Laser diode photoacoustic detection in the infrared and near infrared spectral ranges

Laser diode photoacoustic detection in the infrared and near infraredspectral ranges

V. Horka,a S. Civis,*a Li-Hong Xub and R. M. Leesb

Received 15th March 2005, Accepted 17th May 2005

First published as an Advance Article on the web 10th June 2005

DOI: 10.1039/b503838c

A new technique for high resolution photoacoustic detection based on application of laser diodes

has been developed. This method was tested and compared using identical photoacoustic

instrumentation (cell and microphone) to study gas absorption in three different spectral regions,

namely: the infrared range near 2100 cm21, CO and OCS fundamental band absorption; the

ranges near 4200 and 4350 cm21, CH4, NH3 and N2O overtone and combination band

absorption; the near infrared range near 6500 cm21, CO, CO2 and NH3 overtone absorption.

Several types of diode laser operating at room temperature or at liquid nitrogen temperature were

compared. The optimum gas pressures for the maximum sensitivity of the photoacoustic signals

were found and the detection limits were estimated for all of the gases studied. The best sensitivity

was achieved for NH3 at 100 ppbv. The sensitivity of the developed system was tested on detection

of traces of NH3 and CO2 gases from car exhaust.

1. Introduction

The principles of photoacoustic detection (PA) have been

known for more than a century.1 The greatest renaissance of

this technique occurred in the 1970’s and 1980’s at the time of

the development of laser technology. Photoacoustics have long

been used in combination with high-power lasers (mostly CO

and CO2).2,3 However, a disadvantage of these lasers is that

their wavelength cannot be tuned over a broad range, so that it

is necessary that the laser frequency be coincident with an

absorption line of the studied gas. This is not a restriction with

diode lasers, which can be easily tuned.

Technological development in the near infrared has been

of great importance in enabling application of photoacoustic

detection with semiconductor laser sources having sufficient

power to excite a photoacoustic signal.4,5 Vansteenkiste et al.6

carried out the first applications of diode lasers in the infrared

region in combination with photoacoustic detection in 1981.

This work employed a PbS12xSex laser working in the 4.8 mm

region to measure CO molecules, and achieved a detection

limit of 50 ppmv.

Photoacoustic detection has been successfully employed in a

wide range of scientific fields, e.g. in research related to the

environment,7,8 medicine9 and biology.10 Kreuzer11 first

reported the detection of very low concentrations of gases

using photoacoustic detection in combination with a non-

resonating cell. Subsequently, many scientific teams have

demonstrated that a photoacoustic system can be used to

monitor trace amounts of molecules in the atmosphere.12 The

infrared spectral range is exploited very extensively for

such work as most atmospheric pollutants absorb in this

region.13–15 The use of laser sources emitting in the infrared

permits monitoring of absorption in specific rotation–

vibration bands of the studied molecules. The development

of the new technologies of multi-quantum-well (MQW) GaSb

or InAs based laser diodes (LDs) operating in the 2.3 mm

region provides a very attractive source of radiation with high

spectral purity that can be operated and tuned at or above

room temperatures.16

Traditional direct absorption methods are based on

measuring the difference in intensity between the incident

and transmitted radiation. However, in monitoring small

concentrations of absorbing molecules, the differences detected

between the entering and leaving intensities of the radiation

are generally small. Thus, when a laser source of high power is

used, the small changes caused by the molecular absorption are

observed on a relatively large detected background signal. This

approach is disadvantageous from the standpoints both of the

signal-to-noise ratio and the dynamic detection range. In these

cases, the minimum detectable absorbance is given by the

minimum detectable signal and an increase in sensitivity can be

achieved only by lengthening the optical path. In general, it is

more effective to detect small signals in the absence of a large

background. One of the ways of dealing with this aspect is to

abandon the classical scheme of evaluating the difference

between the entering and leaving signals and to measure

the absorbed energy by some other method. Photoacoustic

detection is such a technique, and is well suited to the

monitoring of low concentrations of gases because of its very

low background offset.

The present work was motivated by the increasing need for a

cheap and simple technique to detect gases that is sufficiently

sensitive and selective. Our paper describes the simple con-

struction of a resonant optoacoustic cell in which inexpensive

microphones (Knowles EK-3024) were employed. We report

aJ. Heyrovsky Institute of Physical Chemistry, Academy of Science ofthe Czech Republic, Dolejskova 3, 18223 Prague 8, Czech Republic.E-mail: [email protected]; Fax: +420 286591766;Tel: +420 266053275bCanadian Institute for Photonic Innovations (CIPI), Centre for Laser,Atomic, and Molecular Sciences (CLAMS), and Department ofPhysical Sciences, University of New Brunswick, Saint John, NB,Canada E2L4L5

PAPER www.rsc.org/analyst | The Analyst

1148 | Analyst, 2005, 130, 1148–1154 This journal is � The Royal Society of Chemistry 2005

tests of the sensitivity of the system on selected molecules

absorbing in the 2100 cm21, 4300 cm21 and 6400 cm21 regions

and give a comparison of the results. A lead chalcogenide laser,

requiring cooling by liquid nitrogen, was selected for the

2100 cm21 region, while MQW GaSb based LDs that can be

operated at or above room temperature were employed in the

4300 cm21 region.17,18 As well, measurements were carried out

on gas molecules in the near-IR region of their vibrational

overtones (6300–6500 cm21) where powerful semiconductor

lasers working at room temperature have been developed.

These lasers are advantageous from the standpoints of their

wide tunability and relatively high power. The power of these

diodes is about 10–15 mW, while that of PbSe and GaSb

lasers is about 1 mW or less. However, a disadvantage of

this spectral region lies in the fact that the combination

bands and overtones are much weaker than the fundamental

bands. In investigating the application of diode laser photo-

acoustic detection for the three spectral regions in this

work, our aim was to study the overall performance of our

system and to compare photoacoustic measurements in the

different regions for the transitions of identical selected

molecules.

2. Theoretical background of photoacoustic detection

The fundamental principles of photoacoustic detection have

been described previously in the literature,3,19 hence only a

brief description of this method will be given here. The first

step in generation of the photoacoustic signal is absorption

of energy from a modulated radiation beam by the molecule.

The molecule then has several ways of losing this energy.

At atmospheric pressure, it is probable that the absorbed

energy will be transformed to kinetic energy following

collision with another species, i.e. converted into the energy

of thermal motion. The input of thermal energy, or heat, then

leads to a change in the gas pressure which can be detected

by a microphone. If the sample contains only one kind of

absorbing molecule, the signal S(l) from the microphone is

given by:

S(l) 5 CP(l)Ns(l) (1)

where C is a constant characterising the properties of the

acoustic cell, P(l) is the power of the laser radiation, N is the

number of absorbing molecules and s(l) is the absorption

cross-section. For multicomponent samples, the photoacoustic

signal can be measured at a set of wavelengths, li (i 5 1,...,m),

chosen on the basis of the absorption spectra of the individual

components. Eqn. (1) is then extended for an n-component

mixture to write signal S(li) as the sum of all the contributions

from the individual absorbing components j each having

concentration cj and absorption cross section sij at wave-

length li:

S(li)~Si~CPiNtot

Xn

j0~1

cjsij (2)

where Pi denotes the laser power at lI and Ntot is total number

density. The sum is taken over the n components present in the

sample.

3. Experimental set up and diode lasers

Tunable diode laser spectrometers were used for the measure-

ments. The experimental arrangements differed somewhat for

the individual spectral regions depending on the type of laser

employed.

3.1 The 2100 cm21 spectral region

The construction of our infrared spectrometer has been

described previously.20 The layout is shown as a block diagram

in Fig. 1. The PbSe laser (Laser Components GmbH) was

placed in a liquid-nitrogen cooled Dewar vessel. The laser

frequency was controlled by the temperature and current using

a Laser Photonics unit, model L5731, with the temperature

varied within the range 80–100 K. The power of the PbSe laser

is less than 1 mW.

The laser beam was focussed by a toroidal mirror into a

monochromator (Czerny-Turner). The radiation leaving the

monochromator was directed either into a reference Ge etalon

(0.026 cm21) or into the photoacoustic cell filled with the gas

under study. In the cell, two microphones of the Knowles EK-

3024 type were located in the centre of a glass resonator with a

diameter of 5 mm. The laser beam was focussed between these

two microphones using an optical CaF2 lens. The direct

absorption spectrum was simultaneously recorded by a

photoconductor InSb detector working at the temperature of

liquid nitrogen. The signal from the detector was processed by

a phase-sensitive amplifier (Stanford Research Systems SR530

Lock-in amplifier). The spectra were measured using current

(saw-tooth) modulation of the laser (FM) at a resonance

frequency of 1010 Hz. The demodulated 2f signal (double the

modulation frequency) was converted and subsequently

processed using a PC (Fig. 1).


Lasers working in this spectral region consist of GaInAsSb/

AlGaAsSb material and operate at a temperature of 0–60 uC.

These lasers are amongst newly developed diodes with a power

of about 1 mW that are not yet commercially available.

The lasers were developed at the Centre d’Electronique et de

Microoptoelectronique de Montpellier (CEM2). Because the

lasers are highly monochromatic and single mode, it was not

Fig. 1 Experimental setup of the diode laser spectrometer.

This journal is � The Royal Society of Chemistry 2005 Analyst, 2005, 130, 1148–1154 | 1149

necessary to include a dispersion element in the system. A

Thorlabs TCLDM9 cold head connected to a home-

constructed control unit enabled adjustment of the tempera-

ture between 220 and 60 uC. The current was controlled using

a Laser Photonics unit, model L5731. The absorbed energy

was recorded either by two Knowles EK-3024 type micro-

phones or by a semiconductor InSb detector working at

liquid nitrogen temperature. Similar to the 2100 cm21 spectral

region, the spectra were measured using FM modulation

at 770 Hz.


A Newport 2010A external cavity tunable diode laser

(ECTDL) source working in the near infrared spectral region

(6250–6570 cm21) was used for the measurement. The modular

laser-diode head operates at 20–25 uC, with output power of

about 10 mW. The laser radiation was divided into two beams,

with one part directed by a mirror system into a wavemeter

(Burleigh WA-1000-NIR) and the second part passing through

the photoacoustic cell. In the cell, the absorbed energy was

recorded by two microphones (Knowles EK-3024), with

the laser beam focussed at the point midway between the

microphones using a spherical mirror. To permit comparison

of the photoacoustic and the direct molecular absorption

signals, the transmitted beam was simultaneously detected at

the output of the cell by an InGaAs detector (New Focus,

Nirvana 2017) working at room temperature. The signals

from the two detectors were processed by a phase-sensitive

amplifier (Stanford Research Systems SR810 Lock-in ampli-

fier). The whole system was controlled by a computer using the

UNB in-house SpecM program.21 The spectra were recorded

employing either internal frequency modulation of the laser

(FM modulation) or external modulation using a mechanical

chopper (AM modulation). A survey of the laser properties is

given in Table 1.

3.4 Photoacoustic cell

The photoacoustic cell is based on a glass acoustic resonator,

shown in Fig. 2, which was constructed of a tube with a length

of 200 mm and inner diameter of 5 mm. In the middle of

this resonator two holes were drilled on each side for

placement of the microphone sensors (Knowles EK-3024).

The external jacket was made of glass and was closed on each

end by an aluminium flange fitted with CaF2 windows of 2 mm

thickness.

From the standpoint of application of PA detection in the

various spectral regions, it was necessary to determine

the resonance frequency for each particular region. For the

2100 cm21 region, the resonance frequency was found to be

1010 Hz, for the 4300 cm21 region, 770 Hz, and for the

6500 cm21 region, 570 Hz. The sensitivity of the employed

microphones was 33 mV Pa21 and the value of the Q factor

was estimated to be about 20. The measurements were carried

out with the cell closed at constant gas pressure. All the mea-

surements were carried out at a room temperature of 22 uC.

Both frequency (FM) and amplitude (AM) modulation

techniques were explored at the resonance frequency of the

cell. For photoacoustic detection, we did not observe a great

difference between the intensities of the FM and AM signals

but it was found that AM is more sensitive to external noise. In

comparison of the FM and AM experiments, the intensity of

the measured signal was a factor of two larger for FM than for

AM. A further advantage of FM modulation lies in avoiding

the significant contribution to the surrounding noise caused by

the mechanical chopper itself. The FM signal was processed in

2f mode (second harmonic detection), where f corresponds to

the resonance frequency of the absorption cell.

The infrared high resolution absorption spectra of methane

and ammonia were used for detailed calibration of the gas

absorption. Fourier transform (FT) spectra of the calibration

gases were taken with a Bruker 120 FT spectrometer with a

resolution 0.003 cm21 in a spectral range from 1700–

7000 cm21. The absorption signal was derived from 100

co-added scans using a cell with a long optical path (25 m).

Fig. 3 depicts a comparison of the photoacoustic and

absorption spectra (measured with a liquid nitrogen cooled

Table 1 Laser parameters and obtained results

Spectral range 2100 cm21 4300 cm21 6500 cm21

Composition PbSe based lasers GaInAsSb/Ga AlAsSb/GaS InP based lasersLaser power/mW ,1 y1 10Threshold current/mA 50 45 45Studied gases OCS, CO CH4, NH3, N20 CO, CO2, NH3

Absorption Fundamentals Overtone and combination bands OvertonesResonant frequency for the cell/Hz 1010 770 570Detection limit (ppmv) OCS: 20 NH3: 0.1 NH3: 5

CO: 2000 CH4: 20 CO2: 1000CO: 2000

Upper pressure limit/Torr OCS: 300 NH3, CH4: 760 NH3, CO2, CO: more than 760CO: 500

Fig. 2 Photoacoustic cell and microphones (Knowles EK-3024).


InSb detector) of methane in the 4349–4352 cm21 region

together with the reference spectra obtained using FT high-

resolution spectrometry.

4. Analysis

All of our measured photoacoustic spectra obtained by phase-

sensitive detection were processed numerically using nonlinear

regression. The main goal of the processing was to find the

optimal parameters of eqn. (5) below.

Since the measurements were carried out in a pressure

interval where the absorption coefficient is affected by pressure

broadening, a Lorentz curve was chosen for the regression:

s(v)~s0

1z4(v{vmax)2

w2

(4)

The second derivative of the Lorentz curve was used for

fitting the spectra recorded in 2f mode:

d2s

v2~

128s0(v{vmax)2

1z4(v{vmax)2

w2

!3

w4

{8s0

1z4(v{vmax)2

w2

!2

w2

(5)

where s(v) is the absorption cross-section of the measured gas,

s0 is the value of the absorption coefficient at the line centre,

and w is the line-width (FWHM). Fig. 4 gives an example of a

measured absorption line and the results of the fitting. The

intensity of the signal was taken as the product of the FWHM

and the height of the signal.

5. Results and discussion

The primary goal of this work was to obtain diode-laser-based

photoacoustic signals in the three spectral regions and to

compare the data obtained.

5.1. 2100 cm21 spectral range

The photoacoustic spectra were measured in the 2050–

2120 cm21 spectral region of the PbSe laser diode. This is

the infrared region of the fundamental bands of the CO and

OCS molecules. An atlas of rotation–vibration spectra22 was

employed in determining the wavenumbers of selected lines.

The OCS (line 2095.0144 cm21, 0400–0000, P26) and CO (line

2094.8627 cm21, v 5 2–1, P12) transitions were selected from

the measured spectra as target lines for our experiments in

this region.

The optimal conditions for the performance of the photo-

acoustic cell (optimal acoustic frequency for the fundamental

longitudinal mode, optimal amplitude) were determined using

the OCS line at a pressure of 100 Torr. The dependence on gas

pressure of the optimised phase (in degrees) for phase-sensitive

detection was studied over the range 10–500 Torr in order to

find the best acquisition conditions. The phases changed most

in the pressure range between 10–100 Torr for the mixture of

CO with air (1 : 500), but were almost constant between 100

and 500 Torr.

In addition to the phase, the intensity of the signal also

depends on the total pressure inside the photoacoustic cell.

This dependence was measured for various mixtures of gas

relative to air. The maximum sensitivity was found at a

pressure of about 100 Torr for OCS and 150 Torr for CO2. As

mentioned above, the laser beam was focussed in the centre of

the cell at the point between the microphones in all the

measurements. It is highly probable that at high pressure

significant absorption would occur close to the entrance of the

cell so that the major acoustic signal would be generated some

distance away from the position of the microphones. Thus,

the detected signal would become weaker, explaining why

measurements could not be carried out up to atmospheric

pressure. At our optimum pressures, detection limits of

20 ppmv for OCS and 2000 ppmv for CO2 were achieved.

5.2 4300 cm21 spectral range

Two lasers were employed in this spectral range, working in

the 4100–4350 cm21 spectral region. This is the spectral range

of higher harmonic bands for NH3, N2O and higher harmonic

and combination bands for the CH4 molecule. The lines that

were studied for CH4 were identified using published data,23

while the HITRAN database of molecular spectra was used to

Fig. 3 Comparison of photoacoustic, direct absorption and Fourier

transform spectra of CH4.

Fig. 4 Solid curve—experimentally obtained PA spectrum of pure

CO2 (30 Torr); dotted curve—fitting according to eqn. (5).


identify the NH3 and N2O lines. Specific transitions for

CH4 (4362.0653 cm21, n3 + n4, 11, 11, F21–10, 10, F12),

NH3 (4350.4827 cm21, J94, K94, J05, K05, s) and N2O

(4361.6470 cm21, 0002–0000, P48) were selected from the

measured spectra and were used to investigate the dependence

of the intensity on the gas pressure and to determine the

detection limits.

The optimal conditions (optimal acoustic frequency for the

fundamental longitudinal mode, optimal amplitude) were

obtained using CH4 at a pressure of 100 Torr. The pressure-

dependence of the optimum phase for phase-sensitive detection

was studied in the range 1–760 Torr for a mixture of CO with

air (1 : 500). In this case, the phases changed most in the

pressure range 1–100 Torr. As found for the 2100 cm21

spectral range, almost no change occurred between 100 and

760 Torr. The dependence of the signal intensity on the total

gas pressure was also measured for various mixtures in air. An

optimal pressure of 70 Torr was found for NH3, 95 Torr for

N2O and 90 Torr for CH4. The detection limits were estimated

to be 20 ppmv for CH4, 0.1 ppmv for NH3 (Fig. 5), and

10000 ppmv for N2O.

5.3 6500 cm21 spectral range

The photoacoustic spectra of the CO, CO2 and NH3 molecules

were measured in the 6250–6570 cm21 region using the

external cavity tunable diode laser. The investigated lines of

the CO and CO2 molecules were identified using the HITRAN

database, while the NH3 transitions were found in published

data for NH3 in the 6400–6900 cm21 region.24 Transitions

of CO (6377.4066 cm21, v 5 4–1, R7), CO2 (6362.5043 cm21,

30012–00001, R20) and NH3 (6487.8350 cm21, 2n3, J96, K92,

J07, K03, a) were selected from the spectra.

The optimal conditions (resonance acoustic frequency of the

fundamental longitudinal mode) were found at a pressure of

100 Torr. The phase dependence on pressure was studied in

undiluted CO gas in the range 5–760 Torr. The phase changed

substantially in the pressure range 5–100 Torr and only very

slightly between 100 and 760 Torr. For CO2 and NH3, the

phase change was studied in greater detail and was recorded

at various gas-to-air ratios, namely ratios of 1 : 5, 1 : 10 and

undiluted for CO2 and 1 : 10, 1 : 100 and 1 : 1000 for NH3. For

both gases, the phases changed most in the pressure range

0–200 Torr. For experimental fit the single exponential decay

function was used. We used the phase shift approach for

determination of the lifetime:25

t 5 (2pf)21tan w (6)

where w denotes the maximum phase shift during the pressure

variation, t is a lifetime for the mixture under study of the

particular energy state and f is the modulation frequency. As

shown in Fig. 6, the value of t increases with decreasing

concentration of the test gas in the mixture, with the exception

of NH3 where its value is lower for a 1 : 100 mixture than for a

1 : 10 mixture.

The optimal pressure was found from the pressure

dependence of the intensity of the photoacoustic signal, shown

in Fig. 7. This dependence was measured for various mixtures

of gas relative to air, and a polynomial fit was drawn through

the individual points. The dependence was measured up to

atmospheric pressure but, as can be seen from Fig. 7, the signal

could also be observed at pressures above 760 Torr. The best

performance of our system was found for a total pressure

of 225 Torr for both NH3 and CO2. In the case of CO,

however, the optimal pressure for the photoacoustic signal

varied betwen 250–350 Torr depending on the gas mixture. We

currently have no explanation for this variation. The detection

limits achieved for the selected lines of the gases studied were

2000 ppmv for CO, 1000 ppmv for CO2 and 5 ppmv for NH3.

5.4 Detection of NH3 and CO2 in exhaust gases

In this phase of the study, we collected samples of exhaust

gases directly from automobile exhaust pipes using a short

teflon tube feeding into an evacuated sample bottle (metal

pressure tank). Three different experiments were carried out in

analysis of the mixtures.

In the first experiment, concerned with monitoring the

concentrations of NH3 and CO2, exhaust samples were taken

from a Dodge Shadow S (year of manufacture, 1991) using

two different petrols (Regular and Premium). Fig. 8 depicts the

Fig. 5 Dependence of the intensity of the photoacoustic signal on concentration for mixtures of CH4 (4362.0653 cm21, n3 + n4, 11, 11, F21–10, 10,

F12) and NH3 (4350.4827 cm21 line) with air.


differences in the trace gas concentrations for the two different

fuels. The concentration of NH3 equalled 120 ppmv for

Regular petrol, which was 12 times higher than the concentra-

tion of 10 ppmv found for Premium petrol. The spectrum of

ammonia measured in the laboratory at a concentration of

10 ppmv is also included in Fig. 8. for comparison.

The second experiment was concerned with measuring the

concentrations of NH3 and CO2 in the exhaust gases after

various degrees of heating of the motor. We examined a total

of three samples taken from a Honda Accord (year of

manufacture, 2000) during the same day. The first sample

was taken 5 min after starting the motor with it running out of

gear, and the next mixture was taken after the automobile had

been running for 7 min at a speed of 50 km h21. The last

sample was taken after the vehicle had been travelling for

25 min at a speed of 100 km h21. The first mixture was found

to contain 29 ppmv of ammonia, whereas the next two

mixtures contained almost identical amounts of 9.6 and 9 ppmv

of this gas. The CO2 concentration was identical for all three

measurements and equalled 24.4 6 104 ppmv.

In the last experiment, we compared the combustion

products from the Honda Accord and the Dodge Shadow S

automobiles. The NH3 concentration was 9 ppmv for the

Honda vehicle and 30 ppmv for the Dodge. The CO2

concentration differed by almost ten-fold, equalling 26.8 6104 ppmv for the Honda and 29.4 6 103 ppmv for the Dodge.

6. Conclusions

In this work, we have investigated the application of

photoacoustic detection with laser diodes in the middle and

near infrared spectral ranges. The performance of the lasers

was tested in detailed measurements of well-resolved absorp-

tion spectra of the atmospherically important species CO2,

NH3, CO, CH4, and OCS. A very simple, cheap, portable glass

photoacoustic resonator was constructed. This acoustic cell,

containing two inexpensive electret microphones, gave results

Fig. 6 Dependence of optimum phase (in degrees) on the total gas pressure for CO2 and NH3 mixtures with air. Solid curve—fitting using first

order exponential decay.

Fig. 7 Dependence of photoacoustic signal on the total gas pressure

in the photoacoustic cell for NH3 (6487.8350 cm21). *Curve 1 : 10 was

divided by a factor of 5.

Fig. 8 Spectra of NH3 in samples from automobile exhaust: (a)

Regular petrol, (b) Premium petrol, (c) Laboratory measured spectrum

of NH3 (10 ppmv concentration).


fully equivalent in sensitivity to conventional direct absorp-

tion measurements with an LN2-cooled semiconductor InSb

detector. The parameters of the developed photoacoustic

system, including those involved in the modulation technique,

were explored and optimized. The detection limits at optimum

pressure were determined for all of the gases studied (Table 1),

with the maximum sensitivity being achieved for NH3

measurements. An experimental study of trace concentrations

of NH3 and CO2 in automobile exhaust gases was also

conducted in the near-infrared region.

Our best results were obtained in the 4300 cm21 region,

employing new MQW GaInAsSb/AlGaAsSb lasers operating

at room temperature. We obtained the maximum concentra-

tion sensitivities of 100 ppbv for NH3, 20 ppmv for OCS and

20 ppmv for CH4, respectively.

In the case of CO, we obtained similar detection limits in the

6500 and 2100 cm21 spectral ranges but different results for

the upper pressure limits. We were able to observe photo-

acoustic signals in the 6500 cm21 region at atmospheric

pressure but for the lead salt diodes at 2100 cm21 much lower

pressure had to be employed as shown in Table 1.

In summary, this paper describes a very simple and

cheap technique, applicable to a broad range of rotational–

vibrational spectra, for trace-gas sensing of molecules absorb-

ing in the middle and near infrared spectral ranges. The main

emphasis of the study comprises a detailed comparison of

diode-laser-based photoacoustic measurements in different

infrared spectral ranges (covering molecular overtones and

combination bands together with fundamental vibrational

transitions) using the identical photoacoustic apparatus.

Careful exploration of the system operating parameters and

optimization for maximum detection sensitivity has shown

that the technique can be succesfully applied to detection

of trace amounts of gas pollutants in the atmosphere or in

automobile exhaust gases.

Acknowledgements

This work is a part of the research programs of the Grant

Agency of the Academy of Sciences of the Czech Republic

(Grant No. A4040104) and project No. OC715.50 of the

Ministry of Education, Youth and Sports, project No.

OC723.001 in the frame of the COST723 action. SC thanks

the J. Heyrovsky Institute of Physical Chemistry for financial

support with the 120 Bruker reconstruction. LHX and RML

acknowledge financial support from the Canadian Institute for

Photonic Innovations (CIPI) under the federal Networks of

Centres of Excellence program and from the Natural Sciences

and Engineering Research Council of Canada.

References

1 A. G. Bell, Am. J. Sci., 1880, 20, 305.2 M. A. Gondal, I. A. Bakhtiari and S. M. A. Durrani, J. Anal. At.

Spectrom., 1998, 13, 455.3 Air Monitoring by Spectroscopic Techniques, ed. M. W. Sigrist,

Chemical Analysis Series, John Wiley, New York, 1994, vol. 127,p. 163.

4 A. Veres, Z. Bozoki, A. Mohacsi, M. Szakall and G. Szabo, Appl.Spectrosc., 2003, 57(8), 900.

5 V. Zeninari, B. Parvitte, D. Courtois, V. A. Kapitanov andY. N. Ponomarev, Infrared Phys. Technol., 2003, 44(4), 253.

6 T. H. Vansteenkiste, F. R. Faxvog and D. M. Roessler, Appl.Spectrosc., 1981, 35(2), 194.

7 V. Slezak, G. Santiago and A. L. Peuriot, Opt. Laser Eng., 2003,40, 33.

8 B. S. Ernegger, P. L. Meyer and M. W. Sigrist, Helv. Phys. Acta,1985, 58(5), 829.

9 F. J. M. Harren, R. Berkelmans, K. Kuiper, S. T. Hekkert,P. Scheepers, R. Dekhuijzen, P. Hollander and D. H. Parker, Appl.Phys. Lett., 1999, 74(12), 1761.

10 F. J. M. Harren, J. Reuss and F. Lenz, Gartenbauwissenschaft,1997, 62(5), 193.

11 L. B. Kreuzer, J. Appl. Phys., 1971, 42, 2934.12 P. L. Meyer and M. W. Sigrist, Rev. Sci. Instrum., 1990, 61, 1779.13 Monitoring Gaseous Pollutants by Tunable Diode Lasers, ed.

R. Grisar, H. Preier, G. Schmidte, G. Restelli and D. Seidel,Kluwer, Dordrecht, The Netherlands, 1987, part 1 of a series.

14 Monitoring Gaseous Pollutants by Tunable Diode Lasers, ed.R. Grisar, H. Preier, G. Schmidte, G. Restelli and D. Seidel,Kluwer, Dordrecht, The Netherlands, 1989, part 2 of a series.

15 Monitoring Gaseous Pollutants by Tunable Diode Lasers, ed.R. Grisar, H. Preier, G. Schmidte, G. Restelli and D. Seidel,Kluwer, Dordrecht, The Netherlands, 1992, part 3 of a series.

16 A. Vicet, D. A. Yarekha, A. Ouvrard, R. Teissier, C. Alibert andA. N. Baranov, IEE Proc.: Optoelectron., 2003, 150(4), 310.

17 S. Civis, V. Horka, T. Simecek, E. Hulicius, J. Pangrac, J. Oswald,O. Petrıcek, Y. Rouillard, C. Alibert and R. Werner, Spectrochim.Acta, Part A, in press.

18 S. Civis, V. Horka, J. Cihelka, T. Simecek, E. Hulicius, J. Oswaldand J. Pangrac, Appl. Phys. B, to be published.

19 Yoh-Han Pao, Optoacoustic spectroscopy and detection, AcademicPress, New York, 1977.

20 A. Popov, V. Sherstnev, Yu. Yakovlev, S. Civis and Z. Zelinger,Spectrochim. Acta, Part A, 1998, 54(6), 821.

21 Zhengfeng Liu, MSc Thesis, University of New Brunswick, 2004.22 A. G. Maki and J. S. Wells, Wavenumber Calibration Tables from

Heterodyne Frequency Measurements, NIST Special Publication821, Washington, 1991.

23 V. Malathy Devi, D. Chris Benner, M. A. H. Smith andC. P. Rinsland, J. Quant. Spectrosc. Radiat. Transfer, 1994,51(3), 439.

24 L. Lundsberg-Nielsen, F. Hegelund and F. M. Nicolaisen, J. Mol.Spectrosc., 1993, 162, 230.

25 J. R. Lakowicz, in Principles of Fluorescence Spectroscopy, KluwerAcademics/Plenum Publishers, New York, 2nd edn., 1999, p. 97.


Laser diode photoacoustic detection in the infrared and near infrared spectral ranges

Documents