Maaneli Derakhshani
Phys. 287, Fall 2005
Prof. Harold Metcalf


Discussion of Semester Research on Single-Bubble Sonoluminescence


The following report will outline the continuation of the research 
done on SBSL (Single-Bubble Sonoluminescence) from summer 2005. The
continued research initially focused on improving bubble stability and
luminescence, while also enhancing the observability of the bubble via
a CCTV camera. However, unanticipated variations in the gas conditions
of the bubble slowed progress along these lines.

In the beginning of the semester, the procedure for producing SBSL was
relatively routine. The 100 ml flask would be filled with distilled
water, just up to the bottom of the neck of the flask, to create a
spherically symmetric volume of water. The vacuum absorption pump would
then be used lower the pressure in the spherical flask, via a rubber
stopper that sealed the flask and connected it to the pump via a rubber
valve. The pressure would be lowered until the water began to boil.
This always begins taking place at a pressure of 20 Torr or lower, as
indicated by the gauge meter in use.

As an aside, it should be noted that the process of reducing the gas
concentration in the water is the most delicate part of the experiment.
Initial experiments led to the development of a general procedure of
pumping on the water, which was developed from observations that
degassing the water (that is, pumping it down by vacuum, until all
observable air bubbles in the water are boiled away), makes it
impossible to sustain or even form a trapped air bubble. When a bubble
was injected, it would migrate towards the center of the flask, but
then instantly dissolve. When air was partially dissolved back into the
water (over the course of 6 hours of exposure to STP (Standard
Temperature and Pressure)) the trapped bubble dissolved within several
seconds instead. After some trial and error, the following general
pumping procedure was developed, allowing one to consistently produce
the necessary concentration of gas to trap and stabilize bubbles of the
proper size. The pumping procedure described below can also be found in
a journal entry dated on September 20, 2005:

1. Pump on the water in the flask until the pressure reading from the
gauge meter is down to 20 Torr.

2. Once at 20 Torr, immediately close off the vacuum pump valve and
test to see if you can trap a stable bubble for a longer duration.

3. Repeat above procedure if initial pumping is not enough to form a
stable bubble.

Following the degassing process, the diametrically opposed PZT (Plumbum
Zirconium Titanate) transducers would then be driven by a voltage
signal from the function generator, ranging from 25-27 kHz, and voltage
amplitude of 4.0-6.2 volts. It should be noted that SBSL could never be
observed at a voltage below 4.0 volts or 6.2 volts, suggesting that
this is the range of bubble stability necessary to produce SBSL. The
operational amplifier and 36 volt power supply would of course increase
the strength of the signal from the voltage generator, since otherwise
the voltage signal would be too weak otherwise to induce oscillations
in the PZT transducers.

When the bubbles were injected in the flask, they would migrate towards
the center of the flask and get trapped at the acoustic anti-node. An
oscilloscope which was connected to the microphone transducer located
on the bottom of the flask, via a coaxial cable, would then display the
sinusoidal signal produced by the periodic bubble expansion and
implosion. Evidence of a trapped bubble could be made either visually,
in which case a Fiber-Lite A3200 high intensity light source was used,
or more easily by looking for ripples in the sinusoidal signal. It
should also be noted that the flask and ring stand were and are
currently placed inside a .5 meter-cubed black painted cardboard box, to allow
for easier observance of the SL bubble. Since SBSL is only produced
when the PZT^�s oscillate at the resonant frequency of the flask, the
signal frequency would be varied in steps of 10 Hz, until the amplitude
of the signal on the oscilloscope was at maximum. The ferrite rod core
would also have to be adjusted in the inductance coils, as part of
achieving electrical resonance in the circuit. One resonance was
achieved, the voltage would be incremented by steps of 0.01 Volts, just
below the point where the bubble would pop. Again, this voltage would
be in the range of 4.0-6.2 volts. Once these conditions were met,
sonoluminescence was observed. Initial experiments could maintain a
stable SBSL bubble for as long as 5 minutes.

Subsequent experiments, which continued up to October, focused on
increasing bubble stability by mixing the distilled water with 5 ml's of
glycerin. This idea was based off of a journal entry by the previous SBSL
student, Kenneth Lee, who reported bubble stability for as long as 5
minutes with glycerin. While it was the case that a stable bubble could be
maintained indefinitely, producing SBSL was impossible, regardless of the
driving voltage. The glycerin mixed water was even re-pumped to remove any
excess gas, but this changed nothing. The glycerin doped water was also
replaced with freshly degassed water, but it was still not possible to
produce SBSL, even at a driving voltage of 7 volts, which is quite
excessive. After dozens of trials and attempts to attain SBSL once again,
it was found that the optimal pumping time on the flask was within five
seconds. In spite of this knowledge, it remains difficult to pump on the
water, within the proper duration. One interesting finding that certainly
does affect the pumping procedure is the amount of liquid nitrogen poured
into the absorption pump Styrofoam container. The liquid nitrogen is
acquired from a source in the Nuclear Structure lab. Normally, one small
tank is filled completely and used for the pump. The precise volume of
this tank is not known. Nevertheless, with one tank's worth of liquid
nitrogen in the absorption pump, it takes approximately five minutes for
the pressure in the flask to drop to under 20 Torr. This is an approximate
time because it varies with the precise amount of liquid nitrogen, and
possibly other as of yet unidentified variables. One possible variation in
the vacuum pump that may account for the difficulties in pumping is the
baking of the vacuum pump, which was done just before using glycerin in
water. Other possible variations include the distilled water source.
Contamination of the water could lead to variations in the gas content of
the water, and thus alter the proper pumping duration. Following these
pumping difficulties, the reuse of glycerin in smaller quantities (1-2 ml)
led to indefinite sustenance of the sonoluminescing bubble.

Additions and modifications to the experimental setup were also 
made. An aluminum foil tube was constructed and inserted into the high
intensity light source, to focus the light directly on the flask, thus
prevented unnecessary illumination of the surrounding room. A CCTV
camera was also screwed down into the lab table and connected to a TV
monitor, which magnifies the image of the bubbles in the flask by a
large but unknown factor. The intent was to use this camera to observe
the SL bubble in greater detail.

Unfortunately, even when SL was attained, the CCTV camera did not
capture enough light to make the bubble visible on the TV monitor, even
in pitch black darkness. Attempts were also made to quantitatively
measure the intensity and thermal spectrum of the SL bubble, with a
photomultiplier (PMT) tube. Unfortunately, the PMT tube is currently
missing in the lab, and so this experiment was not possible.  Further
work in the following semester will focus on measuring the light output
of the bubble with the PMT (once it is found), enhancing visible light
output, and experimenting with more exotic components. For example, the
use of the wavelength shifter Luminol, or a novel electrolytic solution
such as Vitamin Water, should predictably increase the intensity of the
SL bubble. Other possibilities include the use of a time-varying
magnetic field from Helmholtz coils to induce a force on the SL bubble.
Reasons to suppose that the bubble would be affected by a magnetic
field are that thermal temperatures of 10,000 Kelvin, which are
temperatures reported from other experiments, as indicated by the
thermal radiation spectrum, also indicate dissociation of oxygen from
hydrogen. Indeed similar results have been confirmed by Suslick in
experiments with sulfuric acid [1]. This is another possible
experiment, the replication of Suslick's experiment with sulfuric
acid. Another possible experiment is the use of a YAG laser to induced
single bubble cavitation luminescence (SBCL) in light water. But
perhaps the most interesting experiment would be the following:

The Nd:YAG laser, which is a pulsed, Q-switch laser, can be used to
vaporize the liquid and produce gas pockets, i.e., bubbles, which then
grow on their own, usually to a diameter of 1-2 mm (as opposed to 100
microns for SBSL!), and then collapse due to their own elastic rebound.
This is called single bubble cavitation luminescence or SBCL, not
sonoluminescence, because no sound waves are used to drive bubble
implosion. These bubbles also tend to be less spherical upon collapse,
and thus break into two, smaller cavitation bubbles which then collapse
and luminesce. However, these experiments have only been done with
light or heavy water. Nevertheless, SBCL also produces brighter
luminescence than SBSL, that is, more photons per second are detected.
However, the thermal spectrum also indicates that maximum bubble
temperatures are actually lower than SBSL. But this can be attributed
to the bubble partitioning. Sulfuric acid on the other hand has a
higher phase-change coefficient (more elastic and fewer gas pockets)
and can therefore allow for the formation of more spherical bubbles
upon collapse. However, no such SBCL experiments have ever been done
with sulfuric acid or any other liquid; and since one can also combine
acoustic standing waves and a YAG laser to further facilitate bubble
expansion and implosion, the use of an acoustic field may make the
subsequent collapse more spherically symmetric, thus reducing the
chance of bubble partitioning. So such an experiment would work in the
following way:

1. At the time of maximum negative acoustic pressure, the YAG laser
would vaporize the sulfuric acid liquid at the acoustic anti-node
to produce a bubble.

2. This spherically symmetric bubble would then expand because the
inner bubble pressure is greater than the outer surface pressure.

3. At maximum outer positive acoustic pressure, the spherically
symmetric bubble would then implode.

4. Because the acoustic pressure difference is uniform around the
bubble, the pressure difference should produce uniform force vectors
pointing towards the bubble center, all around the bubble's outer
surface, which one can easily visualize.

5. The resultant collapse would then produce luminescence in a very
similar way as SBSL (Single Bubble Sonoluminescence). This whole
process repeats when a new bubble is produce by the YAG laser. But
aside from these more ambitious experiments, it will be most important
to better understand the chemistry of dissolved gas in distilled water
and to develop a more consistent and reliable pumping procedure.


References

[1]. Nature 434, 52-55 (3 March 2005)