Mid-Infrared Laser Applications in Spectroscopy

Abstract

The vast majority of gaseous chemical substances exhibit fundamental vibrational absorption bands in the mid-infrared spectral
region (≈ 2–25 µm), and the absorption of light by these fundamental bands provides a nearly universal means for their detection.
A main feature of optical techniques is the non-intrusive in situ detection capability for trace gases. The focus time period of this chapter is the years 1996–2002 and we will discuss primarily
CW mid-infrared laser spectroscopy. We shall not attempt to review the large number of diverse mid-infrared spectroscopic
laser applications published to date. The scope of this chapter is rather to discuss recent developments of mid-infrared laser
sources, with emphasis on established and new spectroscopic techniques and their applications for sensitive, selective, and
quantitative trace gas detection. For example, laboratory based spectroscopic studies and chemical kinetics, which will also
benefit from new laser source and technique developments, will not be considered.

Full text

Mid-Infrared Laser Applications
in Spectroscopy
Frank K. Tittel1, Dirk Richter2,andAlanFried
2
1Rice Quantum Institute, Rice University
Houston, TX 77251–1892, USA
fkt@rice.edu
2The National Center for Atmospheric Research
1850 Table Mesa Dr., Boulder, CO 80305, USA
{dr,fried}@ucar.edu
Abstract. The vast majority of gaseous chemical substances exhibit fundamen-
tal vibrational absorption bands in the mid-infrared spectral region (≈2–25 µm),
and the absorption of light by these fundamental bands provides a nearly universal
means for their detection. A main feature of optical techniques is the non-intrusive
in situ detection capability for trace gases. The focus time period of this chapter
is the years 1996–2002 and we will discuss primarily CW mid-infrared laser spec-
troscopy. We shall not attempt to review the large number of diverse mid-infrared
spectroscopic laser applications published to date. The scope of this chapter is
rather to discuss recent developments of mid-infrared laser sources, with emphasis
on established and new spectroscopic techniques and their applications for sensi-
tive, selective, and quantitative trace gas detection. For example, laboratory based
spectroscopic studies and chemical kinetics, which will also benefit from new laser
source and technique developments, will not be considered.
1 Solid-State Mid-IR Spectroscopic Laser Sources
Mid-infrared laser sources come in many different varieties. For their appli-
cation to spectroscopic measurement, and specifically for quantitative mea-
surements, the ideal source would have the following properties: (1) sufficient
optical power to overcome inherent electronic detection noise and ensure high
laser signal-to-noise ratios, (2) narrow linewidth to obtain high selectivity and
sensitivity, (3) single longitudinal mode operation with low amplified sponta-
neous emission output for high selectivity and elimination of intermode com-
petition noise; (4) ease of tailoring the inherent laser operating wavelength
(design of gain material and/or cavity structure) to access the desired ab-
sorption region; (5) low source noise and low amplitude modulation; (6) high
beam quality, i.e., small beam divergence, small astigmatism, and stable,
predictable beam output direction, for optimum coupling into and through
a gas sampling cell; (7) low temperature and current tuning rates to minimize
wavelength jitter induced by controller noise; (8) rapid wavelength tunability
for fast response and high data acquisition rates; (9) minimal susceptibil-
ity to changing environmental conditions of temperature, pressure, humidity,
I. T. Sorokina, K. L. Vodopyanov (Eds.): Solid-State Mid-Infrared Laser Sources,
Topics Appl. Phys. 89, 445–516 (2003)
c
Springer-Verlag Berlin Heidelberg 2003

446 Frank K. Tittel et al.
and vibrations; (10) no long term changes in laser wavelength and/or spatial
output characteristics; (11) highly reliable performance for many years; and
(12) compact and robust overall sensor package size. Of course it is challeng-
ing to realize all these idealized attributes in any one real world mid-infrared
laser source, and they are listed here to serve as a general guideline. However,
some of the attributes are more important than others for a given applica-
tion to obtain the best possible measurement performance. In the sections
that follow, we will elucidate wherever possible those important attributes
necessary for achieving optimal performance employing the technique being
discussed.
It is common practice in infrared spectroscopy to express transition fre-
quencies in inverse centimeters (cm−1), or wavenumbers, defined simply as
the inverse of the transition wavelength in vacuum, ν=λ−1. Multiplying this
quantity by cgives the frequency in hertz; thus 1 cm−1is roughly 30 GHz.
We shall use both units throughout this chapter, where appropriate.
The spectral coverage of the most frequently employed tunable CW mid-
infrared sources is shown in Fig. 1. In general, one can distinguish between
two classes of mid-IR laser sources. Class “A” includes sources which generate
tunable mid-IR laser radiation directly from gain in gas discharge, semicon-
ductors, rare-earth and transition-metal doped solid-state bulk materials or
optical fibers. Class “B” laser sources are based on nonlinear optical para-
metric frequency conversion of near-infrared (≈0.9to2µm) laser sources.
0.1 1 10
Ant
i
mon
i
de III-V
897
6
5
432
Semicon ductor
Direct Solid-State
Gas
(Line Tunabl e)
Frequency
Conversion
Based Sources
Nd
3+
CO Laser
Rotat
i
on
-
V
i
brat
i
ona
l
OPSL
Solid State and
Fiber Laser /
Amplifier
DFB Diode Laser
Yb
3+
Pr
3+
Er
3+
Tm
3+
/Ho
3+
Overt one Reg
i
on
III
Wavelength ( m)
QC-Laser
QPM GaAs
PPLN / PPKTP / PPRTA
Lead-Sa
l
t
CO
2
Laser
Cr
2+
II-VI
DFG / OPO / OPA
Class‘A’
Class‘B’
"
Pm)
Fig. 1. Laser sources and typical wavelength coverage. Shown are also the wave-
lengths of the two atmospheric windows I (2.9 –5.3 microns) and II (7.6–16 microns),
typically accessed for trace gas sensing. OPSL, optically pumped semiconductor
laser

Mid-Infrared Laser Applications in Spectroscopy 447
Many hundreds of non-linear optical crystals have been developed for this
purpose, however only very few of these are practical and in use. In this
chapter, examples of parametric frequency conversion sources will be lim-
ited to quasi-phase-matched materials (QPM). While class “A” mid-IR laser
sources are inherently more compact, class “B” laser sources can be made
to operate over a much wider wavelength spectrum, often in a single opti-
cal arrangement, which is a key advantage for certain applications. Figure 1
shows the spectral coverage of Class “A” and “B” laser sources in the mid-
IR wavelength region. The merits of the various spectroscopically important
laser sources shown will be discussed in this section. Although line-tunable
laser sources such as CO and CO2gas lasers have and are being successfully
used in trace gas detection [1,2], they are not within the scope of this book,
which focuses specifically on solid-state laser sources.
1.1 Class “A” Laser Sources
Class “A” lasers as defined above, are direct laser sources. In the following
subsections these lasers are discussed with respect to their typical operating
characteristics and optical set-up conditions. The first group of class “A”
lasers is further subdivided into semiconductor based lasers, namely lead-salt,
antimonide and quantum cascade laser sources, which share some common
features such as cryogenic cooling (in some cases thermoelectric) and the
set-up of collection optics. This section is followed by a discussion of tunable
solid-state laser sources.
1.1.1 Lead-Salt Diode Lasers
Lead-salt diode lasers have been developed for operation at wavelengths from
3to30µm and have been available since the mid-1960s. These lasers are com-
prised of PbTe, PbSe, and PbS and various alloys of these compounds with
the same materials above and with SnSe, SnTe, CdS and other materials.
The lead-salt diode laser consists of a single crystal of these semiconductor
materials to form a p–n junction. The crystal is shaped into an optical cavity
employing parallel end faces that are approximately 250 µm square at both
the front and rear faces. Typical cavity lengths range from ≈300 to 500 µm.
Lasing action is achieved by applying a forward bias current which injects
charge carriers (electrons or holes) across the p–n junction, and this in turn
populates the nearly empty conduction band. Stimulated emission across the
band gap between the conduction and nearly full valance band provides the
gain mechanism for lasing action. Since the emission photon energy approxi-
mately equals the band gap energy, which in the case of Pb-salt materials is
small, these lasers require cryogenic cooling to achieve population inversion.
These lasers therefore are subject to temperature extremes (≈10–300 K) and
this places stringent demands on the entire laser package.

448 Frank K. Tittel et al.
The energy band gap, which is dependent upon semiconductor composi-
tion and crystal temperature, determines the output lasing wavelength. One
can tailor the lasing wavelength regime at the time of manufacture by ei-
ther varying the stoichiometry between Pb and the other constituents or by
employing different alloys of Pb. Any given device can be actively tuned in
wavelength over ≈100 cm−1by changing the device temperature, or over
tens of cm−1by changing the injection current. Both tuning mechanisms,
however, produce semicontinuous wavelength coverage, since the laser struc-
ture is a Fabry–P´erot device. Varying the injection current generally allows
continuous tunability over a ≈1–2 cm−1spectral region before the output
jumps to a new longitudinal mode. In some cases the gain is broad enough
to support multiple longitudinal modes simultaneously, resulting in wave-
length regions where the lasing output gradually shifts from one mode to
another. Tuning by temperature takes advantage of the change in band gap,
and hence the wavelength with temperature. Typical rates range between 2
and 5 cm−1/K. Since this mechanism involves changing the temperature of
the entire laser package, including the stage on which the laser is mounted,
temperature tuning is very slow on the order of seconds. Since typical absorp-
tion linewidths are in the 0.001 cm−1range, at reduced pressures of several
tens of Torr typically employed in sample cells, stable operation requires tem-
perature control to better than 1mK over long time periods. Tuning by cur-
rent, which involves ohmic heating of the active region only, causes a change
in the refractive index of this region, which in turn produces a change in
wavelength. This tuning mechanism is very rapid, thus making it possible
to employ high frequency modulation techniques in the kHz to MHz regime.
Typical current tuning rates range between 0.02 and 0.07 cm−1/mA (≈606–
2121 MHz/mA). This rapid tuning mechanism, however, also requires low
noise laser current controllers in order to avoid linewidth broadening. Given
the typical absorption linewidths above, low noise controller operation of
10 µA or better is required. Employing such a low noise temperature and cur-
rent controller, Reid et al. [3] determined that the measured linewidths from
Pb-salt diode lasers varies dramatically from laser to laser, and for any given
laser, depends strongly upon the junction temperature and injection current.
Typical linewidths (FWHM), in the absence of refrigerator shocks caused
by the closed cycle cooling system employed, were 0.6–25 MHz. Vibrations
from the closed cycle cooler degraded this linewidth to ≈60 MHz. However,
linewidths of ≈100 MHz or more were sometimes observed by Reid et al. and
by Lundqvist et al. [4]. Sams and Fri ed [5] discussed the effect of such me-
chanical vibrations on quantitative spectroscopic determinations. Since most
diode lasers now operate at or above liquid nitrogen temperatures, where
liquid nitrogen dewars have replaced closed cycle refrigerator systems, one
should expect linewidths ≤25 MHz.
Older Pb-salt devices were grown by a diffusion process, which unfortu-
nately often resulted in a poorly defined lasing region. In many instances

Mid-Infrared Laser Applications in Spectroscopy 449
lasing occurred over nearly the entire junction width in multiple filaments,
which in some cases produced a laser output in different directions with dif-
ferent wavelengths [6]. Even when such multiple beams were not present, the
main emission lobe was sometimes emitted at an angle relative to the optic
axis of the laser. The resulting poor spatial beam quality, which made it ex-
tremely difficult to optimally collect the output beam, together with other
problems just discussed, has no doubt contributed to the poor performance
reported by many early users of Pb-salt devices. This has led to the develop-
ment of newer Pb-salt lasers based on mesa structure and ultimately buried
heterostructure or double heterostructure devices in which the active lasing
area is highly confined to a region less than 1 µm thick and less than 10 µm
wide. These devices, which are discussed by Preier et al. [7], are prepared by
molecular beam epitaxy using PbEuSeTe or PbSnTe active layers. In addition
to improved output beam quality, which as we will show is very important
for trace gas detection employing multi-pass absorption cells, these newer
devices exhibit threshold currents as low as 1 mA and higher operating tem-
peratures. Most of these devices typically operate at temperatures between
77 and 120 K, which is accessible using liquid nitrogen cryogenic dewars in-
stead of bulky and noisy closed cycle refrigerator systems required by earlier
devices. Despite this significant effort, Pb-salt diode lasers have not been
manufactured to routinely operate CW at temperatures much higher than
the above range and certainly less than the highest reported temperature
of 215 K for CW operation published by Wall [8]. Typical CW single mode
output powers for Pb-salt lasers are in the range 100–500 µW. For compre-
hensive reviews of Pb-salt diode lasers, their electrical and optical properties,
performance characteristics and device materials along with the correspond-
ing manufacturing techniques, the reader is referred to [7,8,9], and the many
references found in Grisar [10]. Brassington also presents an excellent review
of these topics as well as applications of Pb-salt lasers [11].
Unfortunately due to the small market for Pb-salt lasers, significant ad-
vances in device structure based on buried quantum well, distributed feedback
and distributed Bragg reflectors, which appeared so promising in the 1990s,
are not being aggressively pursued at present. Since lead-salt diode lasers can
be modulated at very high frequencies (tens of kHz up to several hundred
MHz), similar to those of near-infrared diode lasers, harmonic detection and
two-tone modulation techniques can be employed as an efficient means of re-
ducing noise (Sect. 1.2.2). As discussed previously, all CW Pb-salt diode lasers
require some form of cryogenic cooling for operation. Although cryogenic op-
eration is feasible and indeed commonplace in field environments, even in
rugged airborne laser systems [12], such operation may impose limitations
that can ultimately affect system performance. For example, all cryogeni-
cally cooled lasers are temperature cycled many times over their lifetime and
may result in unrecoverable changes of the lasing frequency [13]. Other ef-
fects include long-term changes of tuning characteristics, slight changes of the

450 Frank K. Tittel et al.
spatial mode quality, and reduction in laser power. Although all present Pb-
salt devices are temperature-cycled by the manufacturer to minimize these
problems, the finite possibility of such changes is disconcerting in certain
demanding applications where it is imperative to access routinely a specific
absorption feature. Airborne measurements of formaldehyde, where there is
a limited choice of strong and interference-free absorption lines, is one such
example [14,15]. In addition, since liquid nitrogen dewars frequently contain
as many as four lasers, one runs the risk of mechanically disturbing all lasers
in changing out any given laser. Small manipulations to the laser lead wires,
which are unavoidable in this procedure, may result in irreversible changes
to laser performance, similar to that from temperature cycling.
Lead-salt diode lasers also exhibit large beam divergence and astigmatism,
which places critical and stringent alignment requirements on the collection
optics, particularly the first optical element, which is placed in front of the
dewar. Subtle changes in the alignment between the laser and the first collec-
tion optic, due to small mechanical changes in the position of either the laser
or the optical element, necessitates periodic adjustments to the first collection
element, and this may add some mechanical instability to the alignment. Such
instability, even when relatively small, in turn often leads to optical noise in
IR absorption systems. This noise source, which is produced by unwanted
scattering off various optical elements, results in a periodically undulating
background structure of varying amplitude, spacing, and temporal frequency.
Despite these drawbacks of lead-salt diode lasers, spectrometers employ-
ing these sources still yield excellent sensitivity, even in rugged field envi-
ronments. As we will show in a later section, a liquid-nitrogen cooled Pb-
salt diode laser spectrometer can routinely measure ambient formaldehyde
(CH2O) levels as low as 20–50 parts-per-trillion by volume (pptv) on an
airborne platform employing 1-minute integration [16]. This corresponds to
a minimum detectable absorbance, Amin ,of0.7to1.7×10−6(S/N=1)using
a 100-meter absorption pathlength.
Tabl e 1 . Summary of lead-salt diode laser characteristics
Wave l en g th
range ( µm)
Tuning
(coarse/fine)
Power
(mW)
Linewidth Beam profile
characteristics
Operating
requirements
3–30 100 cm−1/
1–2 cm−1
0.1–0.5 1–1000 MHz – Elliptical
– Highly astigmatic
– Highly divergent
Cryogenic
cooling
1.1.2 Antimonide Diode Lasers
Continuous wave lasing at room temperature at wavelengths above 2 µm
with output optical powers up to 20 mW/facet has been achieved using
structures grown by molecular beam epitaxy (MBE) on GaSb substrates

Mid-Infrared Laser Applications in Spectroscopy 451
and employing compressively strained GaInSbAs quantum wells (QWs) be-
tween Ga(Al)Sb(As) barriers in the active region. Narrow ridge Fabry–P´erot
GaInSbAs/GaSb type II electrically pumped QW lasers emitting at 2.35 µm
have been reported [17,18]. For further details on mid-infrared heterojunc-
tion lasers see the chapter by Joulli´e et al. These lasers emit in a fundamental
spatial mode and exhibit single frequency operation over a range of currents
and temperatures. They emit in a spectral region where overtone and com-
bination absorption lines of such gases as CO, CH4,NH
3and NO2can be
accessed conveniently. In efforts to extend coverage to the fundamental spec-
tral region, several groups have reported the development of antimonide diode
lasers in the 2–3 µm spectral range [19] and InAsSb/InAs lasers between 3
and 5 µm[20,21,22,23,24]. However, this technology is not yet robust and
single frequency operation by means of DFB (Distributed Feedback) laser
designs has not been realized. Hence these devices must be operated in an
external cavity configuration to achieve stable and narrow-linewidth opera-
tion. Although room-temperature operation has been demonstrated, reliable
single-frequency operation still requires cryogenic cooling.
Tabl e 2 . Summary of antimonide diode laser characteristics
Wave l en g th
range ( µm)
Tuning
(coarse/fine)
Power
(mW)
Linewidth Beam profile
characteristics
Operating
requirements
2–3; 3–5 1–2 cm−10.1
(single-
mode)
50 MHz – Elliptical
–Astigmatic
– Highly divergent
– Thermoelectric
(2–3 µm)
– Cryogenic cooling
(>3µm)
1.1.3 Quantum Cascade Lasers
Quantum cascade (QC) lasers are unipolar semiconductor injection lasers
based on intersubband transitions in a multiple quantum-well heterostruc-
ture. They are designed by means of band-structure engineering and grown
by molecular beam epitaxy. The emission wavelength of a QC laser depends
on the thickness of the quantum well and barrier layers of the active region
rather than the band gap of diode lasers. These lasers operate either as CW or
pulsed devices. The chapter by Hofstetter and Faist as well as several papers
in [25,26,27,28] provide details of their design and operating characteristics.
In the following, an overview will be given on emission wavelengths, output
powers and design approaches.
Quantum cascade lasers grown in a InGaAs/AlInAs lattice matched to the
InP material system have been fabricated for emission wavelengths from 3.5
to 24 µm. Quantum cascade lasers have excellent spectroscopic properties in
terms of optical power, but their tuning range is limited and their beam diver-
gence is large and astigmatic. Multi-mode devices with 100 stages (quantum

452 Frank K. Tittel et al.
well gain regions in series) have demonstrated peak powers of 0.6 W at room
temperature. Until recently room temperature operation was only feasible for
pulsed operation, but CW multi-spectral mode operation up to temperatures
of 312 K was reported for a Fabry–P´erot style QC laser in 2002 [29]. Reli-
able single frequency operation has been achieved through the integration of
a Bragg grating into the laser waveguide, resulting in a distributed feedback
(DFB) laser, operating at cryogenic temperatures. The latest generation of
QC-DFB lasers is based on a “top-grating” approach that takes advantage of
the characteristics of a mid-infrared waveguide. For mid-infrared wavelengths
below 15 µm, dielectric waveguides built from low-doped semiconductor lay-
ers that have appropriate refractive index modulation are used [30]. At longer
wavelengths, the waveguide is overlaid with metal. In this case the radiation
is guided not only by the dielectric but also by a surface plasmon mode [31].
Continuous wavelength tunability without mode hops is achieved through
the temperature dependence of the waveguide parameters. The temperature
can either be varied by a temperature change of the heat sink on which the
device is mounted or more rapidly by dissipative heating through changing
the direct QC laser excitation current. Characteristic total tuning ranges
per current sweep are typically around 0.4 % of the emission wavelength.
For many spectroscopic purposes, the spectral linewidth of the laser emis-
sion is as important as continuous tunability. The linewidth of selected CW
DFB QC lasers ranges from a few MHz [32] through current stabilization to
a few kHz with frequency stabilization [28], but exceeds 150 MHz (HWHM)
in pulsed operation. Device reliability and long-term wavelength tuning char-
acteristics are excellent as a result of using robust materials, such as InP and
GaAs based heterostructures. To achieve mode-hop free tuning for Fabry–
P´erot-type QC lasers in the mid-IR, a grating-coupled external cavity has
been used to obtain a wavelength tuning of hundreds of nanometers, or up
to 8 % of the central wavelength in the 3–4 µm region for InAs/InAsSb or
GaSb/InAsSb heterostructure lasers with a few hundred milliwatts (mW)
peak power [33]. Fabry–P´erot QC lasers [34]at4.5and5.1µm have been
tuned with an external cavity. The principal technical issue is the need to
deposit a low loss broadband antireflection coating or an angled surface on
one of the laser output facets.
A practical consideration of QC lasers is their operating current require-
ments, often drawing multiple amperes of current in CW operation. In ad-
dition, a QC laser typically requires compliance voltages of 5–10V. The re-
sulting thermal load to the laser is significant and good thermal management
is important to achieve room-temperature operation. Low noise drivers have
been developed based on the modified Libbrecht design [35]. The use of bat-
teries also permits low noise operation of QC lasers and linewidths below
1 MHz are obtained without frequency locking.
To date, QC laser-based chemical sensors primarily use InGaAs/InAlAs
type-I QC-DFB devices. There are two limitations inherent to this kind of

Mid-Infrared Laser Applications in Spectroscopy 453
Tabl e 3 . Summary of quantum cascade laser characteristics
Wave l en g th
range ( µm)
Tuning
(coarse/fine)
Power
(mW)
Linewidth Beam profile
characteristics
Operating
requirements
4.3–24
(Selected
regions
where single-
frequency
operation
has been
demonstrated)
35 cm−1
with
external
grating
cavity/
3cm
−1
1–100
(CW,
single
frequency)
50 (avg.,
pulsed)
0.001–
10 MHz,
CW
(locked–
unlocked)
>150 MHz
(HWHM),
pulsed
– Elliptical
–Astigmatic
–Highly
divergent
– Cryogenic cooling
(CW)
– Peltier cooling
(pulsed)
– High voltage,
high current low
noise driving
electronics
laser for chemical sensing. First, they cannot access the spectral region of
C–H and O–H stretch vibrations near 3000cm−1. This shortcoming can be
overcome by developing QC lasers based on alternative materials and struc-
tures. For example, the 3000 cm−1region is accessible by type-II lasers [36],
but no single frequency devices of this kind have yet been demonstrated. An-
other issue is the limited tunability of each QC-DFB laser, which restricts
the feasibility of multi-component chemical sensing. This requirement can
be addressed by separating the gain medium from the wavelength-selective
element [34,37]. In [34], a QC laser tunability of ≈35 cm−1at a fixed temper-
ature was demonstrated in an external cavity configuration with a diffraction
grating. This is about a ten times wider range than that typically achieved
for QC-DFB lasers by means of current tuning. On-line concentration mea-
surements of ≈20 gaseous compounds and several isotopomers in ambient air
have been realized in the first three years of various QC laser based chemical
sensors.
Mid-infrared semiconductor laser sources as discussed above share several
common characteristics. Low temperature operation is yet the most reliable
means of obtaining CW tunable single frequency emission. For the purpose
of applying the generated radiation to a specific spectroscopic technique and
application, the laser radiation has to be collected and mode matched to the
spectroscopic sampling path or cell. Semiconductor lasers exhibit a large di-
vergence and astigmatism and four representative approaches for beam collec-
tion, shaping and delivery to the spectroscopic sampling path are illustrated
in Fig. 2.
Illustrated in Fig. 2a is the approach used by Fried et al. [16]. Here, all
reflective elements are used to collect, collimate and image the laser radi-
ation to a long-path absorption cell. Fig. 2b shows the approach used by
Nelson [38], in which an all-reflective objective is used for the same purpose.
However, collection efficiency is typically lower than the technique depicted
in Fig. 2a due to center mirror obscuration by as much as 20 %. Care must
be taken to avoid residual feedback of the center mirror, which may increase
the laser noise and at times induce mode-hopping. In approach 2c (Laser
Components, Inc.), a toroidal mirror is used as a first collection element to
correct the astigmatism/emission aspect ratio and lower the divergence of

454 Frank K. Tittel et al.
Fig. 2. Representative approaches for optical beam collection, shaping and imaging
of highly divergent, astigmatic laser radiation. See text for explanation of figures.
OAP, off-axis parabolic mirror; OAE, off-axis ellipsoidal mirror; AP, aperture; RO,
reflecting objective; M, mirror; OAT, off-axis toroidal mirror; L, lens
the beam. An OAE re-images the beam to the sampling path or cell. This
design provides a well-defined beam shape and quality, however at the cost
of complexity and the use of four optical elements, all subject to drift, which
can lead to beam pointing instabilities. The advantage of this configuration
is that the beam can be optimally collected while maintaining a fixed output
axis. The approach shown in Fig. 2d uses two lenses to collimate and re-
image the laser [39], avoiding the spherical aberrations introduced by OAP
and OAEs. Similar to the approach shown in Fig. 2b, feedback may be an
issue to the laser wavelength stability and is subject to creating stronger
optical interference fringes between refractive optical elements.
1.1.4 Tunable Solid-State Lasers
A large and important class of tunable lasers is based on the vibroni-
cally broadened transitions that can occur in certain gain media, such as
color centers and certain transition metal or rare-earth ions in crystalline
hosts [40,41,42,43]. When such a medium is placed in a tunable cavity and
pumped above laser threshold, stimulated emission can be made to occur at
any desired frequency within the emission band. Tunable laser media based on
3d–3d transitions of transition-metal ions and 4f–5d transitions of rare-earth
ions cover the mid-infrared spectral range between 1.0µmand4.7µm. The
tuning range of such lasers can be widely varied by the choice of impurities
and by selecting different hosts. Recent spectroscopic studies demonstrated
that chromium-doped zinc selenide chalcogenides, such as Cr2+:ZnSe and
Cr2+:ZnS, have favorable characteristics as tunable mid-infrared solid-state
materials near 2.5µm[44,45]. These include room-temperature operation,

Mid-Infrared Laser Applications in Spectroscopy 455
broad tunability, the possibility of direct diode-pumping, erbium fiber ampli-
fier and CW operation.
Tabl e 4 . Summary of tunable solid-state laser characteristics
Wave l en g th
range ( µm)
Tuning
(coarse/fine)
Power
(mW)
Linewidth Beam profile
characteristics
Operating
requirements
2–5 600 nm/
1cm
−1
200–1000 →
20 →
800 MHz, 1 s
0.1 MHz, 1 s
–TEM
00
– Cavity subject
to pointing
instability
Low technical
noise
environment
For selective spectroscopic detection at reduced pressures, the output
power of the above mentioned solid-state laser sources is significantly reduced
if one or more frequency selective optical elements are employed to obtain
laser linewidths of less than 10 MHz. For further details see the chapter by
Sorokina on mid-IR crystalline lasers and references therein.
1.2 Class “B” Laser Sources
Class “B” laser sources, as defined earlier, are based on parametric frequency
conversion of near-IR laser source(s). These can be configured either in reso-
nant (cavity) arrangements with a single pump laser or non-resonant (single-
pass) arrangements with two pump lasers and are referred to as optical para-
metric oscillator (OPO) [46] and difference frequency generation (DFG) based
sources, respectively. Figure 3illustrates these two concepts.
Fig. 3. Shown are two representative examples of DFG and OPO. DFG based mix-
ing uses single pass parametric interaction and requires spatial overlap of the input
sources via discrete or optical fiber components. For wavelength tuning, the OPO
ring-cavity is wavelength tuned with a piezo driven cavity mirror M. A frequency
selector, e.g. ´etalon, is synchronously tuned to maintain single-mode operation [47].
M, mirror; ET, ´etalon; L, lens; PPLN, periodically poled lithium niobate; SM,
semi-transparent mirror; BS, beamsplitter/combiner

456 Frank K. Tittel et al.
Before discussing the two approaches of parametric frequency conversion,
we will discuss a variety of commonly used room-temperature near-IR laser
sources that can serve as effective pump sources, followed by a tabulated
overview of these sources and their performance characteristics. These laser
sources are not emitting mid-IR radiation but can operate at wavelengths
ranging from the visible to near-infrared region. However, used in combination
with frequency conversion, their spectroscopic characteristics (linewidth, tun-
ing range, etc.) are partially (OPO) or completely transferred to the mid-IR.
1.2.1 Near-IR Pump Laser Sources
for Parametric Frequency Conversion
Many of the problems in tuning and wavelength stability of Fabry–P´erot
type diode lasers can be enormously reduced by incorporating an external
or internal grating structure to provide more defined feedback [48]. In an
ECDL (External Cavity Diode Laser) , one or both faces of the laser chip
are antireflection-coated to eliminate optical feedback. Instead, the feedback
required for laser action is provided by an external cavity. The cavity acts
as a narrow wavelength selector, which determines a specific operating wave-
length out of the usually broad gain spectrum of the semiconductor laser
material. Several cavity configurations have been developed that differ in the
method of tuning, number of components, output beam characteristics, and
output coupling efficiency. Mode-hop-free single frequency tuning ranges of
over 1000GHz have been demonstrated for an ECDL [49,50,51]. Grating-
tuned external cavity diode lasers with large tuning ranges have been com-
mercialized in the near-infrared from 0.9 to 1.6µm. All ECDL designs have
been based on tuning a frequency selector by mechanical means. In envi-
ronments with technical noise such as vibrations this can lead to additional
frequency jitter. More recent designs have incorporated Micro Electro Me-
chanical Systems (MEMS) technologies, which dramatically reduce the size
(5 mm ×5mm) of the external cavity [52].
Implementation of integrated wavelength mode filters in the form of dis-
tributed feedback (DFB) structures or distributed Bragg reflectors (DBR)
also enable single frequency operation and have been developed for the optical
telecommunication industry at wavelengths ranging from 1.3 to 2 µm. Other
novel developments include broadly tunable monolithic integrated multi-
section diode laser chips employing gain, filter and Bragg tuning elements [53].
Unlike external grating controlled diode lasers, these lasers [54,55,56]offerfast
and versatile electronic coarse wavelength tunability (≈70 nm at 1.56 µm)
of potential pump sources for nonlinear optical frequency conversion devices.
Other forms of single frequency near-IR diode laser include VCSEL (Vertical
Cavity Surface Emitting Laser), which are simpler and more cost effective
to produce, but have not yet reached comparable output powers and IR
wavelength coverage. VCSELs can tune over ≈30 cm−1by changing their
operating current (threshold-maximum). Such tuning rates are attractive for

Mid-Infrared Laser Applications in Spectroscopy 457
certain applications such as rapid combustion diagnostic studies, but not
desirable for precision gas sensing.
Grating stabilized near-IR laser sources have also been developed in the
form of rare-earth doped single mode DFB fibers in the 1µm(Yb)and
1.5µm (Er) wavelength region. The operating wavelength within the dopant
gain region can be precisely engineered to ≈0.1 nm accuracy by writing
a DFB grating via photolithography into a photosensitive doped fiber region.
Fiber lasers of this type can be wavelength tuned by means of temperature
and straining the DFB fiber length with a piezoelectric element [57]. Ultra-
narrow linewidths on the order of a few kHz have been achieved without
the use of frequency locking techniques and hence offer convenient high-
resolution wavelength tuning capability. In addition DFB fiber lasers offer
low relative intensity noise and high side-mode suppression ratios of >60 dB
[http://www.koheras.com][58].
Another important solid-state laser class demonstrated for operation
in the near-IR wavelength region is the monolithic non-planar ring oscil-
lator based on Nd, Yb or Tm:YAG materials. This type of laser offers
single-frequency CW or pulsed output with diffraction-limited beam qual-
ity. Narrow-linewidth CW powers exceeding several watts have been demon-
strated [59]. This laser source is specifically suited and used for pumping opti-
cal parametric oscillators. For further details see the chapters by Vodopyanov
and Ebrahimzadeh.
1.2.2 Sources Based on Difference Frequency Generation (DFG)
Numerous DFG based mid-IR sources have been designed and used for spec-
troscopy [60,61,62,63,64,65,66,67,68,69,70,71,72]. For further details see also
the chapter by Fischer and Sigrist. In the case of DFG, two laser beams
(“pump” and “signal”) at different frequencies combined in a nonlinear mate-
rial with suitable dispersion characteristics generate a beam at the difference-
frequency (“idler”). The narrow emission spectra of the “pump” (highest fre-
quency) and “signal” (middle frequency) are convolved during the frequency
conversion and hence translate into a similarly narrow spectrum of the idler
wave. Idler wavelength tuning is accomplished by tuning of the pump laser, or
signal laser, or both. In order that the idler wave continue to build up as the
beams pass collinearly through the nonlinear material, the three waves must
stay in phase (the “phase matching condition”). This imposes a condition on
the refractive indices of the three waves.
This condition can often be satisfied with a birefringent nonlinear crystal
by having some of the three waves polarized along an ordinary axis and
some polarized along a direction that includes the extraordinary axis. If the
polarization direction that includes the extraordinary axis is not parallel to
it (angle tuning), the three waves will not propagate in the same direction
(double refraction) and the beams will separate as they pass through the
crystal (“walk-off”) limiting the overlap region and the DFG power. In order

458 Frank K. Tittel et al.
Tabl e 5 . Summary of tunable near-IR diode, fiber laser and solid-state source
characteristics
Wave l en g th
range ( µm)
Tuning
(coarse/fine)
Power
(mW)
Linewidth Beam profile
characteristics
Operating
requirements
ECDL 0.650–1.655 10–100 nm/
1cm
−1
(by Piezo)
1–50 1–5 MHz,
1s
– Elliptical
–Astigmatic
–Beampointing
instability
with tuning
(Littrow
design)
Low technical
noise environ-
ment for best
performance.
Vibrations
lead to
increased
linewidth of
>100 MHz
SG-
DBR-
DL
1.55 200 cm−1/
1cm
−1
2–10 25 MHz,
1s
–Singlemode
fiber
diffraction
limited
Gaussian
beam quality
– Inherent beam
pointing
stability
– Telecom
environment
–Requires
complex
electronics
for tuning
DFB-
DL
1.3–2 3 nm/
1cm
−1
2–30 0.1–
1 MHz,
1s
–Singlemode
fiber
diffraction
limited
Gaussian
beam quality
– Inherent beam
pointing
stability
– Telecom
environment
DFB
Fiber
Laser
1.03–1.2
(Yb)
1.528–1.61
(Er)
5cm
−1
(Temp.)/
10 cm−1
(PZT)
2–10 0.1 MHz,
1s
–Singlemode
fiber
diffraction
limited
Gaussian
beam quality
– Inherent beam
pointing
stability
– Telecom
environment
Solid-
State
0.946, 1.064,
1.319, 1.444,
(Nd:YAG)
2.02
(Tm:YAG)
30 GHz
(Temp.)/
100 MHz
(piezo)
150–
2000
(CW)
0.02 MHz,
1s
–TEM
00
(M2<1.1)
–Room
temperature
ECDL, External Cavity Diode Laser ; SG-DBR-DL, Sampled Grating Distributed Bragg
Reflector Diode Laser; DFB, Distributed Feedback Diode Laser
to satisfy phase matching keeping all waves exactly parallel or perpendicular
to the optic axis (“90◦phase matching”), the refractive indices must be tuned
as the difference frequency is tuned by varying the temperature of the crystal
or by tuning pump and signal simultaneously.
In the early demonstration of the DFG method by Pine [73], single mode
argon-ion and dye laser outputs were combined in bulk lithium niobate crystal
to produce narrow-band (15 MHz) radiation tunable from 2.2 to 4.2µmby
temperature tuning the crystal. Simultaneous tuning of both signal and pump

Mid-Infrared Laser Applications in Spectroscopy 459
has been used in AgGaS2[64] to provide tunable single frequency radiation
from 3.5 to 9 µm. For DFG radiation longer than 5 µm it is also possible
to use birefringent bulk nonlinear optical materials, such as AgGaSe2[74],
ZnGeP2[75], or GaSe [76,77,78].
Another approach to phase matching is the introduction of periodic short
(about 10–30 µm wide) regions in which the sign of the second order suscep-
tibility alternates, thus bringing the three waves back into the right phase
relationship. This is called quasi-phase-matching (QPM). It is most easily
achieved in ferro-electric materials, where the direction of the extraordinary
axis can be permanently reversed locally by the application of an external
electric field (≈20 kV) at elevated temperatures [79]. Other advantages of
QPM are no walk-off effects, access to the crystal’s largest diagonal nonlin-
ear coefficient (not accessible by birefringent phase-matched crystals), wide
acceptance bandwidth, and relative ease of alignment.
The implementation of diode-pumped mid-infrared frequency conversion
sources received a significant boost from the development of novel periodi-
cally poled nonlinear materials, such as lithium niobate (PPLN) [80], lithium
tantalite (LiTaO3), and ferroelectric crystals of the potassium titanyl phos-
phate (KTiOPO4, or KTP) family at wavelengths in the 2.5–5.2µm spectral
region [81,82,83]. The quasi-phase-matching properties of each of these crys-
tals can be engineered for interaction of any pump and signal wavelengths
within the transparency range of the crystal, allowing significant flexibility
in the choice of laser sources for frequency mixing [84,85,86]. In the future,
quasi-phase matched GaAs [87] should become available greatly extending the
long wavelength region covered by DFG. In the work reported by Eyres et al.
orientation-patterned GaAs films of 200 µm thickness have been grown by
hydride vapor phase epitaxy (HVPE) on an orientation-patterned template
fabricated by molecular beam epitaxy (MBE).
The availability of PPLN permits near-infrared diode or fiber lasers to be
used as pump lasers [68,69,70,71,88,89,90,91,92] instead of much larger dye or
Ti:sapphire lasers making it feasible to construct compact mid-infrared spec-
trometers that operate at room temperature and can generate CW output
powers up to 1 mW [93]. Thus the practicality of near-infrared diode laser and
optical fiber technology are combined to achieve the analytical power of mid-
infrared spectroscopy in a single instrument. Such an instrument inherits the
single-frequency operation and high modulation speed capabilities of diode
lasers, and takes advantage of their relatively wide tuning range. For exam-
ple, a typical 780-nm diode laser can be grating tuned over 20 nm, or 2.6 % in
wavelength without appreciable changes in output power. When the output
of such a laser is down-converted by mixing with a 980 nm diode laser, the
tuning range in frequency units remains the same, in this case a significant
tuning range: 3.6–4.1µm, or 13 % in wavelength.
A detailed quantitative theory of this nonlinear optical process is beyond
the scope of this review chapter. Instead, the reader is referred to a paper by

460 Frank K. Tittel et al.
Zondy [94] and the references contained therein. Therefore, we will simply
state that the maximum idler power generated in a given crystal is propor-
tional to the product of crystal length, pump power, signal power, and the
square of the second-order nonlinear coefficient of the crystal. Optimum DFG
output power is achieved by means of precise spatial overlap and focusing of
the pump and signal beams. There is an optimal focus because a point is
reached at which any further increase in beam intensity through tighter fo-
cusing is offset by a decrease in interaction length due to diffraction, resulting
in loss of output power and in some cases clipping at the crystal aperture.
Although this type of source is routinely used for spectroscopy and gas de-
tection, DFG in bulk nonlinear (either birefringent or quasi-phase matched)
crystals is characterized by low conversion efficiency, typically [95]inthe
range 0.002–0.05 % W−1cm−1.
The tradeoff between beam size and interaction length can be eliminated
in guided-wave DFG. Optical confinement of pump and signal radiation near
the waveguide core creates a region of high intensity and good modal overlap
that can be maintained throughout the length of the waveguide. Thus the
interaction length for tight focusing is now limited by the length of the wave-
guide, not by diffraction. Guided-wave parametric processes, such as OPO,
SHG, and DFG, have been demonstrated [96] in periodically poled LiNbO3,
LiTaO3,andKTP.InLiNbO
3, for example, a waveguide can be formed by
titanium in-diffusion, or by a Li+–H
+ion exchange typically followed by sev-
eral hours of annealing at elevated temperature to create a graded index
distribution.
A DFG waveguide designed to carry a single spatial mode at the idler
wavelength is necessarily multimode at the shorter, pump and signal wave-
lengths. The presence of multiple spatial modes complicates waveguide phase
matching characteristics. For example, a TEM00 (fundamental) mode at the
signal wavelength will interact with TEM02 and TEM10 modes at the pump
wavelength, but not with TEM01 or TEM11 modes. Efficient and repro-
ducible fundamental-mode excitation of a DFG waveguide was first achieved
by Chou et al. [96,97] using a combination of a mode filter and an adiabatic
taper. An improved device featuring separate inputs for the pump and sig-
nal beams followed by a directional coupler has also been demonstrated.
DFG waveguides have been used to build sources of mid-infrared radiation
for spectroscopic purposes [98,99,100]. Surprisingly, though, reported idler
power from waveguide-DFG have not exceeded powers of more than 0.1mW,
whereas with bulk QPM DFG crystals, powers exceeding 1 mW have been
obtained through use of high input power sources. Difficulties in efficiently
coupling the pump laser power into the waveguide and maintaining the wave-
guide properties in the presence of higher power fields have so far prevented
the generation of higher mid-IR power levels that exceed those obtained with
bulk QPM DFG crystals. In addition, waveguide-QPM structures have rel-
atively narrow conversion bandwidths, whereas bulk-QPM can be poled to

Mid-Infrared Laser Applications in Spectroscopy 461
have either multiple grating periods or a fan-out structure for continuous
phasematching over 750 cm−1[69].
Tabl e 6 . Summary of difference frequency generation based sources
Wave l en g th
range ( µm)
Tuning
(coarse/fine)
Power
(mW)
Linewidth Beam profile
characteristics
Operating
requirements
2.3–4.6
(based on
available
near-IR
telecom
laser
wavelengths)
15 cm−1per
diode laser
pair;
wavelength
multiplexing
possible/ 2 cm−1
0.1–1 mW 1 MHz – Near Gau ssian
(fiber pump
beam delivery
only)
–Beampointing
stability
– large f/#≈100
Telecom
environment
1.2.3 Tunable Optical Parametric Oscillators
Optical parametric oscillators (OPOs) are progressing as useful spectroscopic
tools for the generation of coherent radiation that is continuously tunable
over large spectral ranges [101,102,103,104,105,106,107]. For further details
see the chapters by Vodopyanov and Ebrahimzadeh. Unlike DFG, the non-
linear crystal is placed in a cavity and is used to generate output beams at
two new frequencies (signal and idler) ν1and ν2from a single pump beam
at ν3(Fig. 3). Energy conservation requires ν1+ν2=ν3. How the frequency is
divided between the new waves, signal and idler is determined by the phase-
matching condition. The development of pulsed OPOs is mature, and these
devices are available commercially. On the other hand, CW OPOs still have
some practical problems in terms of requiring high power pump sources, ef-
ficient and high quality nonlinear crystals, low loss broadband optics, and
mode-hop-free operation with good frequency stability in order to realize
their potential usefulness in mid-IR spectroscopic power.
OPO devices are arranged so that the signal frequency resonates inside
the cavity (singly resonant OPO), or in addition the idler can also be made
resonant (doubly resonant OPO). In some special cases the pump frequency
can be made resonant as well (triply resonant OPO). In principle, both dou-
bly resonant and singly resonant OPO configurations can be used although
as a practical matter doubly resonant OPOs are difficult to construct and
cumbersome to tune.
As in DFG devices quasi-phase-matching in periodically poled ferroelec-
tric crystals offers several distinct advantages for their use in CW OPOs, such
as non-critical phase matching and a high effective nonlinear coefficient, deff .
A particularly significant development was the demonstration of the use of
PPLN as the parametric gain medium, in which case the oscillation threshold
of externally pumped CW singly resonant OPOs can be reduced to the few
watt level hence making it feasible to use diode pumped solid-state pump

462 Frank K. Tittel et al.
lasers [101,102,103,104,105,106,108,109,110,111,112,113,114]. Pump powers
as low as 800mW were used to pump a PPLN singly resonant OPO in which
both the pump and signal were resonated [115].
The parametric process can also be used in optical parametric amplifiers
(OPAs) to boost infrared output powers. Continuous wave OPOs amplified
by pulsed OPAs offer a competitive alternative to other tunable lasers in
the 1–5 µm spectral region in terms of linewidths, wavelength tunability and
output powers.
Tabl e 7 . Summary of OPO (PPLN based) characteristics [116]
Wave l en g th
range ( µm)
Tuning
(coarse/fine)
Power
(mW)
Linewidth Beam profile
characteristics
Operating
requirements
1.45–2
2.3–4
1900 cm−1/
0.05 cm−1
10–
100 mW
0.15 MHz,
instantaneous
TEM00 Low technical
noise
environment
2 Fundamentals of Absorption Spectroscopy
for Trace Gas Detection
Spectroscopic trace gas detection is a method allowing one to determine the
concentration of a known gas, or gases, from a measured optical absorption
spectrum of the gas mixture (in practice, a small fragment of the spectrum
is measured). The procedure requires a good quantitative knowledge of the
gas absorption characteristics. This knowledge is the realm of molecular spec-
troscopy, a complex and highly developed subject. A few fundamental spec-
troscopic concepts and formulae that are directly applicable to gas detection
are, however, summarized in this section.
Each atom or molecule, small or large, is uniquely characterized by a set of
energy levels. Transitions between levels by absorption or emission of electro-
magnetic radiation result in highly specific spectroscopic features. These fea-
tures allow both the identification and quantification of the molecular species,
such as atmospheric trace gases. Molecules may undergo transitions between
electronic, vibrational, and rotational states when exposed to electromagnetic
radiation, resulting in absorption spectra. These spectra consist of a number
of discrete absorption lines. Each line will have a certain linewidth and shape
that depends on temperature and what surrounds the molecule. The lines
may in some cases be resolved and in other cases the line density may be
too high to be spectrally resolved. Transitions between molecular rotational-
vibrational (“ro-vibrational”) states occur in the infrared “fingerprint” region
of the electromagnetic spectrum, approximately between the wavelengths of
2.5 and 25 µm. Also, overtone and combination-overtone ro-vibrational lines
are possible with significantly lower intensities as compared to those for fun-
damental vibrational bands and the corresponding wavelengths are in the

Mid-Infrared Laser Applications in Spectroscopy 463
CO
N
2
O
CO
2
N
2
O
HCl
CH
2
O
CH
4
HCN
L=0.5 m; p=40 T orr ; Resolution= 2.4 cm
-1
All species plotted at 250
ppm
relative concentration
CO
N
2
O
CO
2
N
2
O
HCl
CH
2
O
CH
4
HCN
L=0.5 m; p=40 T orr ; Resolution= 2.4 cm
-1
All species plotted at 250
ppm
relative concentration
Fig. 4. HITRAN simulation of absorption bands of various molecules in the 3–5 µm
spectral region. All species are plotted with identical relative concentration. Spectral
overlap limits the choices of interference free absorption lines
0.8–2.5µm spectral region. Transitions between electronic states of atoms
and molecules occur in the ultraviolet and visible spectral region.
All polyatomic molecules, with the exception of homonuclear diatomic
molecules (e.g. N2), absorb infrared radiation. The absorption changes the
state of molecular rotation and vibration. An absorption spectrum therefore
depends on the physical properties of the molecule such as size and shape and
hence each molecule is characterized by a unique spectral “signature”. Spec-
tra of linear and some nonlinear polyatomic molecules consist of an array of
individual or small groups of lines. In the case of large polyatomic molecules
(e.g. benzene, C6H6) at atmospheric pressure, there are many lines overlap-
ping each other, resulting in broad spectral features with some occasional
peaks.
There are numerous atmospheric trace gases and their concentrations
are normally in the parts-per-trillion (pptv, 10−12 ), parts-per-billion (ppbv,
10−9) to parts-per-million (ppmv, 10−6) range. However, species such as wa-
ter vapor may have concentrations of up to a few percent (%, 10−2). Because
of this, even weak water features, where absorption cross-sections are as much
as a factor of 10−10 weaker than the molecule of interest, can be a problem.
There is much spectroscopic data available in the literature and in elec-
tronic form which are important tools in the identification and develop-
ment of specific detection strategies, especially in the presence of interfering
species [117,118].
Direct gas phase laser absorption spectroscopy based on the Beer–
Lambert absorption law is often used for quantitative measurements.
In the absence of optical saturation and particulate-related scattering, the
intensity of light I(x) propagating in a homogeneous gas of sample length L

464 Frank K. Tittel et al.
Fig. 5. Illustration of the Beer–Lambert absorption law
follows the Beer–Lambert law:
I(x)=I0exp [−σ(ν)NL].(1)
Here Nrepresents the molecular concentration and σ(ν) the absorption cross-
section. The molecular absorption cross-section depends on frequency and has
units of cm2cm−1per molecule when integrated over the absorption line, and
units of cm2per molecule at the line center. For simplicity we assume only
one absorbing species. The peak absorption cross-section at line center (ν0)
is related to the integrated line strength through a lineshape function Γ(ν).
This function Γ(ν) has the same analytical form for all transitions, and in
mid-infrared spectroscopy, the broadening of an individual transition due to
finite upper-level lifetime is insignificant compared to broadening by the other
two important mechanisms – thermal motion and molecular collisions. Their
individual and combined effects on a molecular transition at a frequency νn
are expressed as follows:
Thermal motion (Gaussian):
ΓD(ν)= 1
γDln 2
πe−ν−ν0
γD2ln 2 ;γD=3.58 ×10−7ν0T
M,(2)
Molecular collisions (Lorentzian):
ΓL(ν)= 1
π
γL
(ν0−ν)2+γ2
L
;γL=γ0
LpT
T0
,(3)
Combined broadening (Voigt):
Γ(ν)=k(ν0)D
y
π
∞

−∞
e−t2
y2+(x−t)2dt;k(ν0)D=σ(ν)dν
γDln 2
π,(4)
with,
x=ν−ν0
γD√ln 2 ; y=γL
γD
√ln 2 ; t=δ
γD√ln 2 .
Here Tis the gas temperature (K), Mthe molecular weight, Pthe gas
pressure (atm), and γ0
Lthe coefficient of pressure broadening (cm−1atm−1),
k(ν0)Dis the peak Doppler cross-section, δis the parameter of integration

Mid-Infrared Laser Applications in Spectroscopy 465
and is used to express the Doppler and Lorentz frequency differences during
the convolution process in terms of a single variable. The quantities γDand
γLare referred to as the Doppler- and pressure-broadened half width at half
maximum (HWHM) linewidths. The line shape that results from the com-
bined effect of Doppler- and pressure-broadening is a convolution of the two
respective line shapes, and it is known as the Voigt profile. The physical sig-
nificance of the convolution is that the Voigt profile has different asymptotic
shapes for very low and very high gas pressure. At low pressure, molecular
collisions are less frequent, leaving thermal motion the dominant broaden-
ing mechanism – the corresponding line shape is near-Gaussian. As the gas
pressure increases the collisions take over, and the resulting line shape is
near-Lorentzian (Fig. 6).
The previous expressions do not include the effect of pressure shift,
which is typically in the range of several megahertz per atmosphere. The
shift is very small compared to the width of an atmospheric-pressure-
broadened line, typically several gigahertz. It can be verified by integrat-
ing the absorption cross-section of an individual transition over frequency,
that the shift is independent of the broadening mechanism and is equal
to the line intensity S, in units of cm2cm−1per molecule. The line in-
tensity is proportional to the lower-state population density of a transi-
tion and thus depends on temperature. These parameters have been mea-
sured and calculated for many lightweight gas molecules in the mid-infrared
2831.5 2831.6 2831.7 2831.8 2831.9 2832.0
99.70
99.75
99.80
99.85
99.90
99.95
100.00
C
Frequency (cm
-1
)
Transmission (%)
B
A
H
2
O
CH
2
O
CH
2
O
A: p=1 Torr, 50 ppm CH
2
O, 40 % H
2
O
B: p=40 Torr, 5 ppm CH
2
O, 4 % H
2
O
C: p=200 Torr, 5 ppm CH
2
O, 4 % H
2
O
Fig. 6. Computed mid-IR absorption spectra of CH2OandH
2Oat1,40and
200 Torr. Lineshapes correspond to near-Gaussian (A), Voigt (B) and near-
Lorentzian (C). Also note the higher relative concentration at lower pressures to
obtain a comparable absorption strength. Optimum sampling pressure with good
signal strength and selectivity ranges typically between 30 and 60 Torr

466 Frank K. Tittel et al.
spectrum, and compiled into extensive databases such as HITRAN [117],
GEISA [118], NIST (http://ois.nist.gov/srmcatalog/datafiles), and
PNNL (http://nwir.pnl.gov). Numerically accurate absorption spectra
can be computed based on these data, not only for single gas species but
for gas mixtures as well.
The analytical formulae apply also to multi-component gas mixtures. The
total absorption cross-section σ(ν) is then a weighted average of absorption
cross-sections of individual species, with the mole fraction Cmof each species
used as the weight coefficient:
σ(ν)=
m
Cmσm(ν),
m
C:m=1.(5)
For each o f t h e mspecies, the pressure broadening coefficients γ0
Lgenerally
depend on the transition. They also depend on the type of molecule with
which the collisions occur. In general, partial pressures in conjunction with
the appropriate pressure-broadening coefficients should be used to compute
the overall pressure broadening from all gases present in the background
(this includes self-broadening). Air-broadening coefficients are useful in cal-
culations, and are listed in spectroscopic databases [117,118].
In trace gas sensing applications, however, the species of interest are often
present in very low concentrations, so that self-broadening and broadening
against other trace gases can be neglected in calculations, and air-broadening
alone will suffice. For the conditions of atmospheric pressure broadening,
γ0
LPγD, the Doppler contribution to the overall linewidth can often be
neglected, and the line shape be treated as pure Lorentzian. Likewise, at
pressures low enough to ensure γ0
LPγD, the line shape can be treated
as pure Gaussian. In either case, calculation of the line profile is simplified
considerably.
At intermediate total pressures, γ0
LP≈γD, which for most lightweight
gases range from 5 to 100 Torr, calculation of the Voigt profile is necessary
to obtain numerically accurate absorption spectra. Methods for approximate
calculation of the Voigt profile, and the related plasma dispersion function,
are now a well-developed subject. The approximations published in [119]are
particularly useful.
3 Spectroscopic Techniques:
Signal Enhancement and Noise Reduction
In direct absorption approaches, quantitative information can be obtained
using the expressions discussed in the previous section.
This section only discusses in situ techniques, where the source, sampling
region and the detector are in close proximity. Active remote sensing such as
differential optical absorption spectroscopy (DOAS) and light detection and
ranging (LIDAR) are well developed, but not covered here and the interested

Mid-Infrared Laser Applications in Spectroscopy 467
reader is referred to the following recent book chapters by Platt [120]andby
Svanberg [121].
For in situ measurements, various sensitivity enhancement and noise re-
duction spectroscopic detection techniques have been developed in order to
achieve quantification of trace gas species at concentrations ranging from
ppmv to pptv. Each detection technique has its distinct advantages and
should be chosen depending on the specific application targeted, in particular
in terms of sensitivity and selectivity requirements.
Although laser source noise is an important aspect for sensitive detection,
in practice various other noise sources affect the measurement of a small
change of signal and can severely limit the detection sensitivity. Various noise
reduction techniques can be employed. These include modulation, balanced
beam and zero-background subtraction detection techniques. In addition, one
can improve the sensitivity by signal enhancement methods based on long
pathlength and cavity enhanced spectroscopy (Sects. 3.3 and 3.4), which are
capable of increasing significantly the effective sample optical pathlength L
to tens of meters and to kilometers, respectively, in absorption cells of typical
physical lengths of 0.3 to 1 m.
It is difficult to compare the expressions of sensitivity in terms of trace
gas detection systems that employ different sources and detection techniques
and their significance for a specific application. In this context, different ex-
pressions of sensitivities reported in the literature are a relative statement
and can be at times misleading if applied or compared to a different sig-
nal enhancement or noise reduction technique. Perhaps the most appropriate
method to compare the performance of trace gas detection systems (laser and
even non-laser based) is to determine the minimum detectable concentration
for the target molecule of interest for a given sampling and acquisition time.
For comparisons with other techniques one can relate this to a minimum de-
tectable absorbance per pathlength for the pathlength conditions employed
in the concentration measurement. We make this distinction since, for exam-
ple, in photoacoustic spectroscopy one obtains very high sensitivity for unit
pathlength, however it does not directly scale with increasing pathlength. In
many cases, such as isotopic ratio measurements, the replication precision is
most important. These and other attributes such as wavelength dependent
absorption strengths should be given detailed consideration before selecting
a laser source and technique for a specific trace gas sensor. Table 8gives an
overview of the most commonly used expressions and depicts examples of
values achieved and reported in the literature.
3.1 Balanced Beam and Balanced Ratiometric Detection
(Noise Reduction)
These techniques have been developed in order to eliminate technical noise
including laser intensity noise to approach the fundamental limit of shot-
noise. By measuring the laser signal with and without the absorption signal

468 Frank K. Tittel et al.
Tabl e 8 . Expressions of detection limits and sensitivity
Parameter Expression Common values/units
Minimum detectable
fractional absorption
∆Pmin
P010−4–10−7
Minimum detectable
fractional absorption
scaled to path length
∆Pmin
P0
1
L10−8–10−12 cm−1
Minimum relative
Detectable
Concentration
per unit volume (MDC)
∆Pmin
P0
1
L
1
σNtot
ppm (1 part in 1 million ) 10−6
ppb (1 part in 1 b illion) 10−9
ppt (1 part in 1 trillion ) 10−12
Minimum detectable
fractional absorption
scaled to path length,
relative concentration,
and per shot
measurement time
∆Pmin
P0
1
L
1
σNtot
1
√BW
10−8–10−12 cm−1Hz−1/2
BW = #pts
Tsamplenor ENBW
n
Bandwidth(BW) without or with
a frequency selective filter
(e.g. lock-in amplifier)
Absolute Laser
Instrument Response
Factor (LIRF)
∆Pmin
P0
1
Lwith
Tmeas,total
10−8cm−1(1 s)–10−11 cm−1(60 s)
Measurement
precision (measured
with a stable
input concentration)
Std.Dev. (LIRF) or
Std.Dev. (MDC)
Legend in order of appearance
P,opticalpower;L, length of effective light–matter interaction; σ, molecular
absorption cross-section [cm2]; Ntot , total number of molecules per unit vol-
ume; #pts, number of acquired points per scan; Tsampl e, time for single scan; n,
number of acquired scans; ENBW, equivalent noise bandwidth of a frequency
selective filter; Tmeas,total, measurement time to generate one concentration
data point
simultaneously, common mode noise can be subtracted and small absorption
signals can be recovered. Several approaches using dual-beam detection have
appeared in the literature and are briefly discussed here.
Conventional dual-beam detection systems use optical balancing
schemes [122]. The detected noise of an equal-intensity replica of a probe
beam, such as that created by a variable-ratio beamsplitter, is subtracted
from noise detected in the probe beam and thus leaving only the uncom-
pensated weak absorption signals of interest. For example, such a beamsplit-
ter can be realized by placing a polarization rotator (a half-wave plate) in
series with a polarizing beamsplitter cube. With the input polarization ro-

Mid-Infrared Laser Applications in Spectroscopy 469
tated about 45◦, the beams emerging from the beamsplitter cube carry equal
amounts of power P,andpowernoise∆P. In the absence of absorption,
the photocurrents generated by identical signal and reference detectors can
be subtracted to cancel each other exactly. If one of the beams is attenuated
due to small absorption a, by a gas, the balance of photocurrents is disturbed,
and a signal is seen at the output of the amplifier. Care must be exercised
to ensure that the signal and reference detectors have equal amplitude and
frequency responses.
An implementation of this method that avoids the need for exact balanc-
ing the signal and reference light was proposed by Hobbs [123]. It is known
as balanced ratiometric detection (BRD). It employs electronic circuitry to
produce a log ratio of photocurrents, rather than their difference, and to
cancel noise currents at the same time. This analog divider uses logarithmic
conformance and tight symmetry of base-emitter curves of a matched tran-
sistor pair. This scheme provides nearly perfect cancellation of noise currents
even when the reference beam carries twice the power of the signal beam.
Since the signal versus reference current balancing is performed by means
of electronic feedback, no physical adjustment of the beam splitting ratio is
necessary. The BRD differential response to absorption signals depends on
the ratio of the signal and reference currents, which changes when the sig-
nal beam is partially absorbed. It also depends on temperature because the
transistor base-emitter voltage does, and additional compensation circuitry
is needed to produce a useful output voltage Vout that is linearly propor-
tional to the absorbance. Noise-equivalent absorbances in the near-infrared
spectral region as low as 2 ×10−7Hz−1/2have been demonstrated by Allen
and co-workers [124], close to the limit imposed by shot noise. Application
of this concept to the mid-IR spectral region using a quasi-CW QC laser
was recently demonstrated by Sonnenfroh et al. [125] However, the sensitiv-
ity was reduced due to inherent differences of the employed mid-IR detectors
with large background currents derived from each detector and low average
photocurrents. Improvements to the design may well translate the results
obtained in the near-IR and yield sensitivities of ≈1×10−5.
Ratiometric detectors have been shown to greatly reduce technical noise
and enhance the sensitivity with short path distances. Technical noise origi-
nating from optical elements that are only present in either sample or refer-
ence beam are not cancelled. This limits the use of a BRD to noise reduction
with common optical paths. For example, technical noise from multi-pass
optical cells cannot be eliminated by this technique and hence extrapolation
of short-path sensitivities does not scale with longer pathlengths. BRD is
therefore well suited for low-noise measurement application using compact
short-path absorption cells.

470 Frank K. Tittel et al.
3.2 Wavelength and Frequency-Modulation Spectroscopy
(Noise Reduction)
As discussed in Sect. 1.1, the output wavelength of diode lasers can be
changed by either changing the entire device temperature or by changing the
injection current. The latter tuning mechanism is very rapid, thus making it
possible to employ high frequency modulation techniques in the kHz to MHz
regime. Traditionally, these two frequencies are used to describe two different
modulation approaches. Modulation employing frequencies in the kHz regime
is denoted as wavelength modulation spectroscopy (WMS). This approach,
which is also known as harmonic detection or derivative spectroscopy, uses
a modulation frequency (≈1–100 kHz) that is much less than the half width
of the laser source employed, which typically ranges between several MHz to
hundreds of MHz. By contrast, frequency modulation spectroscopy (FMS) is
characterized by modulation frequencies greater than the laser line half width
and typically involves frequencies up to several hundred MHz.
Frequency modulation is always accompanied by amplitude modulation,
as the injection current also controls the laser output power:
E(t)=A[1 + mcos(Ωt)] sin[ωt +βcos(Ωt +Φ)] .(6)
Here E(t) is the laser electric field, ω=2πc/λ the laser frequency, and
Ωthe modulation frequency. The quantities mand βare the amplitude
and frequency modulation indices, respectively, and Φis the generally non-
zero phase shift. Sine-wave modulation of the diode laser has the effect of
creating multiple side-bands in its otherwise nearly monochromatic emission
spectrum. Each side-band is separated from the carrier by an integer multiple
of the modulation frequency Ω, and its relative intensity depends on β.
In frequency-modulation spectroscopy, Ωsignificantly exceeds the laser
linewidth that is typically several tens of megahertz, and m, β are both small,
so that only the two first-order side-bands, ω+Ωand ω−Ω, have apprecia-
ble magnitude. After uniform attenuation, such as that encountered in non-
resonant optical systems or media, the side-bands add up coherently with the
carrier and balance each other to produce a beam of nearly constant inten-
sity, A2. If the attenuation strongly depends on frequency, however, as is the
case with most gases, one of the side-bands may become unbalanced and lead
to the appearance of multiple harmonics of Ωin the detected laser intensity.
The strength of absorption determines the magnitude of these harmonics,
which may be measured separately and with high noise immunity, by using
a lock-in amplifier for example. This is usually done while the laser carrier
frequency ωis scanned in the vicinity of the absorption line of interest.
This detection technique was first applied by Bjorklund toaCWdye
laser [126]. It has proved very effective and is used in diode laser spectroscopy
today, sometimes in modified form such as two-tone frequency-modulation
(TTFM) [127], or amplitude-modulated phase-modulation (AMPM) spec-
troscopy [128].

Mid-Infrared Laser Applications in Spectroscopy 471
In WMS, which is really another form of FM spectroscopy, the modulation
frequency Ωis smaller than the laser linewidth, and the modulation indices
mand βare both large [129,130,131]. The side-bands are then present to
a very high order and due to their small separation from each other, merge
into a continuous spectrum. The detection is again performed at the first,
second, or higher harmonics of Ωas the laser carrier frequency ωis scanned
in the vicinity of a gas absorption line. Lock-in amplifiers or mixers are em-
ployed in this approach in selecting the harmonic of choice, which in most
cases is the second harmonic. WMS is used in applications that rely on rel-
atively low-speed detectors, and its inherent sensitivity is typically limited
by the laser amplitude 1/f noise [132]. However, optical noise often limits
the sensitivity achievable using both WMS and FMS techniques in real world
systems. In addition, as discussed by Werle [133], an FM spectrometer can be
interpreted as an optimized single beam interferometer. In this model, as the
modulation frequency is increased the measurements become more sensitive
to small changes in the optical path, thus ultimately resulting in less stability
relative to WMS techniques. Both of these issues may explain the fact that
the expected improvements with FMS over WMS have not been realized on
a routine basis.
As an alternative to modulation spectroscopy, some research groups have
been very successful in achieving noise reduction by sweeping the laser injec-
tion current at kHz rates and detecting the resulting direct absorption spec-
trum using a signal averager (see for example, Zahniser et al. [134]). When
coupled with background subtraction (to be discussed in Sect. 4.1.1) absorp-
tion sensitivities in the 10−6range have been achieved. This approach has the
advantage over modulation techniques in providing a direct measure of the
sample concentration using the Beer–Lambert absorption law without the
need for calibration standards and lock-in amplifiers or mixers, required for
modulation approaches. In addition, as all modulation techniques effectively
smear the peak absorbance at the line center with absorbance in the wings,
the effective line center absorbance is reduced. Fried et al. [135], Iguchi [136]
and references therein indicate that the effective line center absorbance us-
ing modulation approaches is only 30–50 % of that achieved employing di-
rect absorption approaches. The exact reduction depends upon the particular
modulation function employed. On the other hand, modulation spectroscopy
presents some advantages over rapid sweep integration direct absorption spec-
troscopy. In modulation spectroscopy one has some flexibility to choose the
modulation amplitude and frequency to minimize dominant optical noise fea-
tures that may be present in direct absorption techniques. Furthermore, since
modulation techniques rely on a “fast” change in the absorption coefficient
with wavelength, these approaches discriminate against broad featureless ab-
sorptions, such as those from the wings of atmospheric pressure water lines
and those from big unresolved organic molecules. This aspect, which is of-
ten overlooked, becomes important as an added degree of selectivity when

472 Frank K. Tittel et al.
measuring trace gas concentrations in the atmosphere at levels of 100pptv
or less. Even though one may select an isolated absorption line to quantify
a particular gas of interest using the HITRAN database [117], there exists
the possibility that numerous broad featureless organic molecules, that can
be present in the atmosphere, may spectrally overlap the absorption line of
interest. In such cases, direct absorption measurements may yield systematic
errors.
3.3 Long Optical Path Length Spectroscopy
(Signal Enhancement)
One of the most obvious ways to enhance the absorption signal is inherent
to the Beer–Lambert absorption law, where the linear signal improves with
longer optical pathlengths. Traditionally, this has been implemented by the
use of optical multi-pass cells. Four types of multi-pass cells are most com-
monly applied: White cells [137,138], Herriott cells [139], Chernin [140]and
astigmatic mirror multi-pass cells [141]. For all four types of cells, the focusing
mirror curvature, applied to the beam at each reflection, keeps the beam from
diverging. The White cell [137,138] is the oldest arrangement. It consisits of
two semicircular mirrors, called the “D” mirrors, closely spaced along a com-
mon diameter facing a third notched mirror in a nearly confocal arrangement.
The probe beam enters through one notch and emerges through the other.
The number of passes is varied by changing the “D” mirror angle. The Her-
riott cell [139] has two identical spherical mirrors separated by nearly their
diameter of curvature (nearly concentric) facing each other. A probe beam
launched through a hole in one of the mirrors at an angle to the optical axis,
completes a certain number of passes between the mirrors, and exits through
the same hole (or a hole in the other mirror). The beam bounce pattern and
pathlength are controlled by adjusting the mirror separation. For both the
White and the Herriott configurations, the number of passes, if not limited
by attenuation of light due to the finite mirror reflectivity, is limited by the
overlapping of spots on the mirrors. Spot overlapping creates optical interfer-
ences causing base line oscillations superimposed on the absorption feature.
Astigmatic mirror cells [141] are variations of the Herriott cell that spread the
light spots over the entire mirror surfaces. This greatly increases the number
of spots achievable without overlapping spots and therefore the number of
passes. This cell type is also more compact and possesses the smallest cell
volume to effective path length ratio. In such a multi-pass cell, the number of
passes is typically configured for 90–238, which translates to effective optical
pathlengths from 18 m to 210 m for mirror separations of 0.3m–0.9m, thus
providing a commensurate improvement in signal strength. The cell volume
of multi-pass cells scales with the number of passes and mirror separation. Re-
spective volumes for the aforementioned astigmatic cells range from 0.3–5.2l.
Hence, longer optical pathlengths also increase the surface area and flushing

Mid-Infrared Laser Applications in Spectroscopy 473
time and ultimately determine the speed of measurement. The four multi-
pass arrangements are commonly enclosed in a vacuum housing, and used for
the measurement of static gas samples or controlled gas flows, but can also be
operated without enclosure for open-path ambient air trace gas monitoring
applications.
Multi-pass-cell mirrors can be configured for broadband laser wavelength
operation (from the near IR to the long-mid-IR region by use of gold or silver
coated mirrors) or can be optimized for specific wavelength regions by using
dielectrically coated mirrors similar to mirrors used in cavity ring down spec-
troscopy. Metallically coated mirrors have a typical reflectivity of >99 %,
resulting in a cell transmission of 16 % for 182 passes. Cell transmissions can
be greatly enhanced by the use of dielectrically coated mirrors, but at the
expense of a narrower wavelength operation. The beam entrance and output
coupling holes are a few mm in diameter. To avoid extensive beam aperture
clipping resulting in forward and backscattering, the entering beam must be
matched to the f/# of the multi-pass cell. Furthermore, the beam spots on
the mirror are separated by a finite width. Thus, the higher the f/#and
the smaller the beam size, the less optical interference may occur. The beam
pointing stability is another important factor as it can have a multiplicative
effect on the effective pathlength and hence produce jitter in the absorption
signal. Slow minute changes of the beam direction may also superimpose base-
line fluctuations. By flushing the multi-pass cell with “zero-air” and acquiring
the background, this effect can be captured and removed. To obtain a high
measurement duty cycle, the beam pointing instability must be minimized
and the multi-pass cell be mounted on a rugged platform.
3.4 Cavity-Enhanced Spectroscopy Methods
(Signal Enhancement)
Another technique, which takes advantage of long optical path length ab-
sorption in high finesse optical cavities, is called cavity-enhanced spec-
troscopy. Various methods have been developed. Cavity ring-down spec-
troscopy (CRDS), first demonstrated by O’Keefe and Deacon [142], is based
on the observation of the decay rate of an injected laser beam stored in
a cavity comprised of ultra-high reflective spherical mirrors. The rate of de-
cay (inverse of the ring-down time constant) is determined by a) mirror ab-
sorption and scattering, and b) wavelength dependent absorption loss by the
inserted sample gas. If the decay rate of a) is determined in the presence
of a non-absorbing gas or at a non-gas absorbing wavelength, then the gas
concentration is exactly proportional to the difference of the observed inverse
sample decay rates. Figure 7illustrates the concept of CRDS. Also shown is
the inherent difference to direct absorption spectroscopy. The decay rate is
independent of laser amplitude, which relaxes the requirements of the laser
source, in particular for pulsed sources with high pulse-to-pulse variations.

474 Frank K. Tittel et al.
Fig. 7. Concept of CRDS. Shown are the laser signal as a function of time and
wavelength before an absorption cell, and the time evolving signal at the detector,
after the laser radiation has stopped or is blocked from entering an absorption cell,
with and without a sample present. The difference of the decay rate of the sample
and ‘empty’ cell is recorded. The integrated area or difference of the inverse decay
rates is directly proportional to the gas concentration
Due to the large effective pathlengths (1 to 10 km is typical, 100 km has
been demonstrated), this technique offers significantly higher signal enhance-
ment than is obtainable in conventional absorption spectroscopy. A typical
CRDS cell has a base pathlength ranging from 0.5 m to 1 m. The mirrors in
such a cell can be designed for specific wavelength regions from the ultra-
violet (≈300 nm) to the mid-IR spectral region (10 µm). High reflectivity
coatings of 99.95–99.99 % in the mid-infrared are commercially available. Al-
though these are very small differences in the mirror reflectivity, the resulting
effective pathlength can be significant. For example, a 0.5 m cavity equipped
with identical 99.98 % r eflectivity mirrors yields a pathlength of ≈2500 m,
whereas with 99.99% reflectivity, an effective pathlength of 5000 m may be
obtained. These pathlengths give ring-down times exceeding 10 µs. If the
minimum detectable fractional absorption determined by the decay rates is
constant, this results in a 50 % change of minimum detectable concentration.
The high reflectivity is usually retained over ≈10 % of the center wavelength,
thus over a typical absorption scan no changes of pathlength need to be con-
sidered and even multiple species detection can be implemented with a single
set of mirrors. Tests with a variety of gases indicate stability of the mirror re-
flectivity without noticeable degradation [143]. Condensation on mirrors can
be avoided by heating the mirrors to 70 ◦C or higher without damage [144].
Nevertheless, one should avoid deposits of any kind on the mirrors, and one
method successfully used is to employ a clean air purge flow over the mirrors.
A second practical consideration relating to the very high mirror reflectivity
deals with the amount of light that can be injected into the cavity and the
amount that can be extracted and detected. For CRDS, the output intensity
can be on the order of 10 % of the input intensity. For non-resonant cavity
enhanced spectroscopy methods, the transmitted intensity is on the order of
the transmission through a single mirror.
In the following two sections, we provide an overview of the inherent mer-
its of this absorption measurement technique in the mid-IR wavelength re-
gion. We describe technical challenges and solutions of several approaches. We
will also give examples on how this technique compares to established spectro-

Mid-Infrared Laser Applications in Spectroscopy 475
scopic techniques. Recent articles will provide a more detailed discussion on
this subject than is possible within the scope of this chapter [145,146,147,148].
3.4.1 Cavity Ring-Down Spectroscopy
Cavity ring-down spectroscopy (CRDS) is a direct absorption technique,
which can be performed with pulsed or continuous light sources. This tech-
nique was derived from the characterization process of high reflectors, in
which the ring down rate indicated directly the reflectivity of the mirrors.
Since the original demonstration for use in spectroscopic measurements [142],
many papers have since reported the improvement and applicability of this
technique and extended it to longer near-IR and mid-IR wavelengths
[142,145,149,150,151,152,153,154,155]. Today, CRDS is used extensively in
the visible and near infrared. In one example, a pulsed laser is employed for
measurements of hydrogen-bonded clusters formed in molecular beams [156].
The progress in the mid-IR has been stimulated by the availability of ultra-
low loss cavity mirrors and convenient mid-IR solid-state sources, such as
pulsed quantum cascade laser and parametric conversion sources. With pulsed
lasers, CRDS requires a short laser pulse to be injected into a high finesse
optical cavity to produce a sequence of pulses leaking out through the end
mirror from consecutive traversals of the cavity by the pulse. Typically, the
laser pulse is short and has a small coherence length compared to a rela-
tively large physical cavity length. Under these conditions interference effects
are avoided and the intensity of the cavity pulses decays exponentially with
a time constant
τ=l
c
1
αl −ln R(7)
where αis the absorption coefficient of the intracavity medium, lis the cavity
length and Ris the mirror reflectivity (both mirrors are assumed to have the
same reflectivity and the refractive index of the medium is assumed to be 1).
By measuring the ring-down time, τ, without and with the absorbing gas
present, the value of αcan be determined. This technique is simple and
immune to laser power fluctuations.
In the previous discussion, pulsed laser sources with low coherence length
were considered and as a result of high-mirror reflectivities, the light levels
reaching the detector are small. An alternative approach utilizes the reso-
nance of the cavity by employing CW laser sources with a long coherence
length. This is illustrated in the “wavenumber time domain” depicted in
Fig. 8a. The laser line represented by the dashed curve is broader than the
cavity mode spacing, or free spectral range FSR = c/2l. The cavity through-
put can be made much higher if the laser linewidth ∆ νLFSR. This con-
dition can be satisfied if a narrow-line CW laser is used. When the laser
frequency coincides with one of the cavity modes as shown in Fig. 8b, the cav-
ity throughput is approximately T=∆νC/∆νL,where∆νCis the spectral

476 Frank K. Tittel et al.
(a)
(b)
Fig. 8. Two situations of laser radiation filtering by an op-
tical cavity (idealized one-dimensional consideration). (a)
The laser linewidth is much broader than the cavity mode
spacing (i.e. pulsed). In this case, the cavity throughput
does not depend on the laser linewidth but is solely de-
fined by the cavity finesse. (b) The laser linewidth is less
than the cavity mode spacing. The cavity throughput is
determined by the ratio of the cavity mode width to the
laser linewidth
width of the cavity mode (1–10 kHz). The laser emission can be interrupted
and the ring-down decay measured in the same manner as it is done with
a pulsed laser. The use of CW laser sources for ring down was first proposed
by Lehmann in 1996 [157], followed by Romanini et al. [158,159]usingCW
near-infrared DFB diode lasers. In order to maintain a good overlap of the
laser with the cavity mode, the two must be locked to each other.
The first work on CRDS measurements with a QC-DFB laser was reported
by Paldus et al. [160]. The authors used a CW laser generating 16 mW at
λ=8.5µm. The measured ring-down time of the empty three-mirror cavity
was 0.93 µs. An acousto-optic modulator was used to interrupt the cav-
ity injection for ring-down time measurements. The system was tested on
diluted ammonia mixtures, and a noise-equivalent sensitivity of 0.25 ppbv
achieved. An estimated 1.0×10−9cm−1detectable absorbance limit was re-
ported. A spectroscopic gas sensor for nitric oxide (NO) detection at 5.2µm
basedonCDRSwasreportedbyKosterev et al. [161]. Measurements of parts
per billion (ppb) NO concentrations in N2with a 0.7 ppb standard error for
an 8 s data acquisition time were performed. Interesting work [162,156]has
been done combining a novel OPO with cavity ring-down spectroscopy.
A practical advantage of pulsed CRDS over CW CRDS is its applicability
to study transient species formed for example by laser ablation sources [163].
3.4.2 Cavity-Enhanced Absorption Spectroscopy
A simpler (as compared to CDRS) method to exploit a high finesse op-
tical cavity for increasing the sensitivity to absorption has been devel-
oped [164,165,166,167,168,169] and is called “integrated cavity output spec-
troscopy” (ICOS) or “cavity-enhanced absorption spectroscopy” (CEAS).
Here, laser light is coupled into the high-finesse cavity via accidental coinci-
dences of the light with the cavity eigenmodes by dithering the cavity length.
The time-integrated intensity radiation leaking out of such an optical cavity,
averaged over many cavity modes, can be used to determine the absorption of
the intracavity medium. Effectively this is equivalent to a time integration of
the ring-down curve. Just as in cavity ring-down spectroscopy, an effective op-
tical pathlength of several kilometers can be obtained in a very small volume.

Mid-Infrared Laser Applications in Spectroscopy 477
However, in its most simple configuration, the noise levels are relatively high
since the cavity transmission varies significantly, depending on whether the
laser is on or off resonance with the cavity. The ability to effectively average
over these frequency response functions is the limiting noise factor to date.
A novel approach by Paul et al. [167] uses an off-axis cavity alignment similar
to the mirror configuration in astigmatic Herriott cells. This effectively lowers
the FSR of the cavity and generates a very dense cavity mode spectrum. In
fact, as the reentrant condition becomes longer (many multiple paths until
the input beam again overlaps with the cavity beam) the mode spacing will
become so dense that the cavity transmission will be almost independent of
the laser wavelength. This effectively eliminates the requirement of any laser
wavelength (as in CEAS) or cavity dither by a piezoelectric transducer (as in
ICOS). However, this in turn also collapses the Fabry–P´erot condition and
the transmission is significantly lowered and thus higher laser powers or more
sensitive detectors may be needed. This technique has been demonstrated
first in the visible wavelength region [167] and has been recently extended to
the near-IR region for measurements of a variety of species (CH4,C
2H2,CO,
CO2,NH
3) and demonstrated sensitivities of up to 3 ×10−11 cm−1Hz−1/2
at 1.51 µm[170]. A minimum detectable concentration of 0.3 ppb (S/N = 3)
of C2H2at 50 Torr was obtained (1 s averaging time). This latest develop-
ment demonstrates the ability of cavity enhanced spectroscopic techniques
to achieve similar detection sensitivities as obtained with traditional multi-
pass cell absorption spectroscopy. However, the true figure of merit for a real
world measurement is the ability to replicate the same result when one sam-
ples a constant input concentration. Since low frequency noise sources from
a variety of causes may also be important in the measurement process, the
ability to replicate the same answer in many successive measurements can be
far worse than the inherent sensitivity, and this will be further discussed in
Sect. 4.1.
One of the most advanced methods, which utilize cavity enhancement is
called “Noise-Immune Cavity-Enhanced Optical Heterodyne Spectroscopy”
(NICE-OHMS) technique [171,172]. This technique combines the power of
signal enhancement of cavity enhanced spectroscopy with the noise reduction
of FM spectroscopy. In this technique, the laser frequency is locked to the fre-
quency of a cavity mode. This method has the potential to provide shot-noise
limited sensitivity with an effective pathlength determined by the cavity ring-
down time. Ma et al. [171] reported a sensitivity of 10−14 cm−1. This spec-
tacular sensitivity is superior to that achieved with CRDS. To achieve such
sensitivity, the laser requires active frequency stabilization below the kHz
linewidth level, which is comparable or less than the spectral width of the
cavity mode. In addition NICE-OHMS typically requires milli-Torr sample
pressure, which effectively reduces the gas density by two orders of magni-
tude, and hence the effective minimum detectable concentration. High optical
saturation (≈300 W of power inside the cavity) makes the Beer–Lambert ab-

478 Frank K. Tittel et al.
sorption law no longer valid and hence the extraction of quantitative data
difficult. The first implementation of this technique using a QC-DFB laser
has been reported [173]. The NICE-OHMS approach is technically highly
sophisticated and is only suitable for fundamental laboratory applications.
The application and suitability of this technique for quantitative trace gas
measurements has not yet been demonstrated.
3.5 Photoacoustic and Photothermal Spectroscopy
(Signal Enhancement)
Photoacoustic spectroscopy (PAS) has found its principal use in sensitive
trace gas detection. It is based on the photoacoustic effect, in which acous-
tic waves result from the absorption of radiation. In its application to laser
spectroscopy, the laser beam impinges on a selected target gas in a specially
designed cell [174,175,176,177,178,179,180,181,182,183].
In contrast with other mid-IR absorption techniques, PAS is an indi-
rect technique in which the effect on the absorbing medium and not di-
rect light absorption is detected. Light, from either pulsed or chopped CW
laser sources produces a transient temperature rise in an absorbing medium
via non-radiative relaxation processes, which then translates into a pressure
change or sound wave as illustrated in Fig. 9. This is detected with a sensi-
tive microphone(s). The acoustic signal is directly related to the concentra-
tion of the absorbing molecules in the cell. For CW laser sources, there are
two modes of operation for PAS, either the exciting light can be modulated
at a frequency away from any cell resonance or it can be adjusted to coin-
cide with an acoustic resonant frequency. The in-resonance mode is usually
employed with the low-power pump lasers to provide larger signals. How-
ever, precautions may be necessary to minimize changes of the instrument
response due to the change of the speed of sound caused by temperature
and gas compositional changes. PAS is ideally a background-free technique:
the absorbing gas generates the signal, and in the absence of an absorbing
gas there is no acoustic signal. In real PAS experiments, background signals
Fig. 9. Principle of photoacoustic spectroscopy. The incoming photons excite the
target molecule at a resonant wavelength. Collisional de-excitation converts the
absorbed energy into local heating and pressure waves, which can be detected by
a microphone. (Illustration by courtesy of M. Webber, Pranalytica, Inc.)

Mid-Infrared Laser Applications in Spectroscopy 479
can originate from nonselective absorption of the gas cell windows (coherent
noise) and from outside acoustic (incoherent) noise, and from scattering of
the laser radiation by aerosols onto the microphone. PAS signals are pro-
portional to the pump laser intensity and therefore PAS is mostly used with
high-power laser sources, in particular CO2and CO lasers [2,184]. In addi-
tion, diode (and in combination with high power optical fiber amplifiers) [185]
and QC lasers, solid-state lasers, DFG and OPOs in the infrared have been
applied to photoacoustic trace gas detection. Recently, a new approach called
quartz enhanced PAS has been developed by Kosterev et al. [186]. Instead of
using a gas-filled resonant acoustic cavity, the sound energy is accumulated
in a high-Q crystal element. Feasibility experiments utilizing a quartz watch
tuning fork demonstrate a sensitivity of 1.2×10−7cm−1W/√Hz.
However, in trace gas monitoring applications, PAS is limited to extractive
point monitoring due to the requirement of an absorption cell. In addition,
PAS requires sufficient sampling pressures (≈100 Torr −1 atm) for efficient
collisional transfer and generation of the acoustic waves, thus limiting the
selectivity in some cases. Furthermore, the effective collisional transfer can
depend on the relative composition of the gas sample. For example, the de-
excitation rates and hence the strength of the generated acoustic wave can
differ by a factor of ≈2, as shown by Fried and Berg for the mid-IR detec-
tion of HCl under condition of dry samples and samples with 44 % relative
humidity [187].
Another form of the photoacoustic effect is called photothermal absorp-
tion spectroscopy. Here, the photoacoustic signal is detected by means of
recording the phase change in, for example, an unfolded Jamin interferome-
ter. In this case, a non-resonant probe laser beam (e.g. He–Ne) is spatially
overlapped with the resonant excitation beam (e.g. CO2-laser). The probe
beam is split and directed through a parallel path without the absorber. The
modulated phase difference of these probe beams is measured [188]. Such
a system has been configured for the detection of NH3and demonstrated
a2σprecision of 250 ppt and 31 ppt in a 1 s and 100 s integration time, re-
spectively.
The key features of the photoacoustic technique include (1) excellent de-
tection sensitivities down to sub-ppbv concentrations with powers in the
watt range, (2) a large dynamic range, (3) PAS detector responsivity is al-
most independent of the pump wavelength, and (4) a PAS signal that is
directly proportional to the absorbed radiation intensity, but does not scale
with pathlength as with the previously discussed signal enhancement tech-
niques. Indeed, the PAS signal will increase if a laser beam passes through the
same volume/detection area of a microphone. However, it must pass through
the same limited small interacting volume, otherwise more microphones are
needed to be co-located along the laser path. Each added microphone will
add to the noise floor. Therefore, only a moderate increase of the signal to

480 Frank K. Tittel et al.
noise per unit Hz−1/2measurement time can be attained. On the other hand,
the photothermal approach does scale linearly with pathlength.
Implementation of a QC-DFB laser to target fundamental absorptions
has the potential of considerably improved flexibility and allows one to ac-
cess many absorption features, whereas line-tunable CO2laser sources depend
upon accidental overlap, and are restricted to the 9–11 µm region. Ammonia
and water vapor photoacoustic spectra were obtained using a CW cryogeni-
cally cooled QC-DFB laser with a 16 mW power output at 8.5µmasreported
by Paldus et al. [189]. A PAS cell resonant at 1.66 kHz was used. The QC-DFB
was used for frequency scans using temperature tuning and for real-time con-
centration measurements with a fixed laser temperature. Measured concen-
trations ranged from 2,200 ppmv to 100 ppbv. A detection limit of 100 ppbv
ammonia (≈10−5noise-equivalent absorbance) at standard temperature and
pressure was obtained for a 1 Hz bandwidth and a measurement interval of
10 min.
Recently, Hofstetter et al. [190] reported PAS measurements of ammonia,
methanol and carbon dioxide using a pulsed 10.4µm QC-DFB laser operated
at 3–4 % duty cycle with 25 ns long current pulses (2 mW average power)
and close to room temperature with Peltier cooling. Temperature tuning
resulted in a wavelength range of 3cm−1with a linewidth of 0.2cm
−1.This
sensor used a 42 cm long PAS cell with a radial 16-microphone array for
increased detection sensitivity. In addition the cell was placed between two
concave reflectors resulting in 36 passes through the cell (with an effective
pathlength of 15 m). The laser beam was mechanically chopped at a resonant
cell frequency of 1.25 kHz, with a PAS signal enhancement by a Q factor
of 70. A pyroelectric detector recorded the QC laser power to normalize the
PAS signal. Detection of ammonia concentrations at the 300 ppbv level with
a SNR of 3 was achieved at a pressure of 400mbar.
4 Mid-Infrared Spectroscopic Applications
Tunable mid-infrared spectroscopic sources and spectroscopic techniques pro-
vide four important performance characteristics: sensitivity, selectivity, fast
response time, and compactness. These are also based on similar optical com-
ponent design and hence offer the unique ability to mix and match laser
sources and techniques to be most useful for a given application. Recent
progress in this field and growing optical industrial resources (e.g. optical
fiber telecommunication) has led to the evolution and utilization of mid-IR
spectroscopic techniques to a wide range of gas sensor applications. These
include such diverse fields as: 1) environmental monitoring (CO, CO2,CH
4
and CH2O are important gas species in various aspects of atmospheric chem-
istry studies); 2) industrial emission measurements (e.g. fence line perimeter
monitoring in the petrochemical industry, combustion sites, waste inciner-
ators, down gas well monitoring, gas–pipe and compressor station safety);

Mid-Infrared Laser Applications in Spectroscopy 481
3) urban (e.g. automobile traffic, power generation) and rural emissions (e.g.
horticultural greenhouses, fruit storage and rice agro-ecosystems); 4) chemical
analysis and control for manufacturing processes (e.g. in the semiconductor,
pharmaceutical, and food industries); 5) detection of medically important
molecules (e.g. NO, CO, CO2,NH
3,C
2H6and CS2), toxic gases, drugs, and
explosives relevant to law enforcement and public safety; and 6) spacecraft
habitat air-quality and planetary atmospheric science (e.g. such planetary
gases as H2O, CH4,CO,CO
2and C2H2).
4.1 Detailed Examples Using Selected Spectroscopic Sources
and Techniques
There are literally hundreds of examples that can be cited where various mid-
IR sources and spectroscopic techniques as discussed in the previous sections
have been employed, including studies in the laboratory and industrial set-
tings, and on ground-based, aircraft, balloon-borne, and rocket-borne plat-
forms. Rather than cite examples of each, we present in this section three rep-
resentative examples of spectroscopic sources utilizing lead-salt diode lasers,
quantum cascade lasers, and difference-frequency generation coupled with ei-
ther long-path and dual-beam absorption, wavelength modulation, or cavity-
enhanced spectroscopy.
4.1.1 Lead-Salt Diode Laser Based Spectrometer
for Airborne Atmospheric Chemistry Studies
In the following, we describe a lead-salt diode laser based trace gas sensor,
which incorporates advances collectively developed by many research groups
over the years, and thus illustrate the high performance that can be achieved
routinely in a rugged airborne field setting. Aircraft measurements present
unique and demanding challenges for one has to contend with: (1) changes
in cabin temperature by as much as 20 ◦C over relatively short time periods
of 1 h or less; (2) changes in cabin pressure by as much as 100 to 300 mbar
over time periods of several minutes; (3) changes in system attitude; and (4)
changes in aircraft vibrations that can couple beneficially or detrimentally
into the system. As will be shown, by careful attention to numerous details,
one can routinely measure absorbances as small as 0.7 to 1.7×10−6for 1-min
integration times employing pathlengths of 100 m on aircraft platforms.
Figure 10a illustrates the optical layout for a dual channel airborne spec-
trometer, which contains two Pb-salt diode lasers mounted in a liquid ni-
trogen dewar. Figure 10b shows a three-dimensional diagram of this system
mounted in a temperature stabilized enclosure. This system has successfully
acquired ambient measurements of the important atmospheric gas, formalde-
hyde (CH2O), on numerous airborne campaigns. As shown, this system was
also configured for simultaneous measurements of hydrogen peroxide (H2O2),
another important atmospheric trace gas. The performance of this second

482 Frank K. Tittel et al.
Fig. 10. (a) Optical set-up of a dual channel absorption laser spectrometer
(DCALS); (b) three-dimensional digitally rendered model of DCALS configured
for airborne operation
channel, however, was significantly inferior to that of the CH2O channel, and
will not be discussed here. Comprehensive details regarding the airborne sys-
tem, the associated airborne measurements, as well as detailed background
information on calibration, sampling, measurement accuracy, inlet tests, and
ground-based and airborne comparisons studies can be found in Fri ed et al.
[16] and references therein.
The IR radiation from the lead diode laser operating at 3.5µm is directed
through a multi-pass astigmatic Herriott cell (Aerodyne Research, Inc.) using
a series of off-axis mirrors (two parabolic and one elliptical mirror), as shown
in Fig. 10a. The IR beam, which traces out a Lissajous pattern in the cell,
achieves a total optical pathlength of 100m in a 3-l sampling volume. Upon
exiting the cell, the IR beam is directed onto sample and reference indium-
antimonide photovoltaic detectors. The optical system employs a minimum
number of components, and of those, only two are readily adjustable. Each
component, furthermore, is mounted on a rigid mount with as low a center
height as possible. Both measures attempt to avoid subtle mechanical align-
ment changes. The entire optical enclosure, including the optical bench, is
temperature stabilized to around 30 ◦C, typically to better than ±1◦Cover
time periods of many hours and significantly better than this over shorter
time periods. A series of heaters mounted in the lid (not shown) are used
for this purpose. Sheets comprised of an aluminum-balsa wood sandwich
(partially removed in drawing) are mounted to the frame structure shown
in Fig. 10b. These sheets provide both good thermal insulation and struc-
tural support. All the above precautions are critical for high performance
aircraft measurements and are essential in extending the system stability pe-
riod (to be discussed) out to 1-min and longer. Without temperature control,
for example, one encounters rather significant changes in optical alignment,

Mid-Infrared Laser Applications in Spectroscopy 483
background optical structure, and detector dark count signals, as the air-
craft repeatedly ascends and descends throughout a typical flight pattern. In
addition, owing to repeated changes in cabin pressure, it is equally critical
to control the liquid-nitrogen pressure using some type of absolute pressure
control valve. Without such control, the liquid-nitrogen boiling point would
change, resulting in rather significant shifts in laser wavelength as the cryo-
gen base temperature changes. Such temperature changes could also result
in changes in dewar dimensions, and this may produce a consequent change
in diode laser position which is integrally mounted to the dewar casing.
Absorption data are acquired using second harmonic detection coupled
with sweep integration, as discussed by Fried et al. [12]. In this procedure, the
diode laser wavelength is repetitively scanned across an isolated absorption
feature of CH2O (2831.6417 cm−1) using a 200-point sawtooth ramp applied
to the laser tuning current at a frequency of 50Hz. A 50-kHz quasi-square
wave modulation waveform is simultaneously applied to the laser tuning cur-
rent, and the 2f signal at 100 kHz is detected using a digital lock-in amplifier.
Second harmonic signals from the reference arm, which contains a high con-
centration cell of pure CH2O, are treated identical to the sample arm. The
lock-in amplifier outputs are then digitized (16-bit A/D converters) and co-
averaged by a computer. The line center of the reference arm, which has
very high signal-to-noise and serves as a wavelength reference, is determined
on every scan. Each scan is then appropriately shifted in memory to align
the peak centers before co-averaging. This fast spectral shifting is extremely
important for high instrument performance [191]. Upon completion of each
ambient measurement block, typically 60 s (3000 scans), the wavelength ref-
erence line center is determined using a polynomial fit, and an appropriate
correction voltage is applied to the laser current controller to keep the ab-
sorption feature centered in the scan window. In addition, the mean value
for the amplitude of each scan is forced through zero before co-averaging.
This procedure effectively removes small scan-to-scan dc variability and thus
improves the co-averaging effectiveness. The co-averaged spectrum at the end
of each scan cycle is then transformed into the frequency domain employing
an FFT algorithm, band pass filtered, and transformed back into the time
domain. This approach helps to reduce both high and low frequency noise
without significantly affecting the retrieved ambient signals.
By far the most dramatic improvement in instrument performance
is achieved using rapid background subtraction. Fr ied et al. [15,12], Zah-
niser et al. [134], and Werle et al. [191] among others have presented the mer-
its of this approach for tunable diode laser absorption spectroscopy. If carried
out correctly, rapid background subtraction effectively captures and removes
optical noise, which ultimately limits the performance of most if not all tun-
able diode laser instruments. As discussed previously, such noise is caused by
light scattering from various optical elements, and this generates a somewhat
random undulating background structure. Often such structure contains mul-

484 Frank K. Tittel et al.
tiple frequencies, amplitudes, and time constants originating from multiple
scattering sources. As addressed by Werle et al. [191], acquisition of sample
and background spectra along with the associated cell flush times need to be
accomplished within a characteristic system stability period, topt,inorderfor
background subtraction to be effective. One can characterize this time period
using the Allan variance, as first presented by Werle et al. [191] and subse-
quently by Fried et al. [12]. During time periods of constant optical enclosure
temperature and constant pressure (not presently controlled in our airborne
system), system stability periods typically range between 40 and 100 s before
drifts become prevalent.
During airborne operation ambient air is drawn through a heated Teflon
inlet at typical flow rates between 8 and 10 standard liters min−1(slm, where
standard conditions are defined as 273 K and 1 atmosphere pressure) and
through the Herriott cell at sampling pressures around 40 Torr. Background
spectra are acquired by passing ambient air from a second inlet through
aheatedPd/Al
2O3scrubber, which removes CH2O without significantly af-
fecting the ambient water concentration, and this airflow is re-routed to the
inlet tip at flow rates exceeding the sample flow. During a typical sampling
sequence[16], six successive 10 s ambient spectra are acquired, and each 1 min
ambient sampling block is preceded and followed by a 10 s background ac-
quisition. An appropriate delay period of 9 s (≈5 inlet/cell e-folding resi-
dence times) is employed after each switch before data are acquired. The
backgrounds surrounding each 1 min block are averaged (time weighted) and
subtracted point by point from each of the 10 s ambient spectra. The average
laser power is determined for each 1 min sampling block using a third data
acquisition channel to continuously record the sample detector dc voltage.
The detector dark voltage is obtained every few 1-minute ambient cycles by
blocking the laser beam with a shutter for a few seconds. The laser power
is determined from the difference of the two measurements, and the ratio of
the laser power obtained during calibration to that during a sample measure-
ment is applied. Calibration spectra are typically acquired every 30–60 min by
adding CH2O standards, from a permeation calibration system, to the zero
airflow near the inlet entrance. At the cell flows employed, typical CH2O
standard concentrations of 12–14 parts-per-billion by volume (ppbv) are gen-
erated at the Herriott cell entrance. Periodically, 7 ppbv standards are also
added on top of ambient to check for inlet loss and to check the veracity of
the data retrieval algorithm. Each 10 s background-subtracted ambient spec-
trum is fitted in real time to a background-subtracted calibration spectrum
(acquired for 20 s) employing a multiple linear regression approach [192,193].
Each complete ambient acquisition cycle, which includes the acquisition of
a 10 s background, 1 minute of ambient averaging, two 9 s delay periods and
computer-processing overhead, typically takes 90 s, and this typically falls
within the system stability period.

Mid-Infrared Laser Applications in Spectroscopy 485
1000
900
800
700
600
500
400
300
200
100
0
Number of Measurements
2001901801701601501401301201101009080706050403020100
1-Minute Measurement Precision [pptv ]
A = 1 x 10
-6
Fig. 11. Formaldehyde (CH2O) measurement precision diagram at the 1σlevel for
airborne measurements throughout the TOPSE campaign [16]
Figure 11 shows the instrument performance obtained during the TOPSE
(Tropospheric Ozone Production about the Spring Equinox) airborne cam-
paign [16]. Here the 1-minute (1σ) measurement precision, obtained from
the standard deviation of replicate measurements upon sampling relatively
constant and low ambient CH2O levels, is given in terms of a histogram.
The histogram reflects the fact that the measurement precision during air-
borne operation is somewhat variable, depending upon the exact alignment
stability and the degree to which the acquired backgrounds truly represent
the actual backgrounds underlying the ambient spectra. As the real precision
for any technique will vary, even in a laboratory setting, the histogram ap-
proach gives a more realistic assessment of overall instrument performance
than a single figure of merit often reported for many techniques.
In addition, the results of Fig. 11 are obtained from replicate measure-
ments, and not the precision of any given measurement, which in the case
of tunable diode laser measurements, can be expressed in terms of an in-
dividual fit precision. As discussed by Fr ied et al. [12] this fit precision is
proportional to the square root of the fit deviations, and in most instances
was found to be a factor of 3 to 4 too optimistic. In this case, like that
for many other techniques, there are additional sources of variance, which
produce ambient results that have larger run-to-run variability than the pre-
cision of any individual measurement. Thus, from these two standpoints the
results of Fig. 11 truly represent the meaningful performance that is obtain-

486 Frank K. Tittel et al.
able with an airborne tunable diode laser system. As can be seen, most of the
1 min measurements (which require 90 s to acquire) yield 1σreplication pre-
cisions in the 20–50 pptv concentration range (median value = 40 pptv) for
CH2O during airborne operation. Employing a sampling pressure of 40 Torr,
a temperature of 303 K, a 100 m pathlength, an integrated absorption cross-
section of 5.44 ×10−20 cm2cm−1molecule−1and an air-broadening coefficient
of 0.107 cm−1atm−1, one calculates using a Voigt function that a line center
absorbance of 1×10−6equatestoaCH
2O concentration of 30 pptv. Thus the
precisions above correspond to minimum detectable line center absorbances
of 0.7 to 1.7×10−6. This in turn computes to pathlength-normalized values
of 7 ×10−11 to 1.7×10−10 cm−1employing a 90 s sampling sequence.
4.1.2 Quantum-Cascade Laser Based Trace Gas Sensors
in the Life Sciences
Laser spectroscopy is finding increasing applications in medicine and the life
sciences [194,195,196,197,198,199]. A particular role for spectroscopy is in the
monitoring of small molecules that have been shown to be important. The
role of simple molecules such as nitric oxide (NO) in physiological processes
has received considerable attention in recent years and was a subject of the
1998 Nobel Prize in Medicine.
One application is the measurement of NO in human breath samples, since
exhaled air is an indicator of several processes taking place in the human
body, in particular in assessing the severity of airway inflammation [200]. To
observe NO in breath, a cavity-enhanced absorption spectroscopy (CEAS)
sensor [169] with a CW QC-DFB laser operating at 5.2µm with an output
power of 80 mW was used. A direct performance comparison was carried out
between a sensor configuration where the CES optical cavity was replaced
with a 100 mpathlength multi-pass cell. It was found that in spite of having an
effective pathlength of 670 m, CES had a lower absorption sensitivity because
of baseline noise of ≈1 % (averaging 104QC laser scans). These baseline
fluctuations are intrinsic to CES and result from the mode structure of the
cavity transmission spectrum. Some improvement in the CES baseline noise
can be achieved with the recently developed off-axis ICOS technique [167].
In [161] a spectroscopic gas sensor for NO detection based on a cavity ring-
down technique is described. NO is the major oxide of nitrogen formed during
high-temperature combustion as well as an important nitrogen-containing
species in the atmosphere (NO is a precursor of smog and acid rain). NO is
also involved in a number of vital physiological processes, and its detection
in human breath has potential applications (e.g. as a marker for diseases like
asthma or inflammatory processes) in noninvasive medical diagnostics. A CW
QC-DFB laser operating at 5.2µm was used as a tunable single-frequency
light source. The technique used consists of the following features:

Mid-Infrared Laser Applications in Spectroscopy 487
1) The laser frequency is slowly scanned across the absorption line of interest.
2) One of the cavity mirrors is dithered back and forth to ensure periodic,
random coincidences of the laser frequency with a cavity mode.
3) Once such a resonance occurs and the cavity is filled, the laser beam
entering the cavity is abruptly interrupted or set off-resonance, and the
decay rate of the exiting light is measured.
From (7) in Sect. 3.4, the absorption coefficient can be determined as
α=1
c1
τ−1
τempty (8)
where τempty is the decay constant of the cavity without absorber.
The sensor schematic is shown in Fig. 12. It depicts a simpler design than
the first CRDS experiment with a QC-DFB laser described in [160]. In that
work a variable temperature cryostat for laser frequency tuning was used
with the inherent complexity and additional cost of an acousto-optic modu-
lator (AOM) to interrupt the QC laser beam. In the work reported in [161]
both frequency tuning and the laser emission interruption were realized by
manipulation of the QC laser pump current. No active temperature control
was applied to the QC-DFB laser located in a liquid nitrogen optical cryostat:
The laser current was supplied by a low-noise current source and monitored
using the 0.5 Ω resistor denoted by r1in Fig. 12. The laser frequency was
tunable from 1922.9 to 1920.8cm
−1when the pump current changed from
300 mA (lasing threshold) to 660 mA. At higher current levels the laser emis-
sion was multimode. The tuning range permitted NO detection by accessing
absorption lines at 1921.599 cm−1and 1921.601 cm−1(R(13.5) components of
Gas flow
PZT Driver
Function
Generator
Function
Generator
Current
Source
Reference Cell
Ringdown Cell
IR Detectors
A/D Converter
Trigger Circuit
QCL
Removable
mirror
r
r
2
1
Fig. 12. Schematic of a QC laser based cavity ring-down spectrometer [161]

488 Frank K. Tittel et al.
the fundamental absorption band). Absorption lines of water vapor and CO2
were also observed. The l= 37 cm long, linear high-finesse optical cavity was
formed by two concave mirrors with a 6 m radius of curvature. The measured
ring-down time of the cavity without absorber was τ≈3.5µs, corresponding
to ∆ νC≈45 kHz. The laser linewidth was estimated to be ∆ νL=3MHz.
When a certain level of the detector signal was reached signaling that
aTEM
00 cavity mode is coupled to the laser, a triggering circuit (TC) opened
a metal-oxide semiconductor field-effect transistor (MOSFET) to shunt the
laser current, thereby reducing it to a subthreshold value. At the same time,
the TC triggered an analog–digital converter, and the detector signal show-
ing the cavity ring-down was digitized for ≈30 µs and stored in computer
memory. The MOSFET was kept open for 35 µs, so that the laser radiation
would not interfere with measurements of the cavity decay constant. This
triggering-and-acquisition process was repeated and consecutive ring-down
transients were stacked in the A/D memory until a desired number of tran-
sients was acquired. The results were post-processed to fit each transient
with an exponential decay function yielding a ring-down time τ.Theinverse
ring-down time plotted as a function of the laser current provides the absorp-
tion spectrum. An example of the NO absorption in a mixture with pure N2
is presented in Fig. 13. The noise-equivalent sensitivity was estimated to be
0.7 ppbv for an 8 s data acquisition time. It was not possible to use this sensor
directly for measurements of NO concentration in exhaled air (≈10 ppbv)
because of strong CO2interference, which can be avoided if the appropriate
NO absorption line is chosen (e.g. like R(7.5) components at 1903.123 and
1903.134 cm−1) with a QC laser that accesses this wavelength.
-0.04 -0.02 0.00 0.02 0.04
0.28
0.30
0.32
0.34
0.36
0.38
0.40
(a)
1/
W
,
P
s
-1
Rela tive fr eque ncy, cm
-1
Fig. 13. NO absorption in a mixture with pure N2obtained with the sensor depicted
in Fig. 12

Mid-Infrared Laser Applications in Spectroscopy 489
Recent work indicates that other gases, such as carbon monoxide (CO),
can also play a very significant physiological role. CO is produced from heme
catabolism by the enzyme heme oxygenase. Previous work has shown that
CO promotes blood flow by inhibiting vascular tone and platelet aggregation
and that neuronal CO production may modulate the NO-cGMP (guanosine
3,5
-cyclic monophosphate) signaling system, demonstrating important bio-
chemical interactions between the two diatomic gases. The extremely low
levels of gas production in living cells and the relatively short in vivo lifetime
of cell cultures have complicated detailed understanding of the kinetic, or
time-dependent processes responsible for their generation. A typical produc-
tion rate of CO, for example from vascular smooth muscle cells (VSMCs), is
1to10pmol/min/10
7cells. Instrumentation for in vivo measurement of gas
production should have sensitivities on the parts per billion (ppb) level in
order that the dynamics of gas production can be followed with laboratory-
scale cell sample populations [201]. Because of low CO production rates from
biological tissues, measurements of CO concentrations have been limited to
gas chromatography and radioisotope counting techniques. Although these
methods are highly sensitive, they cannot measure CO directly, requiring
several time-consuming intermediate steps requiring ≈15 min, and may be
affected by interference from water, oxygen, and carbon dioxide.
Infrared laser absorption spectroscopy is an attractive alternative ap-
proach for the detection of biological CO at the parts-per-billion (ppb) level
in real time [202,203]. A compact gas sensor measured endogenenous CO pro-
duction from vascular cells using a mid-IR spectroscopic laser source based
on differencefrequency generation (DFG) of two near-IR lasers. The CO ab-
sorption was detected in the fundamental vibration band near 4.6µm. In this
work, an extractive technique was used with gas samples taken from the flask
containing the cell culture to an 18m pathlength optical multi-pass cell so
that the measurements could be performed at a reduced pressure of 100 Torr.
Kosterev et al. [203] reported an improved design and performance of an
optical mid-IR CO sensor intended for continuous monitoring of cell culture
activity at ambient atmospheric pressure. The same fundamental absorption
band region was used for CO detection, but a pulsed quantum cascade laser
with a distributed feedback structure (QC-DFB) [25] was employed instead
of the DFG source. The high output power of the QC-DFB laser and an
advanced data analysis approach made it possible to detect biological CO and
CO production rates with ≈1 m optical pathlength folded above a standard
culture flask of VSMCs.
A further improvement of the pulsed QC-DFB based sensor was reported
by Kosterev et al. [204]. The laser beam was split into two channels, one be-
ing used to probe the gas absorption and the other as a reference to measure
the laser pulse energy. The subsequent normalization eliminated pulse-to-
pulse energy fluctuations as an error source, which was the predominant
cause of error previously [203]. This automated sensor was used for contin-

490 Frank K. Tittel et al.
uous monitoring of CO in ambient air detected by its R(3) absorption line
at 2158.300 cm−1(λ≈4.6µm). A noise-equivalent detection limit of 12 ppbv
was experimentally demonstrated with a 1 m optical pathlength. This sensi-
tivity corresponds to a standard error in fractional absorbance of 3×10−5.
All the measurements were carried out at atmospheric pressure, and hence
it was not possible to periodically acquire a baseline with an evacuated sam-
ple container. In order to keep the baseline (which included weak unwanted
interference fringes from optical elements) stable during multi-hour measure-
ments, the slow drifts of the laser frequency were actively compensated by
computer-controlled corrections to the subthreshold current. A constant CO
production rate of 44 ppbv/hour was observed, taking into account the 0.5
liter volume of the cell culture container. This corresponds to a net CO pro-
duction rate of 0.9 nmol/107cells/hour, which is in agreement with previous
measurements [202] obtained with similar cells and treatment regimes.
A compact mobile ammonia sensor based on a thermoelectrically cooled
pulsed QC-DFB laser operating at ≈10 µmwasdescribedin[205]. High sen-
sitivity detection of NH3is also of interest in the control of deNOxchemistry,
industrial safety and medical diagnostics of kidney related diseases. The opti-
cal configuration of this sensor was similar to that described in [206], but the
multi-pass cell was replaced with a simple 50 cm long double pass gas cell,
and no zero air subtraction employed. The laser housing was improved by
replacing the previous beam-shaping optical system consisting of two off-axis
parabolic mirrors and a lens with a single aspheric lens. The laser was scanned
over two absorption lines of the NH3fundamental ν2band. The sensor was
completely automated and only required the LN2dewar of the detector to
be refilled every 12 h. This sensor was applied to the continuous monitoring
of NH3concentration levels at the ppm level present in bioreactor vent gases
in a water reprocessing system located at NASA’s Johnson Space Center in
Houston, TX. A sensitivity of better than 0.3 ppmv was estimated which was
sufficient to quantify expected ammonia levels of 1 to 10ppmv.
4.1.3 Design and Applications
of Fiber Based Difference Frequency Based Mid-IR Gas Sensors
In the following, the optical architecture and performance of several field
portable gas sensors based on difference frequency sources are discussed. The
gas sensors described here utilize fiber based near-IR lasers, high-power rare
earth doped fiber amplification and single-pass difference-frequency genera-
tion in QPM-PPLN crystals (Sect. 1.2).
Difference frequency generation utilizing optical fiber coupled and fiber
based pump sources allows great flexibility in designing a robust mid-infrared
source with power levels typically ranging from 0.1mW to 1mW. Figure 14a
depicts four representative fiber based DFG-source configurations, which have
been developed and applied to trace gas detection. In order to evaluate and

Mid-Infrared Laser Applications in Spectroscopy 491
Fig. 14. (a) Fiber based difference-frequency generation sources using narrow-
linewidth diode laser sources and high-power optical fiber amplifiers. Shown are
respective wavelengths, input and generated power levels [207]. (b) Optical fiber
pumped DFG beam profile: measured 37.5 cm from PPLN exit facet, beam diameter
@1/e2:x=0.80 mm, y=0.80 mm Gaussian fit >95 % [208]
confirm robust performance of these sources itself and their applicability, di-
rect long-pathlength absorption as the simplest form of sensitive detection
was used. Advanced signal enhancing techniques and noise reduction tech-
niques have been applied and laboratory results indicate unique advantages
of these sources.
In the examples shown, low-power seed diode laser sources are used in
combination with high-power, wide bandwidth fiber amplifiers. Several inex-
pensive low-power diode laser sources whose wavelengths overlap with the
gain region of the fiber optic amplifier can be used, one at a time, or can
also be multiplexed to be amplified simultaneously or sequentially in time.
This permits easy modification of a fixed optical fiber platform by choosing
any desirable seed wavelength within the fiber amplifier(s) gain bandwidth
in the difference frequency mixing process and generate the desired mid-IR
wavelength(s) (Fig. 1). Fiber amplifiers retain the spectroscopic properties
of the seed laser sources, and thus decouple the high-power requirement for
efficient DFG from low-noise, narrow-linewidth operation. This approach is
more cost effective and technically easier than constructing a pump laser
source which meets all of the three requirements at the same time, namely

492 Frank K. Tittel et al.
narrow-linewidth, high-power (W), and multiple-wavelength operation. In ad-
dition, near-IR diode lasers have similar tuning rates with current and tem-
perature. This in turn provides predictable ease of exchange of sources and
mid-IR frequency stabilization. For example, an ambient temperature change
will introduce the same wavelength shift to two near IR diode laser sources
with similar tuning rates and use of similar types of current/temperature
controller. If those laser sources are difference-frequency mixed, the drift is
subtractive and affords high inherent mid-IR wavelength stability, provided
the diode lasers and controllers are carefully selected and operate at the same
ambient conditions. The use of fiber optics for DFG pump beam delivery also
provides a stable and inherent spatial overlap to produce a circular, homoge-
neous beam with a near-Gaussian intensity distribution as shown in Fig. 14b.
This offers small beam sizes (≈2mm) with f/#≈100, and are obtained by
the use of only one relay optical element to collect and image the DFG beam
to the spectroscopic cell (e.g. multi-pass cell).
Quasi-phase matched periodically poled LiNbO3(QPM-PPLN) has been
shown to be efficient for parametric frequency down conversion (see Sect. 1.2).
It also offers flexibility in the conversion bandwidth by either using crys-
tals incorporating multiple QPM periods or a fan-out design for continuous
QPM [69].
The DFG sources depicted in Fig. 14a have been used in several field ap-
plications to demonstrate their feasibility in real-world environments. These
include remote volcanic gas measurements in Nicaragua, operation in an in-
dustrial setting and analysis of a space rated trace contaminant control sys-
tem (using the source shown in Fig. 14aA)[209,210]. The DFG source de-
picted in Fig. 14aB),showninmoredetailinFig.15, is configured for high
sensitivity dual-beam long-path absorption spectroscopy of urban pollutants.
Using this source, Rehle et al. [211] conducted extensive urban gas detec-
tion of CH2O over extended time periods and achieved a detection limit of
0.32 ppbv of atmospheric formaldehyde at 3.53 µm (2832 cm−1). This corre-
sponds to a sensitivity of 1 ×10−9cm−1in combination with a 100 m path-
length low-volume (3.3 l) astigmatic mirror Herriott gas cell. A dual-beam
absorption configuration that employs two dc-coupled Peltier-cooled HgCdTe
(MCT) detectors was used to eliminate the optical interference fringes origi-
nating from the refractive optical elements of the DFG conversion stage and
optical fiber components. A typical absorption spectrum obtained is shown in
Fig. 16. To further enhance the signal-to-noise, mainly limited by electronic
noise and technical noise from the multi-pass cell, wavelength modulation
and zero-background subtraction could be used as described in the lead-salt
diode laser based detection system above. However, these techniques were
not used with this device, because of relatively high urban formaldehyde
concentrations ranging from 1 to 50 ppbv. The sensor as shown in Fig. 15
has been operated autonomously for a continuous nine and five-day period
at two separate field sites in the Greater Houston area, administered by

Mid-Infrared Laser Applications in Spectroscopy 493
Fig. 15. High-power continuous-wave DFG source employing dual-beam spec-
troscopy used for urban pollution monitoring of CH2O. DL, diode laser; DFB-DL,
distributed feedback-DL; DBR-DL, distributed Bragg reflector-DL; OI, optical iso-
lator; WDM, wavelength division multiplexer
the Texas Natural Resource Conservation Commission (TNRCC) and the
Houston Regional Monitoring Corporation (HRM). The acquired spectro-
scopic data were compared with results obtained by a well-established wet-
chemical o-(2,3,4,5,6-pentafluorobenzyl) hydroxylamine (PFBHA) technique
with good agreement. While the accuracy of the determined concentrations
is comparable with results from conventional wet-chemical techniques, the
described DFG sensor offers excellent time resolution on the order of seconds
and permits unattended continuous operation for long periods of time. The
maintenance-free design of a tunable infrared DFG based diode laser spec-
trometer and the capability of remotely controllable computerized operation
makes such instrumentation a convenient, robust tool for mobile trace gas
detection. Thus, formaldehyde concentration measurements using direct ab-
sorption laser spectroscopy have proved to be a sensitive and effective method
for online trace gas monitoring in an urban setting. Further testing of a similar
DFG system by Richter et al. [208] utilizing additional wavelength modula-
tion spectroscopy and zero air background subtraction has generated labo-
ratory based replicate detection precisions (1σ, 1 min average) of better than
2.5×10−10 cm−1, which corresponds to 74 pptv minimal detectable CH2O
concentration.

494 Frank K. Tittel et al.
0.99970
0.99975
0.99980
0.99985
0.99990
0.99995
1.00000
1.00005
2831.61 2831.64 2831.66 2831.69 2831.72
Wavenumbers [cm
-1
]
Transmission
concentration:
(8.49 ± 0.57) ppbv
goodness of fit:
F
2
= 3.4272 E-10
V = ± 1.852 E-05
Fig. 16. Absorption spectra of CH2O using the DFG based gas sensor depicted in
Fig. 15
The recent rapid advances in the optical fiber telecommunication indus-
try has produced a number of new laser sources and components to con-
tinuously expand the DFG wavelength coverage and generate higher mid-
IR power at the mW level. Many of the requirements for telecom applica-
tions overlap and directly enhance the spectroscopic performance of DFG
based mid-IR sources. These include narrower linewidth (<1MHz,1s), sta-
ble higher power using highly saturated two-stage fiber amplifiers, and wave-
length stability. Relatively simple, yet powerful, mid-IR sources can be de-
signed. Fig. 14a D) shows the optical design of an all telecom wavelength
based mid-IR source [212]. Here, an optically pumped VCSEL operating at
980 nm provides over 500 mW coupled into a single-mode optical fiber with
linewidths of ≈10 MHz. The other DFG pump source is a standard low noise
and narrow-linewidth (<1 MHz) DFB diode laser amplified by a 500 mW
Er/Yb fiber amplifier. Mid-IR powers in excess of 0.25 mW at 2.64 µmare
easily obtained with this DFG mixing configuration in combination with a
2 cm long QPM-PPLN crystal. Using this source, spectroscopic detection of
fundamental absorption lines of hydrogen fluoride (HF) has been performed.
Employing direct absorption spectroscopy, fractional absorption sensitivities
of 1 ×10−4was obtained. The DFG frequency stability of this source was
measured by tracking a high resolution HF absorption line over several hours,
whichindicatedamaximumpeak-to-peak wavelength drift of 20 MHz during
a 2 h time interval. Given the strong absorption cross-section of HF at this
wavelength range (S=1×10−18 cm molecule−1), sub-ppb detection sensi-
tivity can be obtained over relatively short pathlengths of several meters.
The spectroscopic performance characteristics of DFG sources and their
inherent flexibility to access specific fundamental mid-IR absorption lines
can be further facilitated for precision ratio measurements of, for example,
isotopic 12CO2/13CO2. The ability of measuring isotopic ratios with a high

Mid-Infrared Laser Applications in Spectroscopy 495
precision and selectivity makes this technique particularly attractive for a va-
riety of applications, such as carbon cycle research, volcanic gas emission
studies, and the use of isotopic tracers in medical diagnostics. In order to
be useful, measurement precisions of 0.180
/00 to 10
/00 are required and are
defined as [(Rsample −Rstd )/Rstd]×1000, where R=[
13CO2]/[12 CO2]. Iso-
topic ratios can be measured by detection of closely paired absorption lines
of 12CO2/13 CO2found for example in the 4.3 micron wavelength range and
comparing this spectroscopic signatures with a known reference gas. To ob-
tain high precision, several key requirements of the source and spectroscopic
technique have to be addressed. Most important are the inherent stability
and measurement precision of absorption spectra and the temperature de-
pendence of the probed absorption lines. Closely paired absorption lines of
12CO2/13 CO2with similar absorption strengths usually originate from differ-
ent ground energy levels and hence have different temperature dependences.
Figure 17 shows three suitable line pairs for the detection of 12 CO2/13CO2
isotopic ratios. Also indicated are the temperature coefficients based on the
Boltzman distribution [213].
Using the DFG source depicted in Fig. 14a A) in combination with
a single-pass 20 cm long dual chamber absorption cell, Erdelyi et al. [214]have
obtained a 1σmeasurement precision of ≈0.80
/00. The dual chamber absorp-
tion cell was built from a solid piece of brass and incorporated two parallel
small bore extrusions and was end-fitted with common Brewster angle win-
dows. This allowed rapid comparison measurements of sample gases with
2286 2289 2292 2295 2298 2301 2304
95
96
97
98
99
100
2295.8 2295.9 2296.0 2296.1
98
99
100
0
/
00
K
-1
b: 18.8
0
/
00
K
-1
a: 22.9
c: 18.9
0
/
00
K
-1
L=0.1 m
p=40 Torr
c
CO2
= 350 ppm
Voigt lineshape
c
b
a
Transmission (%)
Frequency (cm
-1
)
12
CO
2
13
CO
2
Fig. 17. HITRAN simulation of isotopic CO2lines

496 Frank K. Tittel et al.
a known reference gas mixture. The cell design also ensured good thermal
mixing of the sample and reference gas standard and because of the small
volume, a minimal amount of expensive reference gases. The measurement
precision of this system was mainly limited by electronic noise due to low
pump power from an ECDL operating at the edge of its tuning range, and in
turn only provided a mid-IR power of ≈0.2µW. However, the achieved preci-
sion is sufficient for many applications including volcanic gas emission studies.
Other approaches with similar detection levels ranging from 0.20
/00 to 20
/00
have been reported in the literature, including lead-salt diode, color-center-,
CO2-, and near-IR diode laser sources in combination with various multi-pass
cell designs and cavity enhanced spectroscopy [213,215,216,217,218].
5 Summary and Outlook
This chapter has attempted to survey the current status of various tunable
CW solid-state laser based sources and techniques suitable for mid-infrared
laser applications in spectroscopy. The emphasis of our discussion has been
to acquaint the reader with the key fundamentals and options in realizing
optimized performance, different available sensitivity enhancement schemes
and minimum detectable absorbances (10−4to 10−6). Wherever possible, we
tried to elucidate the factors that are important for real-world sensors and
applications. In this regard we presented three representative examples of
laser sources, techniques and their specific applications. Each of these critical
areas will constantly undergo incremental improvements of the underlying
enabling technologies and lead to further advances in in situ and remote gas
sensing techniques.
Reliable mid-IR sources in combination with cavity-enhanced spec-
troscopy and optical fiber technologies will improve detection limits. These
new technologies and approaches will also lead to new effective sensor con-
figurations. For example, one could imagine the use of low-loss single mode
fibers with high reflective coatings employed as a ring down cavity and com-
bined with evanescent field absorption of a partially stripped fiber. These
potential and other new inventions will improve the simplicity, the cost and
robustness of spectroscopic laser based gas sensors and broaden their range
of applications [219,220,221].
As these new laser sources and spectroscopic techniques evolve in ma-
turity, an emphasis on the instrument replicate sensitivity and precision of
quantitative trace gas detection must be given, because it is the unequivocal
merit of usefulness in many applications. Such issues may not solely depend
on the laser or measurement principle employed, but also on numerous other
factors which may be application dependent, such as ambient temperature
and pressure conditions. It will thus be important to address these impor-
tant key factors without increasing the system complexity. Such spectroscopic

Mid-Infrared Laser Applications in Spectroscopy 497
laser based systems will have a large impact on the means and quality with
which we can sense the world around us.
Acknowledgements
The authors would like to thank Dr. Douglas S. Baer (Los Gatos Research,
Inc.), Dr. James F. Kelly (Pacific Northwest National Laboratories), Dr. Ana-
toliy A. Kosterev and Dr. Robert F. Curl (Rice University), and Dr. Michael
E. Webber (Pranalytica, Inc.) for their helpful ideas, comments and invalu-
able scientific discussions during the preparation of this manuscript.
References
1. L. R. Narasimhan, W. Goodman, C. K. N. Patel: Correlation of breath am-
monia with blood urea nitrogen and creatinine during hemodialysis, Proc.
National Acad. Sci. 98, 4617–4621 (2001) 447
2. M. E. Webber, M. Pushkarsky, C. K. N. Patel: Fiber-amplifier enhanced pho-
toacoustic spectroscopy using near-infrared tunable diode lasers, Appl. Opt.
42, 2119–2126 (2003) 447,479
3. J. Reid, D. T. Casidy, R.T. Menzies: Linewidth measurements of tunable
diode lasers using heterodyne and ´etalon techniques, Appl. Opt. 21, 3961–
3965 (1982) 448
4. S. Lundqvist, J. Margolis, J. Reid: Measurements of pressure-broadening coef-
ficients of NO and O3using a computerized tunable diode laser spectrometer,
Appl. Opt. 21, 3109–3113 (1982) 448
5. R. L. Sams, A. Fried: Microphone triggering circuit for elimination of mechan-
ically induced frequency-jitter in diode laser spectrometers: implications for
quantitative analysis, Appl. Opt. 26, 3552–3558 (1987) 448
6. R. Sams, A. Fried: Potential sources of systematic errors in tunable-diode-laser
absorption measurements, Appl. Spectrosc. 40, 24–29 (1986) 449
7. H. Preier, Z. Feit, J. Fuchs, D. Kostyk, W. Jalenak, J. Sproul: Status of lead-
salt diode laser development at spectra-physics, in Monitoring of Gaseous
Pollutants by Tunable Diode Lasers, Proc. Int. Symposium, Freiburg, Ger-
many (1988), G. S. R. Grisar, M. Tacke, G. Restelli (Eds.) (Kluwer Academic,
Dordrecht 1989) pp. 85–102 449
8. D. L. Wall, J. C. Sproul, Z. Feit, G. W. Sachse: Development of IR tunable
diode lasers and source assemblies for atmospheric monitoring and related ap-
plications, in Tunable Diode Laser Spectroscopy, Lidar, and DIAL Techniques
for Environmental and Industrial Measurements, A. Fried, D. K. Killinger,
H. I. Schiff (Eds.) (SPIE, Atlanta, GA 1993) pp. 2–11 449
9. M. Tacke: Recent results in lead-salt laser development at the IPM, in Moni-
toring of Gaseous Pollutants by Tunable Diode Lasers, Proc. Int. Symposium,
Freiburg, Germany (1988), G. S. R. Grisar, M. Tacke, and G. Restelli (Eds.)
(Kluwer Academic, Dordrecht 1989) pp. 103–118 449
10. R. Grisar: Monitoring of Gaseous Pollutants by Tunable Diode Lasers,Proc.
Int. Symposium, Freiburg, Germany (1991) (Kluwer Academic, Dordrecht
1992) 449

498 Frank K. Tittel et al.
11. D. J. Brassington: Tunable diode laser absorption spectroscopy for the mea-
surement of atmospheric species, in Advances in Spectroscopy,R.E.Hester
(Ed.) (John Wiley, New York 1995) pp. 83–148 449
12. A. Fried, B. Henry, B. Wert, S. Sewell, J. R. Drummond: Laboratory, ground-
based, and airborne tunable diode laser systems: performance characteristics
and applications in atmospheric studies, Appl. Phys. B 67, 317–330 (1998)
449,483,484,485
13. P. Werle: A review of recent advances in semiconductor laser-based gas mon-
itors, Spectrochim. Acta 54, 197–236 (1998) 449
14. A. Fried, S. McKeen, S. Sewell, J. Harder, B. Henry, P. Goldan, W. Kuster,
E. Williams, K. Baumann, R. Shetter, C. Cantrell: Photochemistry of
formaldehyde during the 1993 tropospheric OH photochemistry experiment,
J. Geophys. Res. Atmospheres 102, 6283–6296 (1997) 450
15. A. Fried, S. Sewell, B. Henry, B. P. Wert, T. Gilpin, J. R. Drummond: Tun-
able diode laser absorption spectrometer for ground-based measurements of
formaldehyde, J. Geophys. Res. Atmospheres 102, 6253–6266 (1997) 450,
483
16. A. Fried, Y. Wang, C. Cantrell, B. Wert, J. Walega, B. Ridley, E. Atlas,
R. Shetter, B. Lefer, M. T. Coffey, J. Hannigan, D. Blake, N. Blake,
S. Meinardi, B. Talbot, J. Dibb, E. Scheuer, O. Wingenter, J. Snow, B. Heikes,
D. Ehhalt: Tunable diode laser measurements of formaldehdye during the
TOPSE 2000: Distributions, trends, and model comparisons, J. Geophys. Res.
108, D4, 8365 (2003) 450,453,482,484,485
17. A. N. Baranov, Y. Cuminal, G. Boissier, C. Alibert, A. Joulli´e: Low-threshold
laser diodes based on type-II GaInAsSb/GaSb quantum-wells operating at
2.36 µm at room temperature, Electron. Lett. 32, 2279–2280 (1996) 451
18. Y. Rouillard, F. Genty, A. Perona, A. Vicet, D. A. Yarekha, G. Boissier,
P. Grech, A. N. Baranov, C. Alibert: Edge and vertical surface emitting lasers
around 2.0–2.5µm and their applications, Philos. Trans. R. Soc. London A
359, 581–597 (2001) 451
19. J. C. Nicolas, A. N. Baranov, Y. Cuminal, Y. Rouillard, C. Alibert: Tunable
diode laser absorption spectroscopy of carbon monoxide around 2.35 µm, Appl.
Opt. 37, 7906–7911 (1998) 451
20. B. Lane, D. Wu, A. Rybaltowski, H. Yi, J. Diaz, M. Razeghi: Compressively
strained multiple quantum well InAsSb lasers emitting at 3.6µmgrownby
metal-organic chemical vapor deposition, Appl. Phys. Lett. 70, 443–445 (1997)
451
21. P. Werle, A. Popov: Application of antimonide lasers for gas sensing in the
3–4-µm range, Appl. Opt. 38, 1494–1501 (1999) 451
22. P. Werle, R. Mucke, F. D’Amato, T. Lancia: Near-infrared trace-gas sensors
based on room-temperature diode lasers, Appl. Phys. B 67, 307–315 (1998)
451
23. P. Werle: Spectroscopic trace gas analysis using semiconductor diode lasers,
Spectrochim. Acta A 52, 805–822 (1996) 451
24. T. C. Hasenberg, R. H. Miles, A. R. Kost, L. West: Recent advances in Sb-
based midwave-infrared lasers, IEEE J. Quant. Electron. 33, 1403–1406 (1997)
451
25. F. Capasso, C. Gmachl, R. Paiella, A. Tredicucci, A. L. Hutchinson,
D. L. Sivco, J. N. Baillargeon, A. Y. Cho, H. C. Liu: New frontiers in quan-

Mid-Infrared Laser Applications in Spectroscopy 499
tum cascade lasers and applications, IEEE J. Sel. Topics Quant. Electron. 6,
931–947 (2000) 451,489
26. C. Gmachl, A. Straub, R. Colombelli, F. Capasso, D. L. Sivco, A. M. Ser-
gent, A. Y. Cho: Single-mode, tunable distributed-feedback and multiple-
wavelength quantum cascade lasers, IEEE Quantum Electron. 38, 569–581
(2002) 451
27. C. Gmachl, F. Capasso, D. L. Sivco, A. Y. Cho: Recent progress in quantum
cascade lasers and applications, Rep. Prog. Phys. 64, 1533–1601 (2001) 451
28. R. M. Williams, J. F. Kelly, J. S. Hartman, S. W. Sharpe, M. S. Taubman,
J. L. Hall, F. Capasso, C. Gmachl, D. L. Sivco, J. N. Baillargeon, A. Y. Cho:
Kilohertz linewidth from frequency-stabilized mid-infrared quantum cascade
lasers, Opt. Lett. 24, 1844–1846 (1999) 451,452
29. M. Beck, D. Hofstetter, T. Aellen, J. Faist, U. Oesterle, M. Ilegems, E. Gini,
H. Melchior: Continuous wave operation of a mid-infrared semiconductor laser
at room temperature, Science 295, 301–305 (2002) 452
30. R.Kohler,C.Gmachl,A.Tredicucci,F.Capasso,D.L.Sivco,S.N.G.Chu,
A. Y. Cho: Single-mode tunable, pulsed, and continuous wave quantum-
cascade distributed feedback lasers at lambda congruent to 4.6–4.7µm, Appl.
Phys. Lett. 76, 1092–1094 (2000) 452
31. J. Faist, D. Hofstaetter, M. Beck, T. Aellen, M. Rochat, S. Laser: Bound-to-
continuum and two-phonon resonance quantum cascade lasers for high duty
cycle, high temperature operation, IEEE J. Quant. Electron. 38, 533–546
(2002) 452
32. M. S. Taubman, T. L. Myers, B. D. Cannon, R. M. Williams, F. Capasso,
C. Gmachl, D. L. Sivco, A. Y. Cho: Frequency stabilization of quantum-
cascade lasers by use of optical cavities, Opt. Lett. 27, 2164–2166 (2002)
452
33. H. Q. Le, G. W. Turner, J. R. Ochoa, M. J. Manfra, C. C. Cook, Y. H. Zhang:
Broad wavelength tunability of grating-coupled external cavity mid-infrared
semiconductor lasers, Appl. Phys. Lett. 69, 2804–2806 (1996) 452
34. G. P. Luo, C. Peng, H.Q. Le, S. S. Pei, W. Y. Hwang, B. Ishaug, J. Um,
J. N. Baillargeon, C. H. Lin: Grating-tuned external-cavity quantum-cascade
semiconductor lasers, Appl. Phys. Lett. 78, 2834–2836 (2001) 452,453
35. K. G. Libbrecht, J. L. Hall: A low noise high-speed diode laser controller, Rev.
Sci. Instrum. 64, 2133 (1993) 452
36. R. Q. Yang, J. L. Bradshaw, J. D. Bruno, J. T. Pham, D. E. Wortman: Mid-
infrared type-II Interband cascade lasers, IEEE Quantum Electron. 38, 559–
568 (2002) 453
37. G. P. Luo, C. Peng, H. Q. Le, S. S. Pei, H. Lee, W. Y. Hwang, B. Ishaug,
J. Zheng: Broadly wavelength-tunable external cavity mid-infrared quantum
cascade lasers, IEEE J. Quant. Electron. 38, 486–494 (2002) 453
38. D. D. Nelson, M. S. Zahniser, J. B. McManus, C. E. Kolb, J. L. Jimenez: A tun-
able diode laser system for the remote sensing of on-road vehicle emissions,
Appl. Phys. B 67, 433–441 (1998) 453
39. D. E. Cooper, C. B. Carlisle: High-sensitivity FM spectroscopy with a lead-salt
diode laser, Opt. Lett. 13, 719–721 (1988) 454
40. L. F. Mollenauer, White, J. C.: General principles and some common features,
in Tunable Lasers, L. F. Mollenauer, White, J. C. (Eds.) (Springer, Berlin Hei-
delberg 1987) pp. 1–18 454

500 Frank K. Tittel et al.
41. P.F. Moulton: Tunable solid-state lasers, Proc. IEEE 80, 348–364 (1992) 454
42. N. P. Barnes, F. Allario: Tunable solid-state lasers: an emerging technology
for remote sensing of planetary atmospheres, Proc. SPIE 868, 107–116 (1988)
454
43. S. K¨uck: Laser-related spectroscopy of ion-doped crystals for solid-state lasers,
Applied Physics B-Lasers & Optics B 72, 515–562 (2001) 454
44. A. Sennaroglu, A. O. Konca, C. R. Pollock: Continuous-wave power perfor-
mance of a 2.47-µmCr
2+: ZnSe laser: Experiment and modeling, IEEE
J. Quant. Electron. 36, 1199–1205 (2000) 454
45. I. T. Sorokina, E. Sorokin, A. Di Lieto, M. Tonelli, R. H. Page, K. I. Schaffers:
Efficient broadly tunable continuous-wave Cr2+ : ZnSe laser, J. Opt. Soc. Am.
B18, 926–930 (2001) 454
46. S. Schiller, J. Mlynek: Special Issue, Continuous-wave optical parametric oscil-
lators, materials, devices, applications, Appl. Phys. B 66, 664–764 (Springer,
Berlin, Heidelberg 1998) 455
47. P. E. Powers, T. J. Kulp, S. E. Bisson: Continuous tuning of a continuous-
wave periodically poled lithium niobate optical parametric oscillator by use
of a fan-out grating design, Opt. Lett. 23, 59–61 (1998) 455
48. R. J. Lang, D. G. Mehyus, D. F. Welch: External cavity, continuously tunable
wavelength source, US patent no. 5,771,252 (1998) 456
49. A. T. Schremer, C. L. Tang: External-cavity semiconductor laser with
1000 GHz continuous piezoelectric tuning range, IEEE Photon. Technol. Lett.
2, 3–5 (1990) 456
50. D.Wandt,M.Laschek,K.Przyklenk,A.T¨unnermann, H. Welling: External
cavity laser diode with 40 nm continuous tuning range around 825 nm, Opt.
Commun. 130, 81–84 (1996) 456
51. T. Kiguchi, A. Uematsu, M. Kitano, H. Ogura: Grating external cavity diode
lasers with broad tunable range and narrow spectral linewidth for high-
resolution spectroscopy, Jap. J. Appl. Phys. 35, 5890–5895 (1996) 456
52. B. Pezeshki: New approaches to laser tuning, Opt. Photon. News 12, 34–38
(2001) 456
53. L. A. Coldren: Monolithic tunable diode lasers, IEEE J. Sel. Topics Quant.
Electron 6, 988–999 (2000) 456
54. B. Schmidt, S. Illek, R. Gessner, M. C. Amann: Design and realization of
a buried-heterostructure tunable-twin-guide laser diode with electrical block-
ing regions, IEEE J. Quant. Electron. 35, 794–802 (1999) 456
55. G. Morthier, B. Moeyersoon, R. Baets: A lambda/4-shifted sampled or su-
perstructure grating widely tunable twin-guide laser, IEEE Photon. Technol.
Lett. 13, 1052–1054 (2001) 456
56. H. Ishii, H. Tanobe, F. Kano, Y. Tohmori, Y. Kondo, Y. Yoshikuni: Broad-
range wavelength coverage (62.4 nm) with superstructure-grating DBR laser,
Electron. Lett. 32, 454–455 (1996) 456
57. E. T. Wetjen, D. M. Sonnenfroh, M. G. Allen, T. F. Morse: Demonstration of
a rapidly tuned Er3+-doped fiber laser for sensitive gas detection, Appl. Opt.
38, 3370–3375 (1998) 457
58. H. Simonsen, J. Henningsen, S. Sogaard, J. E. Pedersen: CO2stabilized Er3+-
doped fibre laser at 1578 nm, Eur. Frequency Time Forum (EFTF) (2000)
457

Mid-Infrared Laser Applications in Spectroscopy 501
59. I. Freitag, A. T¨unnermann, H. Welling: Power scaling of monolithic miniature
Nd:YAG ring lasers to output powers of several watts, Opt. Commun. 115,
511 (1995) 457
60. D. Boucher, W. Chen, J. Burie, P. Peze: A novel CW optical laser-based
difference-frequency infrared spectrometer, Ann. Phys. (Paris) 23, 247–248
(1998) 457
61. W. D. Chen, J. Burie, D. Boucher: Investigation on infrared laser absorption
spectroscopy measurement of acetylene trace quantities, Infrared Phys. Tech-
nol. 41, 339–348 (2000) 457
62. W. Schade, T. Blanke, U. Willer, C. Rempel: Compact tunable mid-infrared
laser source by difference frequency generation of two diode-lasers, Appl. Phys.
B63, 99–102 (1996) 457
63. M. Seiter, M. W. Sigrist: Trace-gas sensor based on mid-IR difference-
frequency generation in PPLN with saturated output power, Infrared Phys.
Technol. 41, 259–269 (2000) 457
64. D. Lee, T. Kaing, J. J. Zondy: An all-diode-laser-based, dual-cavity AgGaS2
CW difference-frequency source for the 9–11 µm range, Appl. Phys. B 67,
363–367 (1998) 457,459
65. H. D. Kronfeldt, G. Basar, B. Sumpf: Application of a CW tunable infrared
spectrometer based on difference-frequency generation in AgGaS2for self-
broadening investigations of NO at 5 µm, J. Opt. Soc. Am. B 13, 1859–1863
(1996) 457
66. M. Seiter, D. Keller, M. W. Sigrist: Broadly tunable difference-frequency spec-
trometer for trace gas detection with noncollinear critical phase-matching in
LiNbO3, Appl. Phys. B 67, 351–356 (1998) 457
67. M. Seiter, M. W. Sigrist: Compact gas sensor using a pulsed difference-
frequency laser spectrometer, Opt. Lett. 24, 110–112 (1999) 457
68. D.G. Lancaster, A. Fried, B. Wert, B. Henry, F.K. Tittel: Difference-
frequency-based tunable absorption spectrometer for detection of atmospheric
formaldehyde, Appl. Opt. 39, 4436–4443 (2000) 457,459
69. D. Richter, D. G. Lancaster, F. K. Tittel: Development of an automated dio de-
laser-based multicomponent gas sensor, Appl. Opt. 39, 4444–4450 (2000) 457,
459,461,492
70. W. D. Chen, F. Cazier, F. Tittel, D. Boucher: Measurements of benzene con-
centration by difference-frequency laser absorption spectroscopy, Appl. Opt.
39, 6238–6242 (2000) 457,459
71. D. G. Lancaster, D. Richter, F.K. Tittel: Portable fiber-coupled diode-laser-
based sensor for multiple trace gas detection, Appl. Phys. B 69, 459–465
(1999) 457,459
72. U. Simon, F. K. Tittel: Nonlinear optical frequency conversion techniques,
in Atomic, Molecular, and Optical Physics: Electromagnetic Radiation,
R. Celotta, R. G. Hulet, F. B. Dunning (Eds.), Experimental Methods in the
Physical Sciences 29, 231–278 (Academic Press, New York 1997) 457
73. A. S. Pine: Doppler-limited molecular spectroscopy by difference-frequency
mixing, J. Opt. Soc. Am. 64, 1683–1690 (1974) 458
74. B. Sumpf, D. Rehle, T. Kelz, H. D. Kronfeldt: A tunable diode-laser spectrom-
eter for the MIR region near 7.2µm applying difference-frequency generation
in AgGaSe2, Applied Physics B-Lasers & Optics 67, 369–373 (1998) 459

502 Frank K. Tittel et al.
75. P. A. Budni, L. A. Pomeranz, M. L. Lemons, C. A. Miller, J. R. Mosto,
E. P. Chicklis: Efficient mid-infrared laser using 1.9-µm-pumped Ho:YAG and
ZnGeP2optical parametric oscillators, J. Opt. Soc. Am. B 17, 723–728 (2000)
459
76. W. C. Eckhoff, R . S. Putnam, S. X. Wang, R . F. Curl, F. K. Tittel: A continu-
ously tunable long-wavelength CW IR source for high-resolution spectroscopy
and trace-gas detection, Appl. Phys. B 63, 437–441 (1996) 459
77. W. Chen, G. Mouret, D. Boucher: Difference-frequency laser spectroscopy
detection of acetylene trace constituent, Appl. Phys. B 67, 375–378 (1998)
459
78. R. S. Putnam, D. G. Lancaster: Continuous-wave laser spectrometer automat-
ically aligned and continuously tuned from 11.8 to 16.1µmbyuseofdiode-
laser-pumped difference-frequency generation in GaSe, Appl. Opt. 38, 1513–
1522 (1999) 459
79. W. K. Burns, W. McElhanon, L. Goldberg: Second harmonic generation in
field poled, quasi-phase matched, bulk LiNbO3, IEEE Photon. Technol. Lett.
6, 252–254 (1994) 459
80. L. E. Myers: Periodically poled materials for nonlinear optics, in Advances in
Lasers and Applications, D. M. Finlayson, B. D. Sinclair (Eds.) (Scottish Univ.
Science Press, 1999) pp. 141–180 459
81. K. Fradkin-Kashi, A. Arie, P. Urenski, G. Rosenman: Mid-infrared difference-
frequency generation in periodically poled KTiOAsO4and application to gas
sensing, Opt. Lett. 25, 743–745 (2000) 459
82. K. Fradkin, A. Arie, A. Skliar, G. Rosenman: Tunable midinfrared source by
difference frequency generation in bulk periodically poled KTiOPO4, Appl.
Phys. Lett. 74, 914–916 (1999) 459
83. K. Fradkin-Kashi, A. Arie: Multiple-wavelength quasi-phase-matched nonlin-
ear interactions, IEEE J. Quant. Electron. 35, 1649–1656 (1999) 459
84. S. Sanders, R. J. Lang, L. E. Myers, M.M. Fejer, R. L. Byer: Broadly tunable
mid-ir radiation source based on difference frequency mixing of high power
wavelength-tunable laser diodes in bulk periodically poled LiNbO3,Electron.
Lett. 32, 218–219 (1996) 459
85. W. Chen, G. Mouret, D. Boucher, F. K. Tittel: Mid-infrared trace gas de-
tection using continuous-wave difference frequency generation in periodically
poled RbTiOAsO4, Appl. Phys. B 72, 873–876 (2001) 459
86. D. Mazzotti, P. De Natale, G. Giusfredi, C. Fort, J. A. Mitchell, L. W. Holl-
berg: Difference-frequency generation in PPLN at 4.25 µm: an analysis of sen-
sitivity limits for DFG spectrometers, Appl. Phys. B 70, 747–750 (2000) 459
87. O. Levi, T. J. Pinguet, T. Skauli, L. A. Eyeres, K. R. Parameswaran, J. S. Har-
ris, Jr., M. M. Fejer, T.J. Kulp, S. E. Bisson, B. Gerard, E. Lallier, L. Be-
couarn: Difference frequency generation of 8-micron radiation in orientation-
patterned GaAs, Opt. Lett. 27, 2091–2093 (2002) 459
88. D. G. Lancaster, D. Richter, R. F. Curl, F. K. Tittel: Real-time measurements
of trace gases using a compact difference-frequency-based sensor operating at
3.5µm, Appl. Phys. B 67, 339–345 (1998) 459
89. D. G. Lancaster, L. Goldberg, J. Koplow, R. F. Curl, F. K. Tittel: Fibre cou-
pled difference frequency generation utilising ytterbium-doped fibre amplifier
and periodically poled LiNbO3, Electron. Lett. 34, 1345–1347 (1998) 459

Mid-Infrared Laser Applications in Spectroscopy 503
90. D. G. Lancaster, D. Richter, R. F. Curl, F. K. Tittel, I. Goldberg, J. Koplow:
High-power continuous-wave mid-infrared radiation generated by difference
frequency mixing of diode-laser-seeded fiber amplifiers and its application to
dual-beam spectroscopy, Opt. Lett. 24, 1744–1746 (1999) 459
91. D. G. Lancaster, R. Weidner, D. Richter, F. K. Tittel, J. Limpert: Compact
CH4sensor based on difference frequency mixing of diode lasers in quasi-
phasematched LiNbO3, Opt. Commun. 175, 461–468 (2000) 459
92. D. Richter, D.G. Lancaster, R. F. Curl, W. Neu, F. K. Tittel: Compact mid-
infrared trace gas sensor based on difference-frequency generation of two diode
lasers in periodically poled LiNbO3, Appl. Phys. B 67, 347–350 (1998) 459
93. D. Richter, M. Erdelyi, F. K. Tittel: Ultra-compact mid-IR spectroscopic
source based on frequency converted Yb-Er/Yb fiber amplified CW diode
lasers, in C. Marshall (Ed.): Adv. Solid-State Lasers, OSA Trends Opt. Pho-
ton. 50 (Opt. Soc. Am., Washington, DC 2001) pp. 96–100 459
94. J. J. Zondy: The effects of focusing in type-I and type-II difference-frequency
generations, Opt. Commun. 149, 181–206 (1998) 460
95. T. Topfer, K. P. Petrov, Y. Mine, D. Jundt, R. F. Curl, F. K. Tittel: Room-
temperature mid-infrared laser sensor for trace gas detection, Appl. Opt. 36,
8042–8049 (1997) 460
96. M. H. Chou, M. A. Arbore, M. M. Fejer: Adiabatically tapered periodic seg-
mentation of channel waveguides for mode-size transformation and fundamen-
tal mode excitation, Opt. Lett. 21, 794–796 (1996) 460
97. M. H. Chou, K. R. Parameswaran, M. M. Fejer, I. Brener: Multiple-channel
wavelength conversion by use of engineered quasi-phase-matching structures
in LiNbO3waveguides, Opt. Lett. 24, 1157–1159 (1999) 460
98. K.P.Petrov,A.T.Ryan,T.L.Patterson,L.Huang,S.J. Field,D.J.Bam-
ford: Mid-infrared spectroscopic detection of trace gases using guided-wave
difference-frequency generation, Appl. Phys. B 67, 357–361 (1998) 460
99.K.P.Petrov,A.P.Roth,T.L.Patterson,T.P.S.Thoms,L.Huang,
A. T. Ryan, D. J. Bamford: Efficient difference-frequency mixing of diode lasers
in lithium niobate channel waveguides, Appl. Phys. B 70, 777–782 (2000) 460
100. D. Hofmann, G. Schreiber, C. Haase, H. Herrmann, W. Grundkotter,
R. Ricken, W. Sohler: Quasi-phase-matched difference-frequency generation
in periodically poled Ti:LiNbO3channel waveguides, Opt. Lett. 24, 896–898
(1999) 460
101. P. E. Powers, T. J. Kulp, S. E. Bisson: Continuous tuning of a continuous-
wave periodically poled lithium niobate optical parametric oscillator by use
of a fan-out grating design, Opt. Lett. 23, 159–161 (1998) 461,462
102. G. M. Gibson, M. H . Dunn, M. J. Padgett: Application of a continuously tun-
able, CW optical parametric oscillator for high-resolution spectroscopy, Opt.
Lett. 23, 40–42 (1998) 461,462
103. T. J. Edwards, G. A. Turnbull, M. H. Dunn, M. Ebrahimzadeh, H. Karlsson,
G. Arvidsson, F. Laurell: Continuous-wave singly resonant optical parametric
oscillator based on periodically poled RbTiOAsO4, Opt. Lett. 23, 837–839
(1998) 461,462
104. E. V. Kovalchuk, D. Dekorsy, A. I. Lvovsky, C. Braxmaier, J. Mlynek, A. Pe-
ters, S. Schiller: High-resolution Doppler-free molecular spectroscopy with
a continuous-wave optical parametric oscillator, Opt. Lett. 26, 1430–1432
(2001) 461,462

504 Frank K. Tittel et al.
105. R. Altahtamouni, K. Bencheikh, R. Storz, K. Schneider, M. Lang, J. Mlynek,
S. Schiller: Long-term stable operation and absolute frequency stabilization
of a doubly resonant parametric oscillator, Appl. Phys. B 66, 733–739 (1998)
461,462
106. M. E. Klein, D. H. Lee, J. P. Meyn, B. Beier, K. J. Boller, R. Wallenstein:
Diode-pumped continuous-wave widely tunable optical parametric oscillator
based on periodically poled lithium tantalate, Opt. Lett. 23, 831–833 (1998)
461,462
107. G. A. Turnbull, M. H. Dunn, M. Ebrahimzadeh: Continuous-wave, intracavity
optical parametric oscillators – an analysis of power characteristics, Appl.
Phys. B 66, 701–710 (1998) 461
108. W. R. Bosenberg, A. Drobshoff, J. I. Alexander, L. E. Myers, R. L. Byer:
Continuous-wave singly resonant optical parametric oscillator based on pe-
riodically poled LiNbO3, Opt. Lett. 21, 713–715 (1996) 462
109. L. E. Myers, R. C. Eckardt, M.M. Fejer, R. L. Byer, W. R. Bosenberg: Multi-
grating quasi-phase-matched optical parametric oscillator in periodically poled
LiNbO3, Opt. Lett. 21, 591–593 (1996) 462
110. L. E. Myers, W. R. Bosenberg: Periodically p oled lithium niobate and quasi-
phase-matched optical parametric oscillators, IEEE J. Quant. Electron. 33,
1663–1672 (1997) 462
111. F. Hanson, P. Poirer, M. A. Arbore: Single-frequency mid-infrared optical
parametric oscillator source for coherent laser radar, Opt. Lett. 26, 1794–1796
(2001) 462
112. M. A. Arbore, M. M. Fejer: Singly resonant optical parametric oscillation in
periodically poled lithium niobate waveguides, Opt. Lett. 22, 151–153 (1997)
462
113. A. J. Henderson, P. M. Roper, L. A. Borschowa, R. D. Mead: Stable, continu-
ously tunable operation of a diode-pumped doubly resonant optical parametric
oscillator, Opt. Lett. 25, 1264–1266 (2000) 462
114. M. van Herpen, S. te LintelHekkert, S. E. Bisson, F. J. M. Harren: Wide single-
mode tuning of a 3.0–3.8µm, 700 mW continuous-wave Nd:YAG-pumped op-
tical parametric oscillator based on periodically poled lithium niobate, Opt.
Lett. 27, 640–642 (2002) 462
115. S. Schiller, K. Schneider, J. Mlynek: Theory of an optical parametric oscillator
with resonant pump and signal, J. Opt. Soc. Am. B 16, 1512–1524 (1999) 462
116. U. Strossner, J. P. Meyn, R. Wallenstein, P. Urenski, A. Arie, G. Rosenman,
J. Mlynek: Single-frequency continuous-wave optical parametric oscillator sys-
tem with an ultrawide tuning range of 550 to 2830 nm, J. Opt. Soc. Am. B
19, 1419–1424 (2002) 462
117. L. S. Rothman, C. P. Rinsland, A. Goldman, S. T. Massie, D. P. Edwards,
J. M. Flaud, A. Perrin, C. Camypeyret, V. Dana, J. Y. Mandin, J. Schroeder,
A. McCann, R. R. Gamache, R. B. Wattson, K. Yoshino, K. V. Chance,
K. W. Jucks, L. R. Brown, V. Nemtchinov, P. Varanasi: The HITRAN molec-
ular spectroscopic database and hawks (HITRAN atmospheric workstation)
1996 edn., J. Quant. Spectrosc. Radiat. Transfer 60, 665–710 (1998) 463,
466,472
118. N. Jacquinet-Husson, E. Arie, J. Ballard, A. Barbe, G. Bjoraker, B. Bonnet,
L. R. Brown, C. Camy-Peyret, J. P. Champion, A. Chedin, A. Chursin, C. Cler-
baux, G. Duxbury, J. M. Flaud, N. Fourrie, A. Fayt, G. Graner, R. Gamache,

Mid-Infrared Laser Applications in Spectroscopy 505
A. Goldman, V. Golovko, G. Guelachvili, J. M. Hartmann, J. C. Hilico, J. Hill-
man, G. Lefevre et al.: The 1997 spectroscopic GEISA databank, J. Quant.
Spectrosc. Radiat. Transfer 62, 205–254 (1999) 463,466
119. J. Humlicek: An efficient method of evaluation of the complex probability
function: the Voigt function and its derivatives, J. Quant. Spectrosc. Radiat.
Tran s fer 21, 309 (1979) 466
120. U. Platt: Air monitoring by differential optical absorption spectroscopy, in
Encyclopedia of Analytical Chemistry, L. E. Myers (Ed.) (Wiley, Chichester
2000) pp. 1936–1959 467
121. S. Svanberg: Atomic and Molecular Spectroscopy: Basic Aspects and Practical
Applications (Springer, Berlin, Heidelberg 2001) 467
122. G. D. Houser, E. Garmire: Balanced detection technique to measure small
changes in transmission, Appl. Opt. 33, 1059–1062 (1994) 468
123. P. C. D. Hobbs: Ultrasensitive laser measurements without tears, Appl. Opt.
36, 903–920 (1997) 469
124. M. G. Allen, K. L. Carleton, S. J. Davis, W. J. Kessler, C. E. Otis,
D. A. Palombo, D. M. Sonnenfroh: Ultra-sensitive dual b eam absorption and
gain spectroscopy: Applications for near-IR and visible diode laser sensors,
Appl. Opt. 34, 3240 (1995) 469
125. D. M. Sonnenfroh, W. T. Rawlins, M. G. Allen, C. Gmachl, F. Capasso,
A. L. Hutchinson, D. L. Sivco, J. N. Baillargeon, A. Y. Cho: Application of
balanced detection to absorption measurements of trace gases with room-
temperature, quasi-CW quantum-cascade lasers, Appl. Opt. 40, 812–820
(2001) 469
126. G. C. Bjorklund: Frequency-modulation spectroscopy: a new method of mea-
suring weak absorption and dispersions, Opt. Lett. 5, 15 (1980) 470
127. D. E. Cooper, T. F. Gallagher: Double frequency modulation spectroscopy:
high modulation frequency with low-bandwidth detectors, Appl. Opt. 24, 1327
(1985) 470
128. R. Grosskloss, P. Kersten, W. Demtroeder: Sensitive amplitude and phase-
modulated absorption spectroscopy with a continuously tunable diode laser,
Appl. Phys. B 58, 137 (1994) 470
129. D. S. Bomse, A. C. Stanton, J. A. Silver: Frequency modulation and wave-
length modulation spectroscopies: Comparison of experimental methods using
a lead-salt diode laser, Appl. Opt. 31, 718–731 (1992) 471
130. F. S. Pavone, M. Inguscio: Frequency and wavelength modulation spectroscop-
ties: comparison and experimental methods of using an AlGaAs diode laser,
Appl. Phys. B 56, 118 (1993) 471
131. D. B. Oh, D. C. Hovde: Wavelength modulation detection of acetylene with
near-IR external cavity diode laser, Appl. Opt. 34, 7002–7005 (1995) 471
132. K. Uehara: Dependence of harmonic signals on sample-gas parameters in
wavelength-modulation spectroscopy for precise absorption measurements,
Appl. Phys. B 67, 517–523 (1998) 471
133. P. Werle: Laser excess noise and interferometric effects in frequency-modulated
diode laser spectrometers, Appl. Phys. B 60, 499–506 (1995) 471
134. M. S. Zahniser, D. D. Nelson, J. B. McManus, P. Kebabian: Measurement of
trace gas fluxes using tunable diode laser spectroscopy, Phil. Trans. R. Soc.
London A 351, 371–382 (1995) 471,483

506 Frank K. Tittel et al.
135. A. Fried, J. R. Drummond, B. Henry, J. Fox: Versatile integrated tunable diode
laser system for high precision: application for ambient measurements of OCS,
Appl. Opt. 30, 1916–1932 (1991) 471
136. T. Iguchi: Modulation waveforms for second-harmonic detection with tunable
diode lasers, J. Opt. Soc. Am. B 3, 419–423 (1986) 471
137. R. G. Pilston, J. U. White: A long path gas absorption cell, J. Opt. Soc. Am.
44, 572–573 (1954) 472
138. J. U. White: Long optical paths of large aperture, J. Opt. Soc. Am. 32, 285–
288 (1942) 472
139. D. Herriot, H. Kogelnik, R. Kompfner: Off-axis paths in spherical mirror in-
terferometers, Appl. Opt. 3, 523 (1964) 472
140. S. M. Chernin, E. G. Barskaya: Optical multi-pass matrix systems, Appl. Opt.
30, 51–58 (1991) 472
141. J. B. McManus, P. L. Kebabian, M. S. Zahniser: Astigmatic mirror multi-pass
absorption cells for long-path length spectroscopy, Appl. Opt. 34, 3336 (1995)
472
142. A. O’Keefe, D. A. G. Deacon: Cavity ring-down optical spectrometer for ab-
sorption measurements using pulsed laser sources, Rev. Sci. Instrum. 59, 2544
(1988) 473,475
143. J. Lauterbach, D. Kleine, K. Kleinermanns, P. Hering: Cavity-ring-down spec-
troscopic studies of NO2in the region around 613 nm, Appl. Phys. B 71,
873–876 (2000) 474
144. D. Kleine, S. Stry, J. Lauterbach, K. Kleinermanns, P. Hering: Measurement
of the absolute intensity of the fifth CH stretching overtone of benzene using
cavity ring-down spectroscopy, Chem. Phys. Lett. 312, 185–190 (1999) 474
145. G. Berden, R. Peeters, G. Meijer: Cavity ring-down spectroscopy: Experimen-
tal schemes and applications, Int. Rev. Phys. Chem. 19, 565–607 (2000) 475
146. R. Peeters, G. Berden, A. Apituley, G. Meijer: Open-path trace gas detection
of ammonia based on cavity-enhanced absorption spectroscopy, Appl. Phys.
B71, 231–236 (2000) 475
147. R. Peeters, G. Berden, G. Meijer: Sensitive absorption techniques for spec-
troscopy, Am. Lab. 33, 60 (2001) 475
148. Y. B. He, B. J. Orr: Rapidly swept, continuous-wave cavity ringdown spec-
troscopy with optical heterodyne detection, Single- and multi-wavelength sens-
ing of gases, Appl. Phys. B 75, 267–280 (2002) 475
149. K. K. Lehmann, D. Romanini: The superposition principle and cavity ring-
downspectroscopy,J.Chem.Phys.105, 10263–10277 (1996) 475
150. B. A. Paldus, J. S. Harris, J. Martin, J. Xie, R. N. Zare: Laser diode cavity
ring-down spectroscopy using acousto-optic modulator stabilization, J. Appl.
Phys. 82, 3199–3204 (1997) 475
151. B.A. Paldus, C. C. Harb, T. G. Spence, B. Wilke, J. Xie, J. S. Harris,
R. N. Zare: Cavity-locked ring-down spectroscopy, J. Appl. Phys. 83, 3991–
3997 (1998) 475
152. J. J. Scherer, J. B. Paul, A. O’Keefe, R. J. Saykally: Cavity ringdown laser
absorption spectroscopy, Chem. Rev. 97, 25–51 (1997) 475
153. R. A. Provencal, J. B. Paul, E. Michael, R. J. Saykally: Absorption spec-
troscopy technique provides extremely high sensitivity, Photon. Spectra 32,
159 ff. (1998) 475
154. J. B. Paul, R. J. Saykally: Cavity ring-down laser absorption spectroscopy,
Anal. Chem. 69, A287–A292 (1997) 475

Mid-Infrared Laser Applications in Spectroscopy 507
155. K. W. Busch, M. A. Busch: Cavity Ring-Down Spectroscopy, An Ultratrace-
Absorption Measurement Technique (American Chemical Society, Washing-
ton, DC 1999) 475
156. J. J. Scherer, D. Voelkel, D. J. Rakestraw, J. B. Paul, C. P. Collier, A. O’Keefe,
R. J. Saykally: Infrared Cavity Ringdown Laser Absorption Spectroscopy (IR-
CRLAS), Chem. Phys. Lett. 245, 273–280 (1995) 475,476
157. K. K. Lehmann: Ring-down cavity spectroscopy cell using continuous wave
excitation for trace species detection, US patent no. 5,528,040 (1996) 476
158. D. Romanini, A. A. Kachanov, N. Sadeghi, F. Stoeckel: CW cavity ring down
spectroscopy, Chem. Phys. Lett. 264, 316–322 (1997) 476
159. D. Romanini, A. A. Kachanov, F. Stoeckel: Diode laser cavity ring down spec-
troscopy, Chem. Phys. Lett. 270, 538–545 (1997) 476
160. 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, A. Y. Cho: Cavity ring-down
spectroscopy using mid-infrared quantum-cascade lasers, Opt. Lett. 25, 666–
668 (2000) 476,487
161. A.A. Kosterev, A. L. Malinovsky, F.K. Tittel, C. Gmachl, F. Capasso,
D. L. Sivco, J. N. Baillargeon, A. L. Hutchinson, A. Y. Cho: Cavity ring-
down spectroscopic detection of nitric oxide with a continuous-wave quantum-
cascade laser, Appl. Opt. 40, 5522–5529 (2001) 476,486,487
162. K. W. Aniolek, P. E. Powers, T. J. Kulp, B. A. Richman, S. E. Bisson: Cavity
ring-down laser absorption spectroscopy with a 1 kHz mid-infrared periodi-
cally poled lithium niobate optical parametric generator optical parametric
amplifier, Chem. Phys. Lett. 302, 555–562 (1999) 476
163. R. J. Saykally, R. Casaes: Cavity ring-down technique measures absorption,
Laser Focus World 37, 159–162 (2001) 476
164. A. O’Keefe: Integrated cavity output analysis of ultra-weak absorption, Chem.
Phys. Lett. 293, 331–336 (1998) 476
165. R. Engeln, G. Vonhelden, G. Berden, G. Meijer: Phase shift cavity ring down
absorption spectroscopy, Chem. Phys. Lett. 262, 105–109 (1996) 476
166. R. Engeln, G. Berden, R. Peeters, G. Meijer: Cavity enhanced absorption
and cavity enhanced magnetic rotation spectroscopy, Rev. Sci. Instrum. 69,
3763–3769 (1998) 476
167. J. B. Paul, L. Lapson, J. G. Anderson: Ultrasensitive absorption spectroscopy
with a high-finesse optical cavity and off-axis alignment, Appl. Opt. 40, 4904–
4910 (2001) 476,477,486
168. M. Murtz, B. Frech, W. Urban: High-resolution cavity leak-out absorption
spectroscopy in the 10-µm region, Appl. Phys. B 68, 243–249 (1999) 476
169. L. Menzel, A. A. Kosterev, R. F. Curl, F. K. Tittel, C. Gmachl, F. Capasso,
D. L. Sivco, J. N. Baillargeon, A. L. Hutchinson, A. Y. Cho, W. Urban: Spec-
troscopic detection of biological NO with a quantum cascade laser, Appl. Phys.
B72, 859–863 (2001) 476,486
170. D. S. Baer, J. B. Paul, M. Gupta, A. O’Keefe: Sensitive absorption measure-
ments in the near-infrared region using off-axis integrated cavity output spec-
troscopy, Appl. Phys. B 75, 261–265 (2002) 477
171. L. S. Ma, J. Ye, P. Dube, J. L. Hall: Ultrasensitive frequency-modulation spec-
troscopy enhanced by a high-finesse optical cavity: Theory and application to
overtone transitions of C2H2and C2HD,J.Opt.Soc.Am.B16, 2255–2268
(1999) 477

508 Frank K. Tittel et al.
172. J. Ye, L. S. Ma, J. L. Hall: Sub-doppler optical frequency reference at 1.064-µm
by means of ultrasensitive cavity-enhanced frequency modulation spectroscopy
of a C2HD overtone transition, Opt. Lett. 21, 1000–1002 (1996) 477
173. M. S. Taubman, R. M. Williams, T. L. Myers, F. Capasso, C. Gmachl,
D. L. Sivco, A. L. Hutchinson, A. Y. Cho: Sub-Doppler NICE-OHMS spec-
troscopy at 8.5 microns using a quantum cascade laser, in OSA Trends in Op-
tics and Photonics (TOPS) Vol 73, Conference on Lasers and Electro-Optics,
OSA Technical Digest (Opt. Soc. of Am., Washington, DC., 2002) pp.625–626
478
174. M. A. Moeckli, C. Hilbes, M. W. Sigrist: Photoacoustic multicomponent gas
analysis using a Levenberg–Marquardt fitting algorithm, Appl. Phys. B 67,
449–458 (1998) 478
175. J. Oomens, H. Zuckermann, S. Persijn, D. H. Parker, F. J. M. Harren: CO-
laser-based photoacoustic trace-gas detection – applications in postharvest
physiology, Appl. Phys. B 67, 459–466 (1998) 478
176. M. A. Moeckli, M. Fierz, M. W. Sigrist: Emission factors for ethene and am-
monia from a tunnel study with a photoacoustic trace gas detection system,
Environ. Sci. Technol. 30, 2864–2867 (1996) 478
177. M. Nagele, M. W. Sigrist: Mobile laser spectrometer with novel resonant multi-
pass photoacoustic cell for trace-gas sensing, Appl. Phys. B 70, 895–901 (2000)
478
178. B. G. Ageev, Y. N. Ponomarev, V. A. Sapozhnikova: Laser photoacoustic spec-
troscopy of biosystems gas exchange with the atmosphere, Appl. Phys. B 67,
467–473 (1998) 478
179. F. J. M. Harren, G. Cotti, J. Oomens, S. L. Hekkert: Photoacoustic spec-
troscopy in trace gas monitoring, in Encyclopedia of Analytical Chemistry,
R. A. Meyers (Ed.) (Wiley, Chichester 2000) pp. 2203–2226 478
180. A. Bohren, M. W. Sigrist: Optical parametric oscillator based difference fre-
quency laser source for photoacoustic trace gas spectroscopy in the 3-µmmid-
IR range, Infrared Phys. Technol. 38, 423–435 (1997) 478
181. Z. Bozoki, J. Sneider, Z. Gingl, A. Mohacsi, M. Szakall, Z. Bor, G. Szabo: A
high-sensitivity, near-infrared tunable-diode-laser-based photoacoustic water-
vapour-detection system for automated operation, Measurement Sci. Technol.
10, 999–1003 (1999) 478
182. F. G. C. Bijnen, F. J. M. Harren, J. H. P. Hackstein, J. Reuss: Intracavity CO
laser photoacoustic trace gas detection – cyclic CH4,H
2OandCO
2emission
by cockroaches and scarab beetles, Appl. Opt. 35, 5357–5368 (1996) 478
183. V. A. Kapitanov, Y. U. N. Ponomarev, K. Song, H. K. Cha, J. Lee: Resonance
photoacoustic spectroscopy and gas analysis of gaseous flow at reduced pres-
sure, Appl. Phys. B 73, 745–750 (2001) 478
184. M. B. Pushkarsky, M.E. Webber, O. Baghdassarian, L. R. Narasimhan,
C. K. N. Patel: Laser-based photoacoustic ammonia sensors for industrial ap-
plications, Appl. Phys. B 75, 391–396 (2002) 479
185. A. Schmohl, A. Miklos, P. Hess: Detection of ammonia by photoacoustic spec-
troscopy with semiconductor lasers, Appl. Opt. 41, 1815–1823 (2002) 479
186. A. A. Kosterev, Y. A. Bakhirkin, R. F.Curl, F. K. Tittel: Quartz-enhanced pho-
toacoustic spectroscopy, Opt. Lett. 27, 1902–1904 (2002) 479
187. A. Fried, W. W. Berg: Photoacoustic detection of HCl, Opt. Lett. 8, 160–162
(1983) 479

Mid-Infrared Laser Applications in Spectroscopy 509
188. M. Avery-Owens, C. C. Davis, R. R. Dickerson: A photothermal interferometer
for gas-phase ammonia detection, Anal. Chem. 71, 1391–1399 (1999) 479
189. B. A. Paldus, T. G. Spence, R. N. Zare, J. Oomens, F. J. M. Harren,
D. H. Parker, C. Gmachl, F. Cappasso, D. L. Sivco, J. N. Baillargeon,
A. L. Hutchinson, A. Y. Cho: Photoacoustic spectroscopy using quantum-
cascade lasers, Opt. Lett. 24, 178–180 (1999) 480
190. D. Hofstetter, M. Beck, J. Faist, M. Nagele, M. W. Sigrist: Photoacoustic
spectroscopy with quantum cascade distributed-feedback lasers, Opt. Lett.
26, 887–889 (2001) 480
191. P. Werle, R. Mucke, F. Slemr: The limits of signal averaging in atmospheric
trace-gas monitoring by tunable diode-laser absorption spectroscopy (TD-
LAS), Appl. Phys. B 57, 131–139 (1993) 483,484
192. S. Sewell, A. Fried, B. Henry, J. R. Drummond: A field diode laser spec-
trometer employing an astigmatic Herriott cell, in Tunable Diode Laser Spec-
troscopy, Lidar, and DIAL Techniques for Environmental and Industrial Mea-
surements, A. Fried, D. K. Killinger, H. I. Schiff (Eds.) (SPIE, Atlanta, GA
1993) pp. 72–80 484
193. A. Fried, B. P. Wert, J. G. Walega, D. Richter, W.T. Potter: Airborne mea-
surements of formaldehyde employing a high-performance tunable diode laser
absorptoion system, in Diode Lasers and Applications in Atmospheric Sensing,
A. Fried (Ed.), Proc. SPIE 4817, 177–183 (2002) 484
194. R. W. Waynant, I. K. Ilev, I. Gannot: Mid-infrared laser applications in
medicine and biology, Phil. Trans. R. Soc. London A 635–644 (2001) 486
195. P. Murtz, L. Menzel, W. Bloch, A. Hess, O. Michel, W. Urban: LMR spec-
troscopy: a new sensitive method for on-line recording of nitric oxide in breath,
J. Appl. Physiol. 86, 1075–1080 (1999) 486
196. H. Dahnke, D. Kleine, P. Hering, M. Murtz: Real-time monitoring of ethane
in human breath using mid-infrared cavity leak-out spectroscopy, Appl. Phys.
B72, 971–975 (2001) 486
197. D. E. Cooper, R. U. Martinelli, C. B. Carlisle, H. Riris, D. B. Bour,
R. J. Menna: Measurements of 12 CO2:13 CO2ratios for medical diagnostics
with 1.6-micron distributed-feedback semiconductor diode lasers, Appl. Opt.
32, 6727–6731 (1993) 486
198. L. R. Narasimhan, W. Goodman, C. K. N. Patel: Correlation of breath am-
monia with blood urea nitrogen and creatinine during hemodialysis, Proc.
National Acad. Sci. United States Am. 98, 4617–4621 (2001) 486
199. C. Roller, K. Namjou, J. D. Jeffers, M. Camp. A. Mock, P. J. McCann,
J. Grego: Nitric oxide breath testing by tunable-diode laser absorption spec-
troscopy, Application in monitoring respiratory inflammation, Appl. Opt. 41,
6018–6029 (2002) 486
200. I. Tsujino, M. Nishimura, A. Kamachi, H. Makita, M. Munakata,
K. Miyamoto, Y. Kawakami: Exhaled nitric oxide – Is it really a good marker
of airway inflammation in bronchial asthma?, Respiration 67, 645–651 (2000)
486
201. T. Ingi, J. Cheng, G. V. Ronnett: Carbon monoxide – an endogenous mod-
ulator of the nitric oxide-cyclic GMP signaling system, Neuron 16, 835–842
(1996) 489
202. Y. Morimoto, W. Durante, D. G. Lancaster, J. Klattenhoff, F. K. Tittel: Real-
time measurements of endogenous CO production from vascular cells using an
ultrasensitive laser sensor, Am. J. Physiol. 280, H483–H488 (2001) 489,490

510 Frank K. Tittel et al.
203. A. A. Kosterev, F. K. Tittel, W. Durante, M. Allen, R. K¨ohler, C. Gmachl,
F. Capasso, D. L. Sivco, A. Y. Cho: Detection of biogenic CO production above
vascular cell cultures using a near-room temperature QC-DFB laser, Appl.
Phys. B 74, 95–99 (2002) 489
204. A. A. Kosterev, F. K. Tittel, R. K¨ohler, C. Gmachl, F. Capasso, D. L. Sivco,
A. Y. Cho, S. Wehe, M. G. Allen: Thermoelectrically cooled quantum-cascade-
laser-based sensor for the continuous monitoring of ambient atmospheric car-
bon monoxide, Appl. Opt. 41, 1169–1173 (2002) 489
205. A. A. Kosterev, R. F. Curl, F. K. Tittel, R. K¨ohler, C. Gmachl, F. Capasso,
D. L. Sivco, A. Y. Cho: Transported automated ammonia sensor based on
a pulsed thermoelectrically cooled QC-DFB laser, Appl. Opt. 41, 573–578
(2002) 490
206. A. A.Kosterev, F. K. Tittel: Chemical sensorsbased on quantum cascade lasers,
IEEE Quant. Electron. 38, 582–591 (2002) 490
207. D. Richter: Portable mid-infrared gas sensors: development and applications,
Dissertation, Rice University, Houston (2001) 491
208. D. Richter, A. Fried, B. P. Wert, J. G. Walega, F. K. Tittel: Development of
a tunable mid-IR difference-frequency laser source for highly sensitive airborne
trace gas detection, Appl. Phys. B 75, 281–288 (2002) 491,493
209. D. Richter, M. Erdelyi, R. F. Curl, F. K. Tittel, C. Oppenheimer, H. J. Duffell,
M. Burton: Field measurements of volcanic gases using tunable diode laser
based mid-infrared and Fourier transform infrared spectrometers, Opt. Lasers
Engin. 37, 171–186 (2002) 492
210. D. Richter, F. K. Tittel, J. A. Bahr, D. T. Wickham, J. D. Wright, J. C. Graf:
Diode laser based formaldehyde measurements in a catalytic trace contami-
nant control system, paper 2302, Proc. 30th Int. Conf. Environmental Systems
(ICES), Toulouse, France (2000) 492
211. D. Rehle, D. Leleux, M. Erdelyi, F. Tittel, M. Fraser, S. Friedfeld: Ambi-
ent formaldehyde detection with a laser spectrometer based on difference-
frequency generation in PPLN, Appl. Phys. B 72, 947–952 (2001) 492
212. D. Richter, A. Fried, G.S. Tyndall, E. Oteiza, M. Erdelyi, F. K. Tittel: Tunable
telecom diode laser based mid-IR source at 2.64 microns, OSA Trends Optics
Photonics, Advanced Solid State Lasers 68, 351–353 (2002) 494
213. P. Bergamaschi, M. Schupp, G. W. Harris: High-precision direct measurements
of 13CH4/12 CH4and 12CH3D/12 CH4ratios in atmospheric methane sources
by means of a long-path tunable diode laser absorption spectrometer, Appl.
Opt. 33, 7704–7716 (1994) 495,496
214. M. Erdelyi, D. Richter, F. K. Tittel: Measurement of 13 CO2/12 CO2isotopic
ratio using a difference-frequency-based sensor operating at 4.35 µm, Appl.
Phys. B 75, 289–295 (2002) 495
215. J. B. McManus, M. S. Zahniser, D. D. Nelson, L. R. Williams, C. E. Kolb: In-
frared laser spectrometer with balanced absorption for measurement of iso-
topic ratios of carbon isotopes, Spectrochim. Acta A 58, 2465–2479 (2002)
496
216. E. R. T. Kerstel, R. v. Trigt, N. Dam, J. Reuss, H. A. J. Meijer: Simultaneous
determination of the H-2/H-1, O-17/O-16, and O-18/O-16 isotope abundance
ratios in water by means of laser spectrometry, Anal. Chem. 71, 5297–5303
(1999) 496

Mid-Infrared Laser Applications in Spectroscopy 511
217. K. Uehara, K. Yamamoto, T. Kikugawa, N. Yoshida: Isotope analysis of envi-
ronmental substances by a new laser-spectroscopic method utilizing different
pathlengths, Sensors Actuators B 74, 173–178 (2001) 496
218. E.R. Crosson, K. N. Ricci, B. A. Richman, F. C. Chilese, T. G. Owano,
R. A. Provencal, M.W. Todd, J. Glasser, Alex A. Kachanov, B.A. Paldus,
T. G. Spence, R. N. Zare: Stable isotope ratios using cavity ring-down spec-
troscopy: determination of 13 C/12C for carbon dioxide in human breath,
Anal. Chem. 74, 2003–2007 (2002) 496
219. L. Hvozdara, N. Pennington, M. Kraft, M. Karlowatz, B. Mizaikoff: Quantum
cascade lasers for mid-infrared spectroscopy, Vibr. Spectrosc. 30, 53–58 (2002)
496
220. D. Richter, A. Fried, F. K. Tittel: Special Issue: Trends in laser sources, spec-
troscopic techniques and their applications to trace gas detection, Appl. Phys.
B75, 143–403 (2002) 496
221. R. F. Curl, F. K. Tittel: Tunable Infrared Laser Spectroscopy, Ann. Rep. Progr.
Chem. C 98, 217–270 (2002) 496

Index
H2O2,481
12CO2,494
13CO2,494
in situ measurement, 467
absorption
– combination, 451
– line, 450,451,462,463,487,488,
490,494,495
– linewidth, 448
–overtone,451
–spectroscopy,462,463,471,473,474,
477,483,489,494
–spectrum,471,488
acousto-optic modulator (AOM), 476,
487
air-broadening coefficient, 466
airborne tunable diode laser system,
486
ambient air, 453,473,484,490
ambient temperature, 492,496
ammonia
– concentration, 480
–mixture,476
–sensor,490
amplitude modulation, 445,470
amplitude-modulated phase-
modulation spectroscopy (AMPM),
470
analog–digital converter, 488
angle tuning, 457
antimonide diode laser, 450,451
astigmatic beam profile, 451,453,458
astigmatic mirror Herriott cell, 492
astigmatic mirror multi-pass cell, 472
balanced beam, 467
balanced ratiometric detection (BRD),
467,469
bandwidth, 459,460,468,480,491,492
base-emitter curve, 469
beam bounce pattern, 472
beam divergence, 445,450,451
beam pointing stability, 458,461,473
beam profile, 451,453,455,458,461,
462
beamsplitter/combiner (BS), 455
Beer–Lambert absorption law, 463,
464,471,472,478
birefringent nonlinear crystal, 457
Boltzman distribution, 495
Bragg grating, 452
Bragg reflector, 449,493
Brewster angle window, 495
broadly tunable monolithic integrated
multi-section diode laser chips, 456
C2H2,477,481
C2H6,481
cavity eigenmode, 476
cavity finesse, 476
cavity ring-down spectroscopy (CRDS),
475,476
cavity transmission spectrum, 486
cavity-enhanced absorption spec-
troscopy (CEAS), 476,477,486
CH2O, 450,465,480,481,483–486,
492–494
CH4,451,477,480
Chernin multi-pass cell, 472
chromium-doped zinc selenide
chalcogenide, 454
CO (carbon monoxide), 447,451,477,
479–481,489,490
CO2,447,477,479–481,488,495

Index 513
CO2laser, 496
collisional de-excitation, 478
combination-overtone ro-vibrational
line, 462
combined broadening, 464
conduction band, 447
continuous wave, 448,450,462
continuous wavelength tunability, 452
cryogenic cooling, 447,449,451,453
CS2,481
CW dye laser, 470
dcc-coupled Peltier-cooled HgCdTe
(MCT), 492
dielectric plasmon mode, 452
difference frequency generation (DFG),
455,457,461,490,492
– beam profile, 491
differential optical absorption spec-
troscopy (DOAS), 466
direct diode-pumping, 455
distributed Bragg reflector (DBR), 449,
456
distributed feedback diode laser (DFB),
458
divergent beam profile, 451,453
dopant gain region, 457
doped solid-state bulk material, 446
Doppler cross-section, 464
double refraction, 457
dual chamber absorption cell, 495
dual channel absorption laser, 482
dual-beam detection, 468
dual-beam long-path absorption
spectroscopy, 492
ellipsoidal off-axis mirror (OAE), 454
elliptical beam profile, 451,453,458
equivalent noise bandwidth (ENBW),
468
Er/Yb fiber amplifier, 494
erbium fiber amplifier, 455
external cavity diode laser (ECDL),
456,458,496
Fab r y–P´erot laser, 448,451,452,456,
477
ferroelectric crystal, 459,461
fiber amplifier, 479,491,494
fiber based pump source, 490
fiber optic amplifier, 491
FM spectrometer, 471
FM spectroscopy, 471
formaldehyde, 450
free spectral range (FSR), 475,477
frequency locking technique, 457
frequency modulation spectroscopy
(FMS), 470,471
FWHM, 448
GaInSbAs quantum well, 451
GaInSbAs/GaSb, 451
Gaussian, 464–466
Gaussian beam quality, 458
Gaussian fit, 491
GEISA databank, 466
guided-wave parametric process, 460
heme oxygenase, 489
Herriott multi-pass cell, 472,477,482,
484
high inherent mid-IR wavelength
stability, 492
high sensitivity, 490
high-power continuous-wave DFG
source, 493
HITRAN database, 463,466,472,495
homonuclear diatomic molecule, 463
HWHM, 452,453,465
hydride vapor phase epitaxy (HVPE),
459
hydrogen fluoride (HF), 494
hydrogen-bonded cluster, 475
idler, 457,460,461
indium-antimonide photovoltaic
detector, 482
integrated cavity output spectroscopy
(ICOS), 476,477,486
intersubband transition, 451
Jamin interferometer, 479
laser
–beam,457,473,478–480,487,489
–CO
2,496
–ECDL,456,458,496
– Fabry–P´erot, 448,451,452,456,477

514 Index
– quantum cascade (QCL), 447,451,
453,475,481,489
– radiation filtering, 476
–source,445–447,453,455–457,459,
475,476,478–480,491,492,494,496
lasing threshold, 487
lead-salt diode laser, 447,449,450,481,
492,496
Libbrecht design, 452
light detection and ranging (LIDAR),
466
LiNbO3,460,492
– periodically poled (PPLN), 455,459,
461,462
linear high-finesse optical cavity, 488
lineshape, 464,465
liquid nitrogen
–dewar,448,450,481
– optical cryostat, 487
LIRF, 468
Lissajous pattern, 482
LiTaO3,459,460
lock-in amplifier, 468,470,471,483
logarithmic conformance, 469
long optical path length spectroscopy,
472
Lorentzian, 466
low-loss single mode fiber, 496
MEMS, 456
methanol, 480
mid-IR detector, 469
mid-IR spectroscopic application, 480
minimal detectable CH2Oconcentra-
tion, 493
minimum detectable concentration, 467
minimum detectable fractional
absorption, 468
minimum relative detectable concentra-
tion, 468
molecular beam, 475
molecular beam epitaxy (MBE), 450,
459
molecular collision, 464,465
molecular rotational-vibrational
(ro-vibrational) state, 462
molecule, 462,463,465,468,471,472,
478,481,486
monochromatic emission spectrum, 470
monolithic non-planar ring oscillator
(NPRO), 457
MOSFET, 488
multi-pass cell, 472,473,477,486,489,
492,496
multiple harmonics, 470
multiple linear regression approach, 484
multiple quantum-well heterostructure,
451
multiple spatial mode, 460
N2,463,476,488
narrow-linewidth diode laser source,
491
near-Gaussian intensity distribution,
492
NH3,451,477,479,481,490
NO, 476,481,486–488
NO-cGMP, 489
noise reduction, 466,467,469–471,477,
491
noise-immune cavity-enhanced op-
tical heterodyne spectroscopy
(NICE-OHMS), 477,478
non-planar ring oscillator (NPRO), 457
non-resonant probe laser beam, 479
off-axis cavity alignment, 477
off-axis mirror, 482
optical absorption spectrum, 462
optical fiber, 446
optical isolator, 493
optical parametric oscillation (OPO),
462
optical parametric oscillator (OPO),
455,457,462,476,479
optical pathlength, 467,472,476,482,
489,490
overtone ro-vibrational line, 462
p–n junction, 447
parabolic off-axis mirror (OAP), 454
parametric conversion source, 475
parametric frequency conversion, 446,
447,455,456
Pb-salt diode laser source, 448–450,481
PbEuSeTe, 449
PbSnTe, 449
Pd/Al2O3,484

Index 515
Peltier cooling, 453,480
PFBHA, 493
phase matching
– condition, 457
photoacoustic spectroscopy (PAS), 467,
478–480
photothermal absorption spectroscopy,
479
photothermal spectroscopy, 478
platelet aggregation, 489
polarization
– rotator, 468
polyatomic molecule, 463
portable gas sensor, 490
potassium titanyl phosphate
(KTiOPO4,orKTP),459
ppbv, 463,476,479,480,484,490,492
PPLN, 455,459,461,462,491
ppmv, 463,467,480,490
pptv, 450,463,467,472,486,493
probe beam, 468,472,479
QPM-PPLN, 490,492,494
quantum cascade laser (QCL), 447,
451,453,475,481,489
quantum well, 449,451,452
quasi-phase-matched material (QPM),
447
quasi-phase-matching (QPM), 459,461
– property, 459
radioisotope counting technique, 489
rapid background subtraction, 483
rapid sweep integration, 471
reference detector, 469
reflecting objective (RO), 454
refractive index
– modulation, 452
resonant excitation beam, 479
ring-down decay, 476
ro-vibrational line, 462
sampled grating distributed bragg
reflector diode laser (SG-DBR-DL),
458
selectivity, 445,465,467,471,479,480,
495
sensitivity, 445,467–469,471,476,477,
479,480,486,490,492,496
short-path absorption cell, 469
signal enhancement, 466,467,472–474,
477–480
single beam interferometer, 471
single-pass difference-frequency
generation, 490
spectroscopic gas sensor, 476,486
spectroscopy
– cavity ring-down (CRDS), 475,476
–FM,471
– photoacoustic (PAS), 467,478–480
stoichiometry, 448
sub-ppb detection sensitivity, 494
surface-plasmon mode, 452
TEM00,455,458,460,462,488
thermal motion, 464,465
toroidal off-axis mirror (OAT), 454
trace gas detection, 445,447,449,462,
467,478,479,490,496
triggering circuit (TC), 488
tunable optical parametric oscillator,
461
tunable solid-state laser characteristics,
455
tunable solid-state laser source, 447,
486
tuning
–´etalon, 455
two-tone frequency-modulation
(TTFM), 470
ultra-high reflective spherical mirror,
473
unipolar semiconductor injection laser,
451
variable-ratio beamsplitter, 468
vascular smooth muscle cells (VSMCs),
489
vertical cavity surface emitting laser
(VCSEL), 456,494
vibronically broadened transition, 454
Voi gt pro file , 465,466
waveguide phase matching, 460
wavelength division multiplexer
(WDM), 493
wavelength modulation, 481

516 Index
–spectroscopy(WMS),470,471,493
wavelength tunability, 445,456,462
wavenumber time domain, 475
White multi-pass cell, 472
YAG material, 457
zero-background subtraction detection,
467