The PA effect is discovered by Alexander
Graham Bell in 1881 (Fig. 1). With the modulated
sunlight focussed into the vessel
containing gaseous sample of interest. Bell has detected (the
intensity of the sound varied from
sample to sample) audible signals at the same periodicity.
Fig. 1 A.G. Bell's experimental
set-up used when discovering the PA effect in 1881. The sunlight
reflected by the plane mirror C,
passes through the opning in rotating wheel B before being
collected and focussed by a parabolic
surface into vessel A.
The excited gas molecules collide
with its neighbouring molecules and return to a ground state
transfering some of its energy.
A periodically varying heat source is generated causing an gas
expansions that can been detected
as pressure wave. For a vessel filled with a gas with an
absorption coefficient s, the amplitude
S of the generated acoustic signal was found proportional
to the product of intensity of
the incoming radiation P, the concentration of the species c and the
cell constant R, i.e. :
S = R P s c
This observation brought Russian
scientist Viengerov sixty years after Bell's discovery to the
idea to construct an instrument
based on a PA effect for the measurement of weak absorbances.
As long as the concentration of
the sample is moderate conventional methods provide
satisfactory results; for A<0.001
standard spectroscopic techniques do not provide reliable data
anymore. On the other hand, PT
techniques are zero-methods (no absorption, no signal) i.e.
amount of absorbed energy is measured
directly (benefits greatly from the availablity of strong
radiation sources). The experiments
have shown that absorbances as low as 10-10 it could
readily be detected without any
preconcentration or preparation of the sample, this indicates the
general potential of PT based methods
for trace detection in general.
Some general considerations concerning
modern PA spectroscopy
Any PA spectrometer comprises a
radiation source, modulator, the cell to accomodate the
sample and a detector. In a modern
PA spectrometer (Fig. 2) the sunlight is replaced by a
suitable laser; mechanical, electronic
(or electro-optical) modulation schemes are used with
sensitive transducer microphone
and lock-in detection. The objective of any PA cell design is to
enhance the amplitude of generated
signal while efficient supressing the undesireable background
contribution.
Fig. 2 Overview of photoacoustic
experiment [1] = nitrogen flow control, [2] = evaporator, [3] =
temperature controlled bath, [4]
= photoacoustic cell, [5] = heater, [6] = exhaust with flow
meter, [7] = CO2 laser, [8] = spectrum
analyzer, [9] = chopper, [10] = photoacoustic resonator
with microphone, [11] = pyroelectric
detector
As far as the dimensions and a shape
form of the cell are concerned, two main types exists.
Resonant cell are constructed to
support one of its acoustic resonances. When modulated with a
frequency corresponding to that
of acoustic resonance itself, the cell acts as an acoustic amplifier
with a high quality factor. The
microphone is placed at the location that corresponds to the
maximum of pressure wave. Since
the frequency of acoustic resonance depends on a density and
the composition of the gaseous
mixture, the resonant cell is very susceptible to temperature
fluctuations. On the other hand
non-resonant cell is much smaller than its resonant counterpart
with lower Q factor.
The Effects of Interference
Just as like in any other spectroscopic
methods, absorption and generation of signals due to
interference affect the full analytical
potential of the PA technique. In particular, when dealing
with the step-tunable sources (spectral
coincidences between the source and the specimen are not
a priori guaranteed), the problem
of interferences deserves additional attention.
An elegant way to suppress interference
is to make use of Stark effect, a phenomenon induced in
polar gases when exposed to the
external electric field. The interaction with the applied field
causes a shift of the molecular
energy levels. Under such conditions in the sample only these
molecules that respond to both,
the incoming radiation and to applied Stark field, can produce a
signal.
The PT heating can produce a refractive
index gradient (RIG). This is due to the lower density of
the medium (sample or coupling
fluid) caused by the local temperature rise, decays in time
following the diffusional decay
of the temperature profile, and remains near the initial optically
excited region. The thermal RIG
generated by the excitation laser affects the propagation of its
beam resulting in "self-defocussing"
or "thermal blooming". Self-defocussing occurs because the
derivative of the refractive index
with respect to the temperature is usually negative (dn/dT < 0);
this technique is called single
beam Thermal Lensing. The dual beam thermal lensing
technique (one excitation laser
(pump beam) and another (very) weak prope beam) provides a
significant enhancement of sensitivity
(especially at loger wavelengths). Several pump/probe
beam configurations possible (parallel,
collinear and perpendicular).
Thermal lens experiments can only
be performed with laser sources having Gaussian intensity
profiles. Instead of monitoring
the intensity changes at the beam axis one can also probe the
thermal lens at there where the
temperature gradient is maximal. Probe beam sent along the
locations corresponding to maxima
of the temperature gradient will be deflected. This method is
known as the photothermal deflection
(PTD) and is the most sensitive among PT techniques. The
angle of deflection is directly
proportional to dn/dT and dT/dr (also to the concentration of
absorbing species) of the sample
and the interaction length between the two beams. The angle of
deflection is measured with a position
sensitive diode (lateral or quadrant).
The operational principle of opto-thermal
window technique (OW), actually a variant of
conventional PA spectroscopy is
as follows: a modulated (laser) radiation passes through the
OW cell before impinging on the
sample. The OW cell is actually an optically transparent disc
(having large thermal expansion
coefficient) the rear side of which is provided with
piezoelectric transducer. Due to
the absorption of radiation, sample's temperature rises and the
generated heat diffuses into the
disc (being in a good thermal contact with the sample) that
expands the induced stress is detected
by a lock-in amplifier.
In comparison to conventional PA
spectroscopy, the OW method offers some attractive features.
At first, the requirement for accomodating
the sample in the sealed cell is no longer an impetus.
In addition, the OW signal not
only remains unaffected by thermal expansion of the sample but is
also less susceptible to the effect
of other sample's thermal parameters. Finally, as long as it
exceeds sample's thermal diffusion
length, the thickness of the sample is not relevant making OW
technique more practical for quantitative
IR analysis of strongly absorbing fluids and semi fluids.
For an optically opaque and thermally
thick (an ideal) sample, making a good thermal contact
with a thermally thick OW , the
amplitude of the normalized, dimensionless optothermal signal is
related to the product of the absorption
coefficient and the thermal diffusion length. This forms
the basis for obtaining the absorption
spectrum of the sample under investigation, provided
optical and thermal properties
of a reference sample at a given wavelength and modulation
frequency are known.
The PT methods outlined above are
applicable to all three physical states of matter. So far PA
method (with strong IR laser) has
been used to monitor low concentrations of several urban and
industrial pollutants, pesticides
an many other gases. In majority of cases the high sensitivity
(below the ppbv level) has been
reached. Measuring in this concentration range requires very
clean equipment. In our own experiments
the PA method was shown capable of detecting very
low emission rates of volatiles
released from the tubing material used to transport gases of
interest from the sampling site
to PA cell. Metabolic processes in refrigerated meat specimens
were studied in order to establish
the onset of spoilage. Production of gases from soil and the
microbiological acitivity rates
were studied too. Likewise numerous attempts to determine the
fluxes of pollutants from the atmosphere
to the vegetation were carried out. Finally, considerable
amount of work was performed to
study the spectra of fatty acids to investigate psychophysical
response functions in humans. The
research is now underway to develop new methodology and a
new instrument for determination
of ultralow amounts of moisture in granulated and porous
samples such as flours, sand or
powdered milk.
As far as the work in liquid phase
is being concerned, the PTD method was used to improve
detection limit for orthophosphate
and ammonium ion in water and soil water solutions. This is
achieved by means of colorimetric
reactions (complexes that absorb strongly at the visible
wavelength); the absorption of
a complex is then directly related to the absorbance of the original
compound.
Other methods related to PA and
TO techniques were also developed throughout the past years;
their detailed description is beyond
the scope of this update. The field is far from being
exhausted and plenty of novelties
are expected to emerge in a near future.
Roslee Abdul Jalil