeschel, S. Oberbauer, and S. Hastings, “A new low-power, open-path instrument for measuring methane flux by eddy covariance,” Appl. Phys. B 102(2), 391–405 (2011). [CrossRef]
43. M. A. Sutton, J. W. Erisman, F. Dentener, and D. Möller, “Ammonia in the environment: from ancient times to the present,” Environ. Pollut. 156(3), 583–604 (2008). [CrossRef] [PubMed]
44. C. R. Flechard, R.-S. Massad, B. Loubet, E. Personne, D. Simpson, J. O. Bash, E. J. Cooter, E. Nemitz, and M. A. Sutton, “Advances in understanding, models and parameterizations of biosphere-atmosphere ammonia exchange,” Biogeosciences 10(7), 5183–5225 (2013). [CrossRef]
45. O. T. Denmead, D. Chen, D. Rowell, Z. Loh, J. Hill, S. Muir, D. W. T. Griffith, T. Naylor, M. Bai, F. Phillips, and S. McGinn, “Gaseous nitrogen emissions from Australian cattle feedlots,” Chapter 3 in M. A Sutton, K. E. Mason, L. J. Sheppard, H. Sverdrup, R. Haeuber, and W. K. Hicks, eds., Nitrogen Deposition, Critical Loads and Biodiversity (Springer, 2013); ISBN 978–94–007–7938–9.
46. B. Arnon, N. Hutchings, F. P. Vinther, P. K. Nielsem, H. D. Poulsen, I. S. Kristensen, and S. Pietrzak, “Analysis of methodologies for calculating greenhouse gas and ammonia emissions and nutrient balances,” in J. Selenius, L. Baudouin, and A. M. Kremer, eds., Eurostat Methodologies and Working Papers; ISSN 1977–0375 (European Union, 2011). [CrossRef]
47. K. von Bobrutzki, C. F. Braban, D. Famulari, S. K. Jones, T. Blackall, T. E. L. Smith, M. Blom, H. Coe, M. Gallagher, M. Ghalaieny, M. R. McGillen, C. J. Percival, J. D. Whitehead, R. Ellis, J. Murphy, A. Mohacsi, A. Pogany, H. Junninen, S. Rantanen, M. A. Sutton, and E. Nemitz, “Field inter-comparison of eleven atmospheric ammonia measurement techniques,” Atmospheric Measurement Techniques 3(1), 91–112 (2010). [CrossRef]
48. A. A. Kachanov, E. R. Crosson, and B. A. Paldus, “Tunable diode lasers: expanding the horizon for laser absorption spectroscopy,” Opt. Photonics News 16(7), 44–50 (2005). [CrossRef]
49. E. D. Thoma, R. C. Shores, E. L. Thompson, D. B. Harris, S. A. Thorneloe, R. M. Varma, R. A. Hashmonay, M. T. Modrak, D. F. Natschke, and H. A. Gamble, “Open-path tunable diode laser absorption spectroscopy for acquisition of fugitive emission flux data,” J. Air Waste Manag. Assoc. 55(5), 658–668 (2005). [CrossRef] [PubMed]
50. Y. He, Y. Zhang, W. Liu, R. Kan, and H. Xia, “Atmospheric ammonia monitoring near Beijing National Stadium from July to October in 2008 by open-path TDLAS system,” in F. Amzajerdian, C-Q. Gao, T.-Y. Xie, eds., International Symposium on Photoelectronic Detection and Imaging 2009: Pt. 1 – Laser Sensing and Imaging, Proc. SPIE 7382, 73821L/1–73821L/7 (2009).
51. Y. He, Y. Zhang, L. Wang, K. You, Y. Gao, A. Zhu, and W. Yang, “An ammonia sensor with high sensitivity in farmland based on laser absorption spectroscopy technology,” in B. Culshaw, ed., Advanced Sensor Systems and Applications V, Proc. SPIE 8561, 85610X/1–8 5610X/7 (2012).
52. R. Peeters, G. Berden, A. Apituley, and G. Meijer, “Open-path trace gas detection of ammonia based on cavity-enhanced absorption spectroscopy,” Appl. Phys. B 71(2), 231–236 (2000). [CrossRef]
53. M. C. Phillips, M. S. Taubman, B. E. Bernacki, B. D. Cannon, J. T. Schiffern, and T. L. Myers, “Design and performance of a sensor system for detection of multiple chemicals using an external cavity quantum cascade laser,” in M. Razeghi, R. Sudharsanan, and G. J. Brown, eds., Quantum Sensing and Nanophotonic Devices VII, Proc. SPIE 7608, 76080D/1–76080D/11 (2010).
54. D. J. Miller, K. Sun, L. Tao, M. A. Khan, and M. A. Zondlo, “Open-path, quantum cascade laser-based sensor for high resolution atmospheric ammonia measurements,” Atmos. Meas. Tech. Discuss. 6, 7005–7039 (2013); http://www.atmos-meas-tech-discuss.net/6/7005/2013/ [CrossRef]
55. L. Lundsberg-Nielsen, F. Hegelund, and F. M. Nicolaisen, “Analysis of the high-resolution spectrum of ammonia (14NH3) in the near-infrared region, 6400–6900 cm–1,” J. Mol. Spectrosc. 162(1), 230–245 (1993). [CrossRef]
56. M. E. Webber, D. S. Baer, and R. K. Hanson, “Ammonia monitoring near 1.5 µm with diode-laser absorption sensors,” Appl. Opt. 40(12), 2031–2042 (2001). [CrossRef] [PubMed]
57. C. L. Bell, M. Dhib, G. Hancock, G. A. D. Ritchie, J. H. van Helden, and N. J. van Leeuwen, “Cavity enhanced absorption spectroscopy measurements of pressure-induced broadening and shift coefficients in the υ1 + υ3 combination band of ammonia,” Appl. Phys. B 94(2), 327–336 (2009). [CrossRef]
58. B. A. Paldus, C. C. Harb, T. G. Spence, R. N. Zare, C. Gmachl, F. Capasso, D. L. Sivco, J. N. Baillargeon, A. L. Hutchinson, and A. Y. Cho, “Cavity ringdown spectroscopy using mid-infrared quantum-cascade lasers,” Opt. Lett. 25(9), 666–668 (2000). [CrossRef] [PubMed]
59. Note in particular Picarro’s G2103 and G1103 near-IR WS-CRDS (Wavelength-Scanned Cavity Ring Down Spectroscopy) NH3 analyzers; see https://picarro.app.box.com/shared/ffey0e0mo8 and http://www.picarro.com/assets/docs/NH3_analyzer_datasheet.pdf, respectively.
60. G. Burba and D. Anderson, A Brief Practical Guide to Eddy Covariance Flux Measurements (Li-COR Biosciences, 2010); http://www.licor.com/env/products/eddy_covariance/ec_book.html
