FRONTIERS ARTICLE
Rapidly swept continuous-wave cavity-ringdown spectroscopy

https://doi.org/10.1016/j.cplett.2011.05.052Get rights and content

Abstract

Continuous-wave (cw) cavity-ringdown (CRD) spectroscopy provides a highly sensitive way to measure optical absorption by observing the decay rate of light from a high-finesse optical cavity containing the sample of interest (usually gas-phase molecules). In rapidly swept cw-CRD spectroscopy, optical build-up and subsequent ringdown decay are initiated by rapidly sweeping the cavity length or the wavelength of the monochromatic tunable cw laser radiation, thereby establishing and interrupting optical resonance between the laser light and the longitudinal-mode frequencies of the cavity. We review the experimental methodology and applications of this technique, indicating its advantages and prospects for spectroscopic sensing.

Highlights

► Continuous-wave cavity-ringdown is able to measure weak optical absorption spectra. ► Rapidly swept cavity length or laser frequency initiates optical buildup and decay. ► Cavity-excitation time needs to be much shorter than cavity-ringdown decay time. ► Highly sensitive noise-equivalent absorption of ∼5 × 10−10 cm−1 Hz−1/2 is realised. ► High-resolution spectra, spanning a wide wavelength range, can be rapidly recorded.

Introduction

The cavity-ringdown (CRD) technique was first used for spectroscopy in 1988 [1] with pulsed lasers, which led to many significant early applications [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], (e.g., to laser-ablated molecules cooled in supersonic beams [2], [5], [6], [8], [9]). Continuous-wave (cw) lasers, on which this article focuses, were not employed for CRD spectroscopy [12], [13], [14], [15], [16], [17] until almost 10 years after O’Keefe’s pioneering work [1]. Reviews [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43], [44] show that CRD spectroscopy is now well established and widely used. The sample of interest (which usually comprises gas-phase molecules) is contained in a high-finesse optical cavity into which coherent (e.g., laser) light is introduced. This light is allowed to build up in intensity and then to decay away (or ‘ring down’) with a characteristic exponential time constant τ once the build-up process is interrupted, by one means or another.

The ringdown time τ depends on the reflectivity of the cavity mirrors and additional intra-cavity optical losses (due to processes such as absorption and scattering). At particular wavelengths λ, optical absorption causes the ringdown rate constant τ−1 to increase relative to that for an absorber-free cavity or for that at non-absorbing wavelengths. This enables optical absorption by constituents of the sample medium inside the cavity to be observed with high sensitivity and photometric accuracy.

The CRD approach is a form of cavity-enhanced laser spectroscopy, stemming from the realization (by Jackson [45] and Kastler [46], soon after the discovery of lasers 50 years ago) that particularly high laser-spectroscopic sensitivity could be achieved by placing an absorbing medium of interest inside a Fabry–Perot optical cavity which was external to the cavity of the laser itself. Assorted forms of cavity-enhanced laser spectroscopy have subsequently evolved, yielding remarkably low noise-equivalent detection sensitivity limits (for example, ∼10−14 cm−1 Hz−1/2 from the NICE-OHMS technique [47], [48]).

Such cavity-enhanced techniques benefit from the fact that, for light resonating in an optical cavity with high finesse F, the equivalent effective absorption pathlength is enhanced (and like wise the detection sensitivity or contrast) by a factor G = 2F/π [47], [48]. Optical interaction over extremely long effective pathlengths x (up to tens of kilometers) can be attained by using highly reflective, low-loss cavity reflectors. For a linear cavity with mean reflectivity R = (R1R2)1/2, in terms of the reflectivities R1 and R2 of its two mirrors, the finesse F is given by πR1/2/(1  R), which reduces to π(1  R)−1 for the very highly reflective mirrors that are used in cavity-enhanced spectroscopy. Optical cavities with mean reflectivities R = 0.9990, 0.99990, and 0.999990 therefore yield pathlength enhancement factors G of 2 × 103, 2 × 104, and 2 × 105, respectively.

An additional advantage of CRD spectroscopy is that it measures temporal decay of radiation inside the cavity, rather than transmitted optical power or energy as in conventional absorption spectroscopy. This allows the CRD approach to be insensitive to amplitude fluctuations of the tunable coherent light. Generally, the CRD decay of the intensity of monochromatic radiation inside the cavity is of a simple exponential form:I(t,λ)=I(0,λ)exp[-t/τ(λ)],where intensities observed when the optical build-up process is interrupted and at a later time t are denoted by I(0,λ) and I(t,λ), respectively.

Spectroscopic information at a particular wavelength λ is derived from the total round-trip intra-cavity optical loss Ltotal, which comprises the sum of a wavelength-dependent optical-absorption contribution Labsorption(λ) for the sample gas in the cavity and additional cavity losses Lcavity (from scattering, mirror reflectivity, etc.). The first-order rate constant τ(λ)−1 in Eq. (1) equals Ltotal divided by the round-trip time, which is (2d/c) for a linear two-mirror cavity with inter-mirror distance d and intra-cavity speed of light c:τ(λ)-1=Ltotal/(2d/c)=Ltotal×FSR=[Labsorption(λ)+Lcavity](c/2d),where FSR is the free spectral range, which equals the interval between the cavity’s longitudinal-mode resonance frequencies and is expressed as (c/2d) (i.e., the inverse of the round-trip time). Intra-cavity optical absorption contributes linearly to τ(λ)−1 and a CRD spectrum is typically derived by measuring τ(λ)−1 as a function of laser wavelength or frequency.

In conventional absorption spectroscopy, losses over an optical pathlength x can be expressed in terms of the absorption coefficient σ(λ) and the molar absorption (or extinction) coefficient ϵ(λ), so that the round-trip optical-absorption loss contribution Labsorption(λ) is then given by:Labsorption(λ)=2dσ(λ)=2dCϵ(λ),and Eq. (2) becomes:τ(λ)-1=Lcavity(c/2d)+cσ(λ)=Lcavity(c/2d)+cCϵ(λ)=τ0(λ)-1+cCϵ(λ),where C is the molar sample concentration (mol L−1). The ‘absorber-free’ ringdown rate constant τ0(λ)−1, equal to Lcavity (c/2d), can be determined by making measurements either of an absorber-free cavity (i.e., with C = 0) or (if scattering and other cavity losses are effectively wavelength-independent) at adjacent wavelengths at which there is zero absorption. It follows that σ(λ) is simply proportional to [τ(λ)−1τ0(λ)−1], the difference between the two ringdown rate constants (apart from small refractivity-based corrections for the dependence of the speed of light c on gas composition or wavelength). It is not necessary to know the cavity length d explicitly.

Accurate absolute concentrations can then be determined if ϵ(λ) is known for a spectroscopic feature at a particular wavelength λ, as follows:C=τ(λ)-1-τ0(λ)-1c-1ϵ(λ)-1

Such measurements are ‘absolute’ only in the sense that the spectroscopic database (e.g., HITRAN [49]) is sufficiently reliable and that adequate correction can be made for the effects (on c, (λ) and spectroscopic lineshape) of temperature, pressure, wavelength and gas composition.

Incidentally, as a precursor to CRD spectroscopy, measurements of τ0(λ) for an absorber-free ringdown cavity were recognized [50] and realized experimentally [51], [52], [53], [54], [55] as a way to determine mirror reflectivity R, for which the single-mirror contribution to Lcavity is (1  R); this early work on CRD reflectometry has been surveyed more comprehensively in Ref. [31].

A recently published book on CRD spectroscopy [42] demonstrates the technique’s scope and relevance, including its basic principles [56] and its operation with cw lasers [57], with broad-band light sources [58] and in optical waveguides [59]; it also reviews key CRD-spectroscopic applications to analytical chemistry [60], astrophysical molecules [61], atmospheric chemistry [62], medicine and the life sciences [63] and surface science [64].

In view of the extensive pre-existing and rapidly growing literature on CRD spectroscopy, we limit the scope of this Frontier Article by concentrating on a variant of cw-CRD spectroscopy, first developed in 1998, in which ringdown decay is measured by rapidly sweeping either the cavity length or the monochromatic tunable laser wavelength λ. Optical resonance between the laser and cavity frequencies during such sweeps enables automatic build-up and subsequent decay of the intensity of radiation in the cavity. We review progress in experimental methodology and in applications for this rapidly swept cw-CRD technique, indicating and examining its advantages and prospects as a convenient spectroscopic sensing method.

Section snippets

Rapidly swept cw-CRD spectroscopy: overview

In pulsed CRD spectroscopy [1], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [32], [42], the arrival of the incident optical pulse in the optical cavity and its subsequent termination naturally initiates the ringdown decay signal. By contrast, it is necessary in cw-CRD spectroscopy to build up the relatively low-intensity light in the optical cavity and then to interrupt the interaction between the cavity and the incident cw radiation; this is often achieved via an active optical

Principles and practical considerations

If the length of a Fabry–Perot cavity with high finesse F is swept very slowly, transmission of light by the cavity is effectively described by Airy’s formulae [21], as for a steady-state situation. However, as the cavity-length sweep rate increases, the cavity transmission profile becomes asymmetric and is modified by cavity-ringdown decay. Optical power builds up inside the cavity whenever a longitudinal-mode resonance frequency νm of the cavity (for a particular integer order m = νm/FSR) is

Principles and practical considerations

In rapidly swept cw-CRD spectroscopy, as explained in Sections 1 Introduction, 2 Rapidly swept cw-CRD spectroscopy: overview, the cw radiation and the optical cavity are swiftly moved in and out of resonance with each other to establish optical build-up and subsequent ringdown decay. One way to achieve this – rapidly sweeping the cavity length – has been discussed in Section 3. An alternative, complementary approach entails rapidly sweeping the wavelength or frequency of the monochromatic

Principles and practical considerations

We now consider quantitative trace-level near-IR spectroscopic sensing of gas mixtures (e.g., greenhouse-gas molecules such as CO2, H2O vapour, and CH4 at ambient concentrations in air) [78]. Our swept-cavity cw-CRD approach is readily amenable to such multi-species detection schemes and, moreover, enables multiple species to be simultaneously monitored, within the millisecond period of a ringdown-cavity sweep cycle.

A conventional (but time-consuming) direct approach to cw-CRD spectroscopy uses

Extensions of rapidly swept cw-CRD spectroscopy

In this section, we briefly survey additional experimental strategies for rapidly swept cw-CRD spectroscopy. Examples considered include: rapidly swept cw-CRD detection of stimulated Raman gain spectra (Section 6.1); rapidly swept liquid-phase and evanescent-wave cw-CRD measurements (Section 6.2); strategies for control of cw-CRD spectroscopic resonances (Section 6.3).

Concluding remarks: utility and potential of rapidly swept cw-CRD spectroscopy

As explained in Section 1, this Frontier Article has focused selectively on rapidly swept variants of cw-CRD spectroscopy, in which ringdown decay is measured by rapidly sweeping either the cavity length or the wavelength of the (monochromatic) tunable laser. Our rapidly swept cw-CRD approaches have previously been described [67], [68], [69], [70], [71], [72], [73], [74], [75], [76], [77], [78] and, excluding our own self-citations, our publications are cited in many review articles [22], [24],

Acknowledgments

We appreciate enthusiastic exchanges of factual information and critical opinion with many colleagues during the decade in which rapidly swept cw-CRD spectroscopy has evolved. We acknowledge financial support from the Australian Research Council and Macquarie University.

Brian Orr is Professor of Molecular and Optical Physics (and, before that, Professor of Chemistry) at Macquarie University in Sydney, Australia. Born and educated in Sydney, he took a PhD at the University of Bristol (UK) with Professor A.D. Buckingham, then worked at NRCC, Ottawa, Canada before pursuing his academic career back in Australia. Brian’s research interests cover a range of chemical, molecular and optical physics, including laser instrumentation, nonlinear optics, molecular

References (196)

  • A. O’Keefe et al.

    Chem. Phys. Lett.

    (1990)
  • G. Meijer et al.

    Chem. Phys. Lett.

    (1994)
  • J.J. Scherer et al.

    Chem. Phys. Lett.

    (1995)
  • J.J. Scherer et al.

    Chem. Phys. Lett.

    (1995)
  • D. Romanini et al.

    Chem. Phys. Lett.

    (1997)
  • D. Romanini et al.

    Chem. Phys. Lett.

    (1997)
  • D. Romanini et al.

    Chem. Phys. Lett.

    (1997)
  • Y. He et al.

    Chem. Phys. Lett.

    (1998)
  • K.J. Schulz et al.

    Chem. Phys. Lett.

    (1998)
  • H.-P. Loock

    Trends in Analyt. Chem.

    (2006)
  • P.B. Tarsa et al.
  • L.S. Rothman

    J. Quant. Spectrosc. Radiat. Transfer

    (2009)
  • Y. He et al.

    Chem. Phys. Lett.

    (2000)
  • Y. He et al.

    Chem. Phys. Lett.

    (2001)
  • Z. Li et al.

    Opt. Commun.

    (1991)
  • Z. Li et al.

    Opt. Commun.

    (1993)
  • J.Y. Lee et al.

    Appl. Phys. B

    (2004)
  • M.D. Levenson et al.

    Chem. Phys. Lett.

    (1998)
  • Z. Majcherova et al.

    J. Molec. Spectrosc.

    (2005)
  • B.V. Perevalov et al.

    J. Molec. Spectrosc.

    (2006)
  • B.V. Perevalov et al.

    J. Molec. Spectrosc.

    (2008)
  • A. O’Keefe et al.

    Rev. Sci. Instrum.

    (1988)
  • D. Romanini et al.

    J. Chem. Phys.

    (1993)
  • J.J. Scherer et al.

    J. Chem. Phys.

    (1995)
  • J.J. Scherer et al.

    J. Chem. Phys.

    (1995)
  • M.G.H. Boogaarts et al.

    J. Chem. Phys.

    (1995)
  • J.J. Scherer et al.

    J. Chem. Phys.

    (1995)
  • P. Zalicki et al.

    Appl. Phys. Lett.

    (1995)
  • B.A. Paldus et al.

    J. Appl. Phys.

    (1997)
  • J.J. Scherer et al.

    Chem. Rev.

    (1997)
  • J.B. Paul et al.

    Analyt. Chem.

    (1997)
  • M.D. Wheeler et al.

    J. Chem. Soc., Faraday Trans.

    (1998)
  • K.W. Busch, M.A. Busch (eds.), ‘Cavity-Ringdown Spectroscopy: an Ultratrace-Absorption Measurement Technique,’ Vol. 720...
  • G. Berden et al.

    Int. Rev. Phys. Chem.

    (2000)
  • C. Wang et al.
  • G. Berden et al.
  • A. McIlroy et al.
  • D.B. Atkinson

    Analyst

    (2003)
  • S.S. Brown

    Chem. Rev.

    (2003)
  • S.M. Ball et al.

    Chem. Rev.

    (2003)
  • C. Vallance

    New J. Chem.

    (2005)
  • A.A. Kachanov et al.

    Opt. Photonics News

    (2005)
  • M. Mazurenka et al.

    Ann. Rep. Prog. Chem., Sec. C: Phys. Chem.

    (2005)
  • B.A. Paldus et al.

    Can. J. Phys.

    (2005)
  • W.B. Yan et al.

    Int. J. Thermophysics

    (2008)
  • G. Friedrichs

    Zeits. Physik. Chemie

    (2008)
  • L. van der Sneppen et al.
  • L. van der Sneppen et al.

    Ann. Rev. Analyt. Chem.

    (2009)
  • C. Wang

    Sensors

    (2009)
  • K.K. Lehmann et al.
  • Cited by (0)

    Brian Orr is Professor of Molecular and Optical Physics (and, before that, Professor of Chemistry) at Macquarie University in Sydney, Australia. Born and educated in Sydney, he took a PhD at the University of Bristol (UK) with Professor A.D. Buckingham, then worked at NRCC, Ottawa, Canada before pursuing his academic career back in Australia. Brian’s research interests cover a range of chemical, molecular and optical physics, including laser instrumentation, nonlinear optics, molecular spectroscopy, chemical analysis, time-resolved optical double resonance, and molecular energetics. His recent work has focused on novel forms of cavity ringdown spectroscopy, narrowband tunable optical parametric oscillators, rotationally resolved collision-induced molecular energy transfer, and spectroscopic sensing applications. His academic distinctions include OSA’s 2004 W.F. Meggers Award “for outstanding work in spectroscopy”; he is currently a Deputy Editor of Optics Express.

    Yabai He was born and educated in China, before his PhD studies at Universität Bonn (Germany) and a postdoctoral period at ETH, Zürich (Switzerland), where he pioneered cw cavity-ringdown techniques. Since 1997, he has worked as a Senior Research Fellow at Macquarie University in Sydney, Australia. There, he and Brian Orr have published extensively on aspects of laser-based instrumentation and spectroscopic sensing, notably narrowband tunable optical parametric oscillators, cavity-ringdown spectroscopy, and fundamental atomic/molecular spectroscopy. In 2004, the Australian Optical Society awarded him its Technical Optics Award for “outstanding achievement in cavity ringdown and optical parametric oscillator system development.”

    View full text