1.1 The apparatus

Generally the right angle PA setup is used for studies with liquid samples, in which the acoustic waves are detected at right angle of the pump beam. The pressure transducer is clamped to the side of the quartz cuvette and acoustic impedance is partly matched by means of a thin layer of either vacuum grease or silicon oil.
An alternative setup is the front-face detection geometry, where the transducer is clamped behind the cuvette, and a dichroic mirror is placed in between cuvette and transducer, in order to reduce the background signal coming from the absorption of the laser pulses at the transducer surface.
 
Lasers

In the photoacoustic experiments, photoexcitation is generally achieved by nanosecond pulsed lasers, although a few applications of picosecond lasers have been also reported.
The laser pulse energy on the sample is usually in the 1-100 mJ range for right angle detection, but can be higher for the front-face geometry, since in this case the beam size is larger and the photon density is lower. In both cases, care must be taken in checking the linearity with laser fluence of the detection system and the photochemical system under examination (vide infra).  Beam shape and pointing stability, and pulse-to-pulse energy stability are key features for lasers used in time resolved PA applications. Although the energy used in the right angle PA experiments is in the mJ range, it is very useful to have mJ pulses from the laser, so that beam shaping can be done without focussing, by simply skimming the incident light in passing through either a pin hole or a slit (vide infra).

Laser pulse energy normalization

The absolute value and the fluctuation of the laser pulse energy incident on the sample is monitored by splitting part of the beam to an energy meter. In the right angle detection, this is generally done before shaping the unfocussed beam with a slit or a pin hole, which are positioned in front of the cuvette.  The absolute value of the energy incident on the cuvette can be determined by measuring the ratio between the energy of the split beam and the unsplit part of it (it is important to check this last value before and after the cuvette and take the average value to compensate for reflections at cuvette surfaces). Monitoring the fluctuations in the energy of the split beam allows the monitoring of the fluctuations in energy of the unsplit beam. Measured signals must be normalized (divided) by the actual energy of the laser pulses during acquisitions.
It is not strictly necessary to have absolute energy values (in J) since the normalization is a relative scaling process.  However, it is highly recommended that the absolute value of the laser pulse energy used in the experiment is known, in order to have a better control on the photophysical and photochemical  processes.
 

Shaping of the beam

In the right angle geometry, the laser beam is generally not focussed and is shaped by means of either a pin hole or a slit (see  a closeup ) positioned in front of the quartz cuvette. Mild focussing, either with spherical or cylindrical lenses, can also be used to shape the beam, expecially when working with low power lasers as, e.g., nitrogen-pumped dye lasers.  Care must be taken, however, not to use lenses with short focal length and, in any case, not to have the focus inside the cuvette, in order to avoid fluence gradients that may lead to nonlinearities.
On the contrary, in front-face detection (see a closeup) the laser beam is generally shaped so that the beam "fills" the cuvette, and in this case laser pointing stability is less critical. Laser fluence is greatly reduced by adopting this geometry.
 

Ultrasonic detectors and signal conditioning

The pressure wave induced in solution is detected by sensitive piezoelectric transducers.  Piezoelectric transducers include ceramic transducers as the popular Lead Zirconate Titanate (PZT) transducers (also used as accelerometers), which are essentially differential transducers, and organic films as polyvinylidene difluoride (PVDF), which are rather broadband in nature.  These transducers are characterized by very high output impedance (> 100 MW ) which requires adequate signal conditioning.  Commercially available transducers have resonance frequencies in the 0.1-100 MHz range (Panametrics), DC-20kHz (PCB). After impedance adapting, the 50 W load can be fed into high gain amplifiers (at least 100 X, bandwidth of at least 10 MHz).  Commercial tranducers can be generally purchased with the appropriate signal conditioning electronics.
 
 

Digital sampling oscilloscopes

Digital sampling oscilloscopes (DSO) are generally used to digitize the signals.  Early applications included boxcar integrators and transient recorders, which have by now been superseded by the outpassing performances of the new DSO.  Typical figures include 50 W DC input coupling, 1-5 mV vertical sensitivity, 8 bit single shot vertical resolution of the ADC converter, averaging capability with 16 or 32 bit processors, > 100 MHz bandwidth at the maximum amplification of the front-end amplifier, > 100 MS/s maximum sampling rate.
Generally, in PA applications it is not necessary to use DSO with deep memories, since the time window digitized in the experiment is about 10 ms.  At a sampling rate of 400 MS/s this would require 4,000 data points.
It is extremely useful to have remote control on the scope, for instance via a GPIB interface, in order to transfer the data to a computer for the following data analysis.
Modern DSO are also equipped with hard and/or floppy disks which can be used for temporary data storage.
Commonly used oscilloscopes include LeCroy, Tektronix and HewlettPackard.

Sample holder

The sample holder is a very critical component of the PA setup.
The most important feature of the sample holder is thermal stability and homogeneity.  The temperature of the sample in the experiment determines the values of the compressibility, the thermal expansion coefficient and the speed of sound, parameters that appear in the expressions describing the PA signal. Uncontrolled temperature gradients inside the sample lead to PA signal generation which is generally non reproducible from sample to sample and render both the signal amplitude and time profile meaningless.  Temperature control inside the cuvette for room temperature studies with aqueous or organic solutions must provide at least 0.2 °C stability; requirements are higher (0.02 °C stability) for lower temperature studies of aqueous solutions, because of the strong temperature dependence of the thermoelastic parameters.
The quartz cuvette and the transducer are placed inside a metal block which is thermostated by means of a water circulating bath. Since the sample holder is frequently kept for long time at temperatures below room temperature, care must be taken in insulating the sample holder by means of suitable material.  Moreover, dry gas must be flown across the cuvette walls in order to prevent moisture condensation which may impair the possibility of performing data analysis of experimental waveforms.
Recently, thermoelectric coolers have been introduced to control the temperature inside the metal black housing the sample cuvette.  In our laboratory we have both a thermostated sample holder using a water circulating bath and a recently implemented  temperature controlled sample holder from Quantum Northwest, Inc. (model TASC 300) which assures a temperature stability of better than 0.02 °C inside the solution.  The sample holder allows magnetic stirring of the solution inside the cuvette and dry gas purge of the atmosphere, in order to prevent condensation at low temperatures. In experiments where the atmosphere must be carefully controlled, it is highly recommended that gas-tight quartz cuvettes are used.
 

Linearity checks
 
Linearity with laser fluence must be checked for every sample under investigation.  Checks must be performed at two levels.
Linearity of reference signals.  The reference signal amplitude (corrected for the absorbance but not for the laser fluence) must plotted as a function of the laser fluence and the resulting plot must be linear with zero intercept.  This assures linearity of the detection system and the photoinduced response of the reference compound.  Lack of fulfillment of this linearity check must be carefully investigated, in order to understand wheter the nonlinearity arises from the detection system (transducer + amplifier + digitizer) or from the reference compound (two-photon effects, saturation, photodegradation,....)
Sample preexponential factors. The preexponential factors must be independent of the laser fluence to ensure that the system doesn't show any saturation or multiphotonic effect.
The use of picosecond lasers is more troublesome with regard to this linearity requirement, since peak powers are much higher than for nanosecond lasers.  This is important since the photoacoustic signals are proportional to the energy of the pulses, therefore requiring higher fluences when using short pulses.

Linearity with the absorbed energy, i.e. with the parameter (1-10-A) is important as well, when interactions between the absorbing solutes should be avoided.  Lack of linearity in this plot may indicate that ground (or excited) state interactions are present.  This plot can also be used to improve the quality of the data analysis.
 

 back to the index