Evidence for a Plasma Inside a Sonoluminescing Bubble

Send a high-intensity ultrasound wave through a container of liquid, and it’s not surprising to find that the alternating cycles of acoustic compression and rarefaction create micron-sized bubbles in the liquid and cause them to successively expand and contract. What is mysterious is to see those bubbles emit light. Somehow, the energy dispersed in an acoustic wave-field becomes sufficiently concentrated to produce visible light.

Multiple-bubble sonoluminescence (MBSL) was first seen in the 1930s, but in recent decades researchers have learned to produce and control a stable single bubble. Single-bubble sonoluminescence (SBSL) has allowed them to study in detail the dynamics of bubble cavitation. Experiments soon revealed other remarkable features: The bursts of light are as short as a few tens of pico-seconds and the time between successive pulses can be synchronized to within a few parts in 10¹¹. (See the articles in Physics Today, by Lawrence A. Crum, September 1994, page 22 , and by Detlef Lohse, February 2003, page 36 )

Most of the current theoretical models ¹ of SBSL predict that the gas bubble will collapse very rapidly to an extremely small radius and that, under some circumstances, the gas within an inner core will form an opaque plasma. That picture has now been strengthened by a recent experiment done by David Flannigan and Kenneth Suslick at the University of Illinois, Urbana-Champaign. ² They provide evidence for the gas-phase light emission from ions, signaling the formation of a plasma.

The Illinois experiment also demonstrates the promise of studying SBSL with argon bubbles in concentrated sulfuric acid. The light emission, as seen in Figure 1, was 3000 times as intense as that from the well-studied system of argon in water. “It’s as bright as a light bulb,” remarked Lawrence Crum of the University of Washington, who had tried sulfuric acid in his own lab after learning about Flannigan and Suslick’s work.

Perhaps because of the increased intensity, Flannigan and Suslick found discrete spectral lines not seen in a water system. From those lines, they were able to make a firmer estimate than previously possible of the temperature outside the plasma core. Consistent with earlier estimates, the new experiment finds that the bubble temperature rises above 15 000 K, several times hotter than the surface of the Sun. No doubt, it’s hotter still in the bubble’s core, but the opacity prevents one from probing inside.

Theorist William Moss of Lawrence Livermore National Laboratory is excited to see theory and experiment pointing in the same direction. Not only is there now evidence for the long-anticipated plasma in SBSL, but there are also some signs that the light-emitting region within the bubble is as small as 200 nm in radius. ³ Moss and his Livermore colleagues had predicted in 1999 that the light emissions in a water system come from the surface of an optically thick region of about the same radius. ⁴

Sulfuric acid, not water

Flannigan and Suslick tried to form single bubbles in sulfuric acid and other liquids with low vapor pressures in the hope of getting brighter emissions than with water. A group from the Istituto Elettrotecnico Nazionale Galileo Ferraris in Turin, Italy, had already reported a few-fold increase in intensity using sulfuric acid. ⁵ The Illinois team was able to push that brightness much higher.

One reason to expect the increased brightness ⁶ is that sulfuric acid is more viscous than water. The higher viscosity helps keep the bubbles both stable and spherical as they oscillate. Furthermore, sulfuric acid’s low vapor pressure prevents most of the acid from evaporating into the bubble’s interior. Thus, the bubble should consist almost entirely of atoms of argon that had been dissolved in the sulfuric acid. Because Ar atoms have no vibrational or rotational degrees of freedom, most of the cavitation energy should go into kinetic energy. Additionally, the high viscosity and low vapor pressures somehow combine to let researchers drive bubbles at much higher acoustic pressures (above 5 bar) than is possible with bubbles formed in water.

Seth Putterman of UCLA is not convinced that the lower vapor pressure can fully account for the increased brightness. He thinks the brightness might have something to do with the bubble dynamics. Unlike the very stationary bubbles in most SBSL studies, the bubble in the Illinois experiment has a jittery motion.

In studies of SBSL using argon or xenon in water, the emission spectrum has been rather featureless. The temperature within those bubbles has been estimated by fitting the spectrum to a blackbody temperature, or sometimes to the spectral distribution expected from bremsstrahlung. Such estimates have not been very satisfactory because they are model dependent.

In MBSL, researchers routinely see both the continuum and discrete spectral lines. The spectral lines give a handle on the temperature because they correspond to chemical species that form only at high temperatures. By knowing the temperature dependence of the reaction rates that produce these species, experimenters have estimated the temperatures reached during cavity collapse to be around 5000 K.

For SBSL in sulfuric acid, Flannigan and Suslick found both the blackbody-like continuum typical of SBSL in water and a series of discrete spectral lines at longer wavelengths, as seen in Figure 2(a). The spectral lines, which are shown in Figure 2(b) with the underlying continuum subtracted, correspond to transitions between highly excited states of Ar. In particular, the lines correspond to jumps between the 4p states (13.1–13.5 eV above the ground state) and the 4s states (11.5–11.8 eV) of the Ar atom.

To populate those excited states requires collisions with high-energy particles, such as electrons. Such particles would most likely come from the high-energy tail of some Boltzmann distribution. By studying the relative population of excited atomic states, Flannigan and Suslick calculated the effective temperature of the bubble outside the core. The experimenters suspect that the Ar emissions stem from the region just outside the opaque plasma core; that’s where charged particles near the core’s boundary might collide with Ar atoms.

Although the temperatures calculated from the Ar atom emissions are in the same ballpark as those determined from the blackbody-like spectrum seen at shorter wavelengths, Suslick cautions against directly comparing the two. The emissions on which the temperatures are based do not necessarily originate at the same time in the bubble or from the same spatial region. Suslick is collaborating with Putterman on an experiment to pin down the respective emission times.

The temperatures determined from Ar emissions were hotter at higher values of the acoustic driving pressure. By 2.9 bar, the temperature had reached 15 000 K. At even higher pressures, the thermal broadening of the atomic lines hinders any estimate of the temperature.

Why are such atomic emissions not seen when Ar bubbles form in water? The blackbody temperatures of SBSL in water can be higher than those seen in the sulfuric-acid system. Thus, one might expect a plasma to form in a water system as well. Perhaps atomic emissions do occur, but they are blurred by thermal broadening. Or perhaps their appearance in sulfuric acid has to do with the jittery motion: Putterman and Suslick note that atomic emissions have not been seen in single bubbles that are more stationary.

Evidence for a plasma

Flannigan and Suslick argue that Ar atoms are unlikely to be kicked into the 4s and 4p levels by thermal processes. Rather, they say, it takes collisions with energetic charged particles. That implies the presence of a plasma. Even stronger evidence comes from the sighting of spectral lines corresponding to the excited state, O₂ ⁺. This species, Flannigan and Suslick assert, could have been formed only by collisions with highly energetic charged particles, and not by thermal processes. That’s because the dissociation energy of the oxygen molecule is much less than its ionization energy. The sighting of O₂ ⁺ indicates that it must have been hit with a charged particle and ionized before it had a chance to dissociate.

Last year, Putterman and two colleagues at UCLA presented indirect evidence for the formation of a plasma in SBSL. They drove an isolated bubble of xenon in water at a very high frequency to produce such a small bubble that its core was no longer opaque. The group fit the emission spectrum with a thermal bremsstrahlung distribution and estimated a temperature of a million degrees. ⁷

The Illinois experiment is a first step toward exploring the inner core of sonoluminescing bubbles. There’s still a lot more to learn, such as how dense the plasma core is, how hot it gets, and how its opacity varies with other conditions. The challenge is for experimentalists to learn how to probe the inner core of optically opaque bubbles.