7
Multipass cell based on confocal mirrors for sensitive broadband laser spectroscopy in the near infrared T. Mohamed, 1,2,3, * F. Zhu, 1 S. Chen, 1 J. Strohaber, 1 A. A. Kolomenskii, 1 A. A. Bengali, 2 and H. A. Schuessler 1,2 1 Texas A&M University, College Station, Texas 77843, USA 2 Science Department, Texas A&M University at Qatar, Doha 23874, Qatar 3 Physics Department, Faculty of Science, Beni-Suef University, Beni Suef 62511, Egypt *Corresponding author: [email protected] Received 10 May 2013; revised 18 July 2013; accepted 12 September 2013; posted 16 September 2013 (Doc. ID 190331); published 9 October 2013 We report on broadband absorption spectroscopy in the near IR using a multipass cell design based on highly reflecting mirrors in a confocal arrangement having the particular aim of achieving long optical paths. We demonstrate a path length of 314 m in a cell consisting of two sets of highly reflecting mirrors with identical focal length, spaced 0.5 m apart. The multipass cell covers this path length in a relatively small volume of 1.25 l with the light beam sampling the whole volume. In a first application, the absorp- tion spectra of the greenhouse gases CO 2 , CH 4 , and CO were measured. In these measurements we used a femtosecond fiber laser with a broadband spectral range spanning the near IR from 1.5 to 1.7 μm. The absorption spectra show a high signal-to-noise ratio, from which we derive a sensitivity limit of 6 ppmv for methane observed in a mixture with air. © 2013 Optical Society of America OCIS codes: (300.1030) Absorption; (300.6270) Spectroscopy, far infrared. http://dx.doi.org/10.1364/AO.52.007145 1. Introduction Absorption laser spectroscopy (ALS) is considered to be a powerful technique for qualitative and quantita- tive studies of atoms and molecules. The importance of ALS stems from its capability to detect gases in trace concentrations, which has many applications not only in physics and chemistry but also in environ- mental monitoring, biology, and medicine. A limita- tion of ALS originates from the requirement of long absorption path lengths especially when detecting low concentrations of gases on the order of parts per million by volume. To reach such a high sensitiv- ity, a variety of techniques exists, which are all based on achieving long optical path lengths. The most sensitive systems employ resonant cavities with high quality factors, such as cavity ringdown spec- troscopy (CRDS) [ 1, 2] and cavity enhanced absorp- tion and leak out spectroscopies (CEAS, CALOS) [ 3, 4], where achieving optical path lengths in the range of kilometer length are routine. More modest absorption lengths are provided by nonresonant de- vices such as integrated cavity output (ICO) [ 5, 6] and ringdown (ICRD) [ 7] spectroscopies as well as various multipass absorption spectroscopies (MPAS) [ 813]. The nonresonant methods are simple to work with and more robust against misalignment and have so far achieved path length of up to about 240 m. Mul- tipass absorption cells have been used in applications including environmental monitoring [ 1417], analy- ses of combustion processes [ 18, 19], performance of medical diagnostics [ 20], and the study of fundamen- tal atomic and molecular physics [ 21, 22]. 1559-128X/13/297145-07$15.00/0 © 2013 Optical Society of America 10 October 2013 / Vol. 52, No. 29 / APPLIED OPTICS 7145

Multipass cell based on confocal mirrors for …sibor.physics.tamu.edu/publications/papers/2013-02.pdfspherical mirrors. The number of reflections can be varied by changing the mirror

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Multipass cell based on confocal mirrors for …sibor.physics.tamu.edu/publications/papers/2013-02.pdfspherical mirrors. The number of reflections can be varied by changing the mirror

Multipass cell based on confocal mirrors forsensitive broadband laser spectroscopy

in the near infrared

T. Mohamed,1,2,3,* F. Zhu,1 S. Chen,1 J. Strohaber,1 A. A. Kolomenskii,1

A. A. Bengali,2 and H. A. Schuessler1,2

1Texas A&M University, College Station, Texas 77843, USA2Science Department, Texas A&M University at Qatar, Doha 23874, Qatar

3Physics Department, Faculty of Science, Beni-Suef University, Beni Suef 62511, Egypt

*Corresponding author: [email protected]

Received 10 May 2013; revised 18 July 2013; accepted 12 September 2013;posted 16 September 2013 (Doc. ID 190331); published 9 October 2013

We report on broadband absorption spectroscopy in the near IR using a multipass cell design based onhighly reflecting mirrors in a confocal arrangement having the particular aim of achieving long opticalpaths. We demonstrate a path length of 314 m in a cell consisting of two sets of highly reflecting mirrorswith identical focal length, spaced 0.5 m apart. The multipass cell covers this path length in a relativelysmall volume of 1.25 l with the light beam sampling the whole volume. In a first application, the absorp-tion spectra of the greenhouse gases CO2, CH4, and CO were measured. In these measurements we useda femtosecond fiber laser with a broadband spectral range spanning the near IR from 1.5 to 1.7 μm. Theabsorption spectra show a high signal-to-noise ratio, from which we derive a sensitivity limit of 6 ppmvfor methane observed in a mixture with air. © 2013 Optical Society of AmericaOCIS codes: (300.1030) Absorption; (300.6270) Spectroscopy, far infrared.http://dx.doi.org/10.1364/AO.52.007145

1. Introduction

Absorption laser spectroscopy (ALS) is considered tobe a powerful technique for qualitative and quantita-tive studies of atoms and molecules. The importanceof ALS stems from its capability to detect gases intrace concentrations, which has many applicationsnot only in physics and chemistry but also in environ-mental monitoring, biology, and medicine. A limita-tion of ALS originates from the requirement of longabsorption path lengths especially when detectinglow concentrations of gases on the order of partsper million by volume. To reach such a high sensitiv-ity, a variety of techniques exists, which are all basedon achieving long optical path lengths. The most

sensitive systems employ resonant cavities withhigh quality factors, such as cavity ringdown spec-troscopy (CRDS) [1,2] and cavity enhanced absorp-tion and leak out spectroscopies (CEAS, CALOS)[3,4], where achieving optical path lengths in therange of kilometer length are routine. More modestabsorption lengths are provided by nonresonant de-vices such as integrated cavity output (ICO) [5,6] andringdown (ICRD) [7] spectroscopies as well as variousmultipass absorption spectroscopies (MPAS) [8–13].The nonresonant methods are simple to work withand more robust against misalignment and haveso far achieved path length of up to about 240m.Mul-tipass absorption cells have been used in applicationsincluding environmental monitoring [14–17], analy-ses of combustion processes [18,19], performance ofmedical diagnostics [20], and the study of fundamen-tal atomic and molecular physics [21,22].

1559-128X/13/297145-07$15.00/0© 2013 Optical Society of America

10 October 2013 / Vol. 52, No. 29 / APPLIED OPTICS 7145

Page 2: Multipass cell based on confocal mirrors for …sibor.physics.tamu.edu/publications/papers/2013-02.pdfspherical mirrors. The number of reflections can be varied by changing the mirror

Common types of multipass cells are Whitecells [8,9], Herriott cells [10], optical matrix systemcells [11], simple cells based on a combination ofWhite and Herriott configurations [12], and astig-matic mirror cells [23]. The White cell [8,9] is the old-est arrangement and is composed of three concavespherical mirrors. The number of reflections can bevaried by changing the mirror angles. The Herriottcell [10] has two identical spherical mirrors withequal radii of curvature. The mirrors are separatedby a distance that is close to the radii of curvature.The laser beam is coupled into the cell through a holein one of the mirrors at an angle to the optical axis; itcompletes a certain number of passes between themirrors and exits through the same hole (or a holein the other mirror). In [10], the beam optical pathlength is fine-controlled by adjusting the separationdistance between the mirrors. Absorption pathlengths of up to 50 m were achieved for a Herriottcell with a base length of 0.5 m. In both the Whiteand the Herriott configurations, many of the spotsare overlapped and the surfaces of the mirrors arenot used efficiently. In order to better fill the mirrorsurfaces, a multipass optical cell formed by two astig-matic concave mirrors has been used [11]. The twomirrors have different radii of curvature, and the la-ser beam is injected and exits from the cell through acoupling hole in the center of the front mirror. Withsuch an astigmatic multipass optical cell, absorptionpath lengths of up to 240 m have been realized for acell with a base length of ≤1 m [13,24]. All reportedmultipass cells [8,10,11] can be enclosed in a vacuumchamber (for measurement in static gas samples orcontrolled gas flows), or be open to the atmosphere(for trace gas monitoring applications). A multipassoptical cell based on two cylindrical mirrors has beenreported in [25]. In addition, Silver [26,27] describedmathematically a cell consisting of two cylindricalmirrors with equal focal lengths. By rotating the rearmirror and adjusting the distance between the frontand rear mirrors, different spot patterns and pathlengths can be achieved. Recently, with a multipasscell based on cylindrical mirrors, path lengths of15 and 50 m were achieved in [28,29], respectively.

In this paper we report on a multipass cell designbased on confocal mirrors with a particular focus onachieving very long optical paths. We demonstrate apath length of about 300 m in a cell with the mirrorsspaced 0.5 m apart. Different spot patterns and pathlengths were achieved by tilting the mirrors with an-gles ≤0.05 rad. In a proof-of-principle experiment,CO2, CO, and CH4 direct absorption measurementswere performed using an M-fiber femtosecond fiberlaser (Menlo Systems GmbH) with a broadbandsupercontinuum spectrum in the near-IR spectralrange from 1.5 to 1.7 μm. Absorption spectra were re-corded using a high-resolution spectrum analyzer(Yokogawa AQ6375) and showed a high sensitivitylevel with good signal-to-noise ratio (S∕N), whichcould be extrapolated to a theoretical maximumsensitivity in the parts per million by volume range.

2. Optical Arrangement of the Multipass Cell

Multipass cells usually involve two or more mirrors.By fine adjustment of the positions of these mirrorsrelative to each other, the desired number of reflec-tions is achieved. One requirement to obtain a maxi-mum number of reflections is employing mirrors ofultrahigh reflectivity. In the reported multipass sys-tem, six mirrors were employed. Each side is formedby three spherical mirrors having 500 mm radii ofcurvatures: two of them are in a square shape of25 mm × 25 mm, while the third one has rectangularshape with dimensions of 25 mm × 50 mm. Togetherthe three mirrors form a square with dimensions of50 mm × 50 mm and act as reflectors on each side ofthe cavity. Eachmirror can be tilted independently inthe horizontal and vertical planes to change thenumber of reflections. Also the position of each mir-ror can be adjusted independently in the horizontaland vertical directions. The gaps between the mir-rors are minimized in such a way that the mirrorscan be tilted without touching each other. Beforeadjusting the number of desired reflections for a par-ticular task, the spot pattern on each mirror is calcu-lated. To achieve this, a computer simulation code inFORTRANwas developed to determine and optimizethe coordinates of the reflections. In the simulation,the two rectangular mirrors (50 mm × 25 mm) areconsidered to be composed of two square mirrors(25 mm × 25 mm). In the usually employed arrange-ment, a simulation is carried out, where on each sidethe reflector area of the multipass is divided into 64squares of 6.25 × 6.25 mm size each; i.e., the numberof squares in each square mirror is 16. The optimiz-ing criteria are as follows: the reflected beam cannotgo outside the mirror areas, it must not go back tothe edge of the incident aperture and must not beincident on the edges of the mirrors, the beam foldingnumber should be as large as possible, and themultireflected beams must completely cover the des-ignated volume. The simulation considers all subar-eas as the locations of the aperture on the mirrorwhere the entrance and the exit of the beam occur,and assumes the input laser beam is always parallelto the system axis. Each mirror has its own center ofcurvature. The image position of an object spot de-pends on the center of curvature of the mirror fromwhich it is reflected. In order to ensure that the im-age locations always stay on the center of a subarea,all mirror centers must be on the corners betweenthe subareas. There are nine corners that are closestto the system center. The simulation scans thesenine corners considering them as the possible mirrorcenter for each mirror. Our computer program uses amultilooping structure to find the optimum sequenceof filling the six mirrors. It works similarly to findingthe best move in a chess game and is not based onmatrix formulation, but only on the law of reflection,where both the image point and the object point aresymmetrical to the center of the reflecting mirror. Itrequires considerable computing effort and for thepresent 64 squares case takes about two days on a

7146 APPLIED OPTICS / Vol. 52, No. 29 / 10 October 2013

Page 3: Multipass cell based on confocal mirrors for …sibor.physics.tamu.edu/publications/papers/2013-02.pdfspherical mirrors. The number of reflections can be varied by changing the mirror

personal computer. As an example, the algorithm of asimulation program for seeking the highest numberof reflections is as follows: (a) the image coordinatesof the first reflection are calculated, (b) these coordi-nates are used as the next object point to determinethe coordinates of the following image point, (c) theprocess is repeated until the beam exits through thesame hole from which it entered the system, and(d) the configuration yielding the highest numberof reflections is saved, while the others are discarded.In cases in which the beam leaves the surface of themirrors, the simulation is terminated, and new loca-tions for the orientation of curvatures are assigned.After having calculated the spot pattern for the de-sired input parameters, the mirrors were adjusteduntil the spots matched their theoretical positionsand the spot pattern fell into its place. The simula-tion shows that dividing each mirror into smallersubareas is an effective way to increase the numberof reflections as shown in Fig. 1, where the number ofreflections increases linearly when increasing thenumber of subareas.

Using this result, the signal enhancement rate(SER) is found to increase when increasing thenumber of reflections. Here the SER is definedas �I1 – I2�∕Iβ, where I1 � Iin × RN and I2 � Iin ×�R − β�N are the output laser beam intensities with-out and with the absorbing medium, respectively,Iβ � Iin × β is the intensity absorbed after a singlepass, Iin is the input intensity, R is the mirror reflec-tivity, N is the number of reflections, and β is theabsorbance coefficient of the medium for a singlepass. In the calculations shown in Fig. 2, mirror re-flectivity ofR � 99.995% and β � 0.00002were used.

3. Experimental Results

The optical setup for carrying out absorption spec-troscopy is shown in Fig. 3. It consists of the multi-pass system, the M-fiber femtosecond laser, and aspectrum analyzer. The multipass cell is composed

of high-reflectivity mirrors from Layertec GmbH.The front confocal mirror M3 has a 2 mm diameterhole through which the light beam enters and exits.Since the light intensity at the output reducesexponentially with the number of reflections, high-reflectivity mirrors are necessary to achieve verylong path lengths with sufficient light exiting the cellfor spectroscopic measurements. All six mirrors thatform the cell have a broadband dielectric coatingwith a reflectivity higher than 99.995% in the wave-length range from 1.5 to 1.75 μm. To achieve preci-sion tilt control, micrometer adjustable rotationstages were attached to the sides of the mirrors.For fine adjustment of displacement, the precisionrotation stages were fixed to XYZ translation stages.The volume occupied by the beams depends on thespace between the mirrors. This volume has roughlya 50 mm × 50 mm cross section and is 500 mm long,yielding a total volume of 1.25 l.

For measurements with the multipass cell, weused a broadband femtosecond laser fromMenlo Sys-tems GmbH having an output power of 200 mW andwith Raman shifting covered the spectral range from1.5 to 1.75 μm. In addition, mode matching of the

Fig. 1. Number of reflections as a function of the number of sub-areas in each mirror: points show the simulation results, and thedashed line is a linear fit.

Fig. 2. SER as a function of the number of reflections.

Fig. 3. Schematic diagram of the multipass system. BS, beamsplitter.

10 October 2013 / Vol. 52, No. 29 / APPLIED OPTICS 7147

Page 4: Multipass cell based on confocal mirrors for …sibor.physics.tamu.edu/publications/papers/2013-02.pdfspherical mirrors. The number of reflections can be varied by changing the mirror

input beam to the cell configurations was carried outin such a way that the beam diameter remained thesame for a round trip in the cell.

The spectrum of the input and output beamswas measured by an optical spectrum analyzer(Yokogawa AQ6375). It covers the range from 1.2to 2.4 μm and has a high sensitivity that enablesmeasurements of powers as low as −70 dBm. It alsooffers a high wavelength resolution, since it usesa double-pass monochromator structure to achievea wavelength resolution of �0.05 nm and a wideclose-in dynamic range (55 dB). Thus, closely locatedsignals near the noise limit can be observed.

In practice, the beam spots on the mirrors wereadjusted to match the theoretical positions by the fol-lowing procedure: initially we cover each mirror’sreflective side with a transparent mask that hasthe subarea divisions indicated, then by tilting themirrors, the first 18 spots on the mirrors were posi-tioned, as required by the theoretical pattern withthe beam spots at the center of each subarea with ac-curacy of about 0.1 mm. The alignment was achievedwith the help of an infrared InGaAs camera fromHamamatsu (Model C10633) that has a spectral sen-sitivity from 0.9 to 1.7 μm and a quantum efficiencyof 80% at 1.5 μm. Once the alignment is achieved,after 628 reflections the beam exits the same holewhere the beam enters the multipass cell, and itspower is measured by a power meter. Then whilekeeping the desired spot pattern, the exit beampower was maximized by additional fine tuning ofthe positions and angles of each mirror. The trans-mission efficiency (power of the exit beam/power ofthe input beam) of the multipass cell reached about80%. This is close to the theoretical transmission ef-ficiency of about 90% obtained from the known reflec-tivity of the mirrors and the number of passes. Thedifference between the theoretical and measuredtransmission efficiency is mainly due to losses fromthe spots that were close to the gaps between the

mirrors. Figure 4 displays the 2D and 3D intensityprofiles of the observed spot patterns for mirrorsM1–4 [Fig. 4(a)] and mirrors M5–8 on the oppositeside [Fig. 4(b)]. In the 2D intensity profile, the spotsizes are not equal. This is mainly due to the fact thatoccasionally the beam falls within the same subareaon the same mirror several times. One of the perfor-mance characteristics of a multipass system is theaverage visiting numberN of the laser beam per spotarea on the mirrors, and in other experiments it hascaused S∕N limitations due to interference fringesof scattered light [23]. This effect is significantlyreduced in our experiment, which used a high-repetition-rate short-pulse femtosecond frequencycomb laser. In this case not only spatial but alsotemporal overlap are required to produce interfer-ence fringes. In addition the rear side of the high-reflectivity mirrors was antireflection coated. In thepresent experimentN � 4.9 and causes the intensityprofiles in Fig. 4.

The sensitivity of the multipass cell was tested bymeasuring the absorption spectra of CH4, CO, andCO2. These molecules have overtone absorptionbands that fall within the range of the used laser ra-diation (1.5–1.7 μm). In order to sample desired gasmixtures, the multipass system is placed in a plexi-glass enclosure at atmospheric pressure of ambientair. The enclosure has two ports; one is used foradmitting the gases through leak valves for eachgas, and the other port has an optical window for theinput and output laser beams. Figure 5 shows theoutput beam spectrum (laser spectrum, LS) recordedby the spectrum analyzer with no gas leaked to themultipass cell.

When the gases were leaked into the cell, theabsorption spectra of CH4, CO, CO2 gases wereobtained and are displayed in comparison with theLS in Fig. 6.

Fig. 4. 2D and 3D intensity profiles of the observed spot patternsfor mirrors (a) M1–4 and (b) M5–8.

Fig. 5. Spectrum of the output laser beamwith no gases leaked tothe multipass cell.

7148 APPLIED OPTICS / Vol. 52, No. 29 / 10 October 2013

Page 5: Multipass cell based on confocal mirrors for …sibor.physics.tamu.edu/publications/papers/2013-02.pdfspherical mirrors. The number of reflections can be varied by changing the mirror

The total multipass absorbance α is obtained from

α � − ln�II0

�; (1)

where I is the light intensity remaining after absorp-tion (the exit beam intensity) and I0 is the outputintensity in the absence of the absorber (when themirror losses are small it is close to the input beamintensity). For small losses, we have variationsδI0 ≈ δI, and when they are determined independ-ently one can write for the variation of the absorb-ance [30,31]

δα ≈ 2δII: (2)

The signal-to-noise ratio, S∕N � α∕δα, by takinginto account that α ∝ �n� 1� and that I0�≈ I� ∝ Rn,where R is the reflectance per pass and n is thenumber of reflections, can be expressed in theform [32]

S∕N � αI2δI

∝�n� 1�Rn

2δI: (3)

The minimum detectable number density ofabsorbing molecules ΔN is given [33] as

ΔN ≥α

σL�S∕N� : (4)

Equation (4) shows that to achieve the highest sen-sitivity to gas concentrations, the absorption pathlength L and the S∕N should be as large as possible,and the transitions with a large absorption crosssection σ are preferable.

The total absorbances α of CH4, CO, and CO2 gasescan be obtained from the observed data shown inFig. 6 and are plotted as a function of the wavelengthas shown in Figs. 7–9.

With the aim of determining the possible maxi-mum sensitivity of the developed optical multipasssystem, the absorption spectrum of CH4 was mea-sured and analyzed quantitatively. From the mea-sured absorption spectrum of the CH4 of Fig. 7, weselected the absorption line having the highest ab-sorbance at λ � 1.645 μm, where α� − ln�I∕I0� � 12.From the known absorption cross section [33] σ �17 × 10−21 cm2∕molecules and the optical path length

Fig. 6. Measured absorption spectra of CH4, CO, CO2 gasesdisplayed in comparison with the laser spectrum (LS). Fig. 7. Total absorbance α of CH4 gas as a function of the wave-

length. Only the central portion of the spectrum containinga strong absorption band is shown.

Fig. 8. Total absorbance α of CO gas as a function of thewavelength.

10 October 2013 / Vol. 52, No. 29 / APPLIED OPTICS 7149

Page 6: Multipass cell based on confocal mirrors for …sibor.physics.tamu.edu/publications/papers/2013-02.pdfspherical mirrors. The number of reflections can be varied by changing the mirror

L � 314 m, we can find the number density of CH4molecules

N�CH4∕cm3� �− ln

hII0

σ�cm2∕molecules�λ × L�cm� ; (5)

i.e., N � 2 × 1016 molecules∕cm3, and the corre-sponding concentration of 800 ppmv. The observedabsorption spectrum of CH4 in comparison withthe HITRAN database spectra in Fig. 10 exhibits ahigh S∕N of about 120. The S∕N is obtained by inte-grating the area under the spectral peak, which isthe signal S, and dividing S by the integral of the in-tensity over a corresponding amount of area where

no signal is detected, which is the noise N. For dis-cussion, reference is made to Figs. 10 and 11; Fig. 11depicts the HITRAN simulation for the six overlap-ping lines of interest, and Fig. 11 is a comparison ofthe observed and simulated signals. The measuredsignal is broadened by the resolution of the spectrumanalyzer. We measure the spectra without anyaveraging in 25 s per spectrum of 250 nm bandwidthwith 0.05 nm resolution and 0.01 nm samplinginterval. The minimal detectable number density ofabsorbing molecules ΔN is then 120 times less thanwhat is given by Eq. (5) and is about 1.6 × 1013

molecules∕cm3, i.e., 6 ppmv. Similarly, the minimaldetectable number densities of CO2 and CO werefound 640 and 320 ppmv. The ambient air concentra-tions of CH4, CO2, and CO are 1.7, 387, and 0.2 ppmv,respectively; i.e., it will be not possible to monitorthese molecules in the ambient air.

4. Conclusion

A multipass optical cell with a large optical pathlength was built and tested. It behaved as a nonreso-nant cavity and is based on six highly reflecting con-focal mirrors of identical focal length. The multipasscell of 50 cm length covers the long optical pathlength of about 314 m in a relatively small volumeof about 1 l where the particles of the gas mediumbetween the reflecting mirrors undergo interactionwith the laser radiation. Different spot patternsand path lengths have been achieved by tilting themirrors with angles ≤0.05 rad. For testing, the devel-oped multipass cell direct absorption measurementsof CO2, CO, and CH4 have been performed using abroadband frequency comb femtosecond fiber laserin the spectral range from 1.5 to 1.7 μm. The absorp-tion spectra for CO2, CO, and CH4 were recorded us-ing a spectrum analyzer and showed good S∕N. Inthe case of methane, S∕N � 120 was found, yielding

Fig. 9. Total absorbance α of CO2 gas as a function of thewavelength.

Fig. 10. HITRAN simulation of 6 CH4 absorption lines betweenλ � 1.645 μm and λ � 1.646 μm at room temperature 296 K, at 1atmospheric pressure, 800 ppm methane, and 314 m absorptionlength.

Fig. 11. Absorption line of CH4 having the highest absorbance atλ � 1.645 μm in comparison with the HITRAN database spectra(six overlapping lines; see Fig. 10).

7150 APPLIED OPTICS / Vol. 52, No. 29 / 10 October 2013

Page 7: Multipass cell based on confocal mirrors for …sibor.physics.tamu.edu/publications/papers/2013-02.pdfspherical mirrors. The number of reflections can be varied by changing the mirror

the estimated theoretical sensitivity of 6 ppmv, whichcan be further improved by optimizing the numberof passes. The optical apparatus is portable and canbe used for a wide range of applications, includingenvironmental monitoring, combustion processes,medical diagnostics, and fundamental atomic andmolecular physics studies.

This work is supported by the Qatar Foundationunder grant NPRP 09-585-1-087, the Robert A.Welch Foundation under Grant No. A1546, and theNSF under Grant No. 1058510.

References1. C. Wang, N. Srivastava, B. A. Jones, and R. B. Reese, “A novel

multiple species ringdown spectrometer for in situ measure-ments of methane, carbon dioxide, and carbon isotope,” Appl.Phys. B 92, 259–270 (2008).

2. K. W. Busch and M. A. Busch, Cavity-Ringdown Spectroscopy:An Ultratrace-Absorption Measurement Technique, Vol. 720 ofACS Symposium Series (Oxford University, 1999).

3. G. Berden, R. Peeters, and G. Meijer, “Cavity-enhanced ab-sorption spectroscopy of the 1.5 μm band system of jet-cooledammonia,” Chem. Phys. Lett. 307, 131–138 (1999).

4. H. Dahnke, D. Kleine, W. Urban, P. Hering, and M. Murtz,“Real-time monitoring of ethane in human breath usingmid-infrared cavity leak-out spectroscopy,” Appl. Phys. B 72,971–975 (2001).

5. J. B. Paul, L. Lapson, and J. G. Anderson, “Ultrasensitiveabsorption spectroscopy with a high-finesse optical cavityand off-axis alignment,” Appl. Opt. 40, 4904–4910 (2001).

6. E. S. Fry, G. W. Kattawar, and R. M. Pope, “Design of an in-tegrating cavity absorption meter,” Appl. Opt. 31, 2055–2065(1992).

7. A. O’Keefe, J. J. Scherer, and J. B. Paul, “Cw integrated cavityoutput spectroscopy,” Chem. Phys. Lett. 307, 343–349 (1999).

8. J. U. White, “Long optical paths of large aperture,” J. Opt. Soc.Am. 32, 285–288 (1942).

9. R. G. Pilston and J. U. White, “A long path gas absorption cell,”J. Opt. Soc. Am. 44, 572 (1954).

10. D. Herriot, H. Kogelnik, and R. Kompfner, “Off-axis paths inspherical mirror interferometers,” Appl. Opt. 3, 523–526 (1964).

11. S. M. Chernin and E. G. Barskaya, “Optical multipass matrixsystems,” Appl. Opt. 30, 51–58 (1991).

12. C. Robert, “Simple, stable, and compactmultiple-reflection op-tical cell for very long optical paths,” Appl. Opt. 46, 5408–5418(2007).

13. J. B. McManus, P. L. Kebbabian, and M. S. Zahniser, “Astig-matic mirror multipass absorption cells for long-path-lengthspectroscopy,” Appl. Opt. 34, 3336–3348 (1995).

14. P. Werle, R. Mücke, and F. Slemr, “The limits of signal aver-aging in atmospheric trace gas monitoring by tunable diode-laser absorption spectroscopy,” Appl. Phys. B 57, 131–139(1993).

15. D. G. Lancaster, A. Fried, B. Wert, B. Henry, and F. K. Tittel,“Difference-frequency-based tunable absorption spectrometerfor detection of atmospheric formaldehyde,” Appl. Opt 39,4436–4443 (2000).

16. A. Silver and W. R. Wood, “Miniature gas sensor for monitor-ing biological space environments,” Proc. SPIE 4817, 82–87(2002).

17. C. Guimbaud, V. Catoire, S. Gogo, C. Robert, M. Chartier,F. Laggoun-Défarge, A. Grossel, P. Albéric, L. Pomathiod,

B. Nicoullaud, and G. Richard, “A portable infrared laserspectrometer for flux measurements of trace gases at thegeosphere–atmosphere interface,” Meas. Sci. Technol. 22,075601 (2011).

18. A. Arnold, H. Becker, R. Hemberger, W. Hentschel, W.Ketterle, M. Kollner, W. Meienburg, P. Monkhouse, H. Neckel,M. Schafer, K. P. Schindler, V. Sick, R. Suntz, and J. Wolfrum,“Laser in situ monitoring of combustion processes,” Appl. Opt.29, 4860–4872 (1990).

19. P. Minutolo, C. Corsi, F. D’Amato, and M. De Rosa, “Self- andforeign-broadening and shift coefficients for C2H2 lines at1.54 μm,” Eur. Phys. J. D 17, 175–179 (2001).

20. M. Gianella and M. W. Sigrist, “Infrared spectroscopy onsmoke produced by cauterization of animal tissue,” Sensors10, 2694–2708 (2010).

21. J. Zhang, Z. H. Lu, and L. J. Wang, “Precision measurement ofthe refractive index of carbon dioxide with a frequency comb,”Opt. Lett. 32, 3212–3214 (2007).

22. D. Qu and C. Gmachl, “Quasichaotic optical multipass cell,”Phys. Rev. A 78, 033824 (2008).

23. J. B. McManus, M. S. Zahniser, and D. D. Nelson, “Dual quan-tum cascade laser trace gas instrument with astigmaticHerriott cell at high pass number,” Appl. Opt. 50, A74–A85(2011).

24. R. F. Curl and F. K. Tittel, “Tunable infrared laser spectros-copy,” Annu. Rep. 98, 219–272 (2002).

25. L.-Y. Hao, S. Qiang, G.-R. Wu, L. Qi, D. Feng, Q.-S. Zhu, andZ. Hong, “Cylindrical mirror multipass Lissajous systemfor laser photoacoustic spectroscopy,” Rev. Sci. Instrum. 73,2079–2085 (2002).

26. J. A. Silver, “Simple dense-pattern optical multipass cells,”Appl. Opt. 44, 6545–6556 (2005).

27. J. A. Silver, “Dense pattern optical multipass cell,” Europeanpatent 1,621,867A1 (21 July 2005).

28. V. L. Kasyutich and P. A. Martin, “Multipass optical cellbased upon two cylindrical mirrors for tunable diode laserabsorption spectroscopy,” Appl. Phys. B 88, 125–130(2007).

29. D. Das and A. C. Wilson, “Very long optical path-lengthfrom a compact multi-pass cell,” Appl. Phys. B 103, 749–754(2011).

30. J. F. Doussin, R. Dominique, and C. Patrick, “Multiple-passcell for very-long-path infrared spectrometry,” Appl. Opt.38, 4145–4150 (1999).

31. D. W. Steyert, J. M. Sirota, M. E. Mickelson, and D. C. Reuter,“Two new long-pass cells for infrared and visible spectros-copy,” Rev. Sci. Instrum. 72, 4337–4343 (2001).

32. X. Cui, C. Lengignon, W. Tao, W. Zhao, G. Wysocki, E.Fertein, C. Coeur, A. Cassez, L. Croize, W. Chen, Y.Wang, W. Zhang, X. Gao, W. Liu, Y. Zhang, and F. Dong,“Photonic sensing of the atmosphere by absorption spectros-copy,” J. Quant. Spectrosc. Radiat. Transfer 113, 1300–1316(2012).

33. L. S. Rothman, I. E. Gordon, A. Barbe, D. C. Benner, P. F.Bernath, M. Birk, V. Boudon, L. R. Brown, A. Campargue,J.-P. Champion, K. Chance, L. H. Coudert, V. Dana, V. M. Devi,S. Fally, J.-M. Flaud, R. R. Gamache, A. Goldman, D.Jacquemart, I. Kleiner, N. Lacome, W. J. Lafferty, J.-Y.Mandin, S. T. Massie, S. N. Mikhailenko, C. E. Miller, N.Moazzen-Ahmadi, O. V. Naumenko, A. V. Nikitin, J. Orphal,V. I. Perevalov, A. Perrin, A. Predoi-Cross, C. P. Rinsland,M. Rotger, M. Šimečková, M. A. H. Smith, K. Sung, S. A.Tashkun, J. Tennyson, R. A. Toth, A. C. Vandaele, and J.Vander Auwera, “The HITRAN 2008 molecular spectroscopicdatabase,” J. Quant. Spectrosc. Radiat. Transfer 110,533–572 (2009).

10 October 2013 / Vol. 52, No. 29 / APPLIED OPTICS 7151