Sonoluminescence: How a Trapped Bubble Turns Sound Into Picosecond Flashes of Light

The Setup: A Bubble Trapped at a Pressure Antinode

Sonoluminescence starts with a standing acoustic wave. A piezoelectric transducer glued to a flask of degassed water drives a bulk resonance — typically the (1,1,1) or a spherical mode near 20–40 kHz. At the pressure antinode (usually the flask center) the acoustic pressure oscillates as p_a(t) = P_a·cos(ωt), with amplitude P_a ≈ 1.2–1.5 atm superimposed on the 1 atm ambient. A small air bubble, once injected, is held there by the primary Bjerknes force, the time-averaged radiation force ⟨F⟩ = −V(t)∇p that pushes a small bubble toward a pressure antinode.

Over one 38 μs acoustic period the bubble breathes wildly. During the negative-pressure half-cycle the ambient pushes below the vapor pressure and the bubble expands from an equilibrium radius R₀ ≈ 4–5 μm out to a maximum R_max ≈ 40–50 μm — a tenfold radius growth, a thousandfold volume increase. Then the pressure swings positive, the overstretched bubble becomes unstable, and it collapses catastrophically. The single flash of light is emitted at the very bottom of that collapse, when the wall velocity peaks and the trapped gas is crushed to minimum size.

The Governing Equation: Rayleigh–Plesset Dynamics

The radius R(t) of a spherical bubble in an incompressible liquid obeys the Rayleigh–Plesset equation, derived by matching the liquid's kinetic energy to the pressure work at the wall:

• ρ(R R̈ + (3/2)Ṙ²) = p_B(t) − p∞(t) − 2σ/R − 4μṘ/R
• ρ ≈ 1000 kg/m³ is the water density; σ ≈ 0.072 N/m is the surface tension; μ ≈ 1.0×10⁻³ Pa·s is the dynamic viscosity.
• p_B is the pressure the gas exerts on the wall; p∞(t) = p₀ + P_a·cos(ωt) is the far-field driving pressure.

The gas pressure follows an approximately adiabatic law, p_B = (p₀ + 2σ/R₀)·[(R₀³ − h³)/(R³ − h³)]^γ, where h is a van der Waals hard-core radius (the gas cannot be compressed below it) and γ ≈ 5/3 for a monatomic-dominated interior. The two nonlinear inertial terms, R R̈ and (3/2)Ṙ², are the crux: as R → 0 during collapse, mass conservation demands the wall accelerate inward, and the velocity Ṙ can exceed the sound speed of water (1482 m/s). At that point the incompressible assumption fails, launching a converging shock and a compression heating pulse — the moment of light emission.

Why the Collapse Concentrates Energy So Extremely

The physics is spherical implosion. In Rayleigh's original 1917 analysis of a collapsing empty cavity, energy conservation gives the wall speed as the cavity shrinks:

• Ṙ² = (2/3)(p∞/ρ)·[(R_max/R)³ − 1]
• As R/R_max → 0, Ṙ diverges like R⁻³ᐟ² — the wall accelerates without bound (until gas pressure or compressibility intervenes).

Because kinetic energy per unit volume scales as ½ρṘ², and that energy is delivered to a shrinking gas core, the energy density skyrockets. A bubble that gathered acoustic energy over a ~40 μs, ~50 μm-radius expansion dumps it into a ~0.5 μm core in under a nanosecond. The compression ratio R_max/R_min can exceed 100 in radius — a factor of >10⁶ in volume — so the internal energy density rises by a comparable factor. The acoustic wavelength in water at 26.5 kHz is λ = c/f ≈ 1482/26,500 ≈ 5.6 cm; the emitting region is ~0.5 μm. The bubble is a geometric funnel converting a wave field 100,000× larger than itself into a point-like hot spot — an energy-focusing factor around 10¹².

How Hot, How Bright, How Fast: The Numbers

The emitted light is a brief, roughly blackbody-like burst. Key measured quantities for a typical air-in-water SBSL bubble:

• Flash duration: 50–350 ps, measured with time-correlated single-photon counting (Gompf, Putterman, et al., late 1990s). The pulse width is essentially independent of wavelength across the visible — a signature of thermal, not chemical, emission.
• Photons per flash: roughly 10⁵–10⁶, peaking in the near-UV; the spectrum rises toward shorter wavelengths and is cut off below ~200 nm by water's absorption, so the true peak (and thus temperature) is partly hidden.
• Inferred temperature: fitting the spectrum to a blackbody or bremsstrahlung/recombination model gives 10,000–20,000 K; noble-gas doping and lower drive can push spectroscopic estimates toward 6,000–15,000 K, while some plasma models argue the very core briefly reaches far higher.
• Power during the flash: ~10⁵–10⁶ photons × ~5 eV ≈ 10⁻¹³ J in ~100 ps ≈ order 1 mW instantaneous, from a source the size of a bacterium.

The Mach number of the wall, M = Ṙ/c_water, reaches order 1–4 at collapse; the interior gas is compressed enough to partially ionize, forming a transient weakly-ionized plasma. The line-free smoothness of SBSL spectra is thought to arise because this dense hot plasma is optically thick and its emission lines are pressure-broadened into a continuum.

The Controlling Variables: Argon Rectification and Ambient Gas

Stable, bright SBSL is finicky, and the decisive variable is dissolved gas content. Air-saturated water gives weak, unstable emission; the recipe that works is water degassed to ~20% of air saturation. The reason is the argon rectification hypothesis (Lohse, Hilgenfeldt, et al., 1996): during each violent collapse the reactive components of air — N₂ and O₂ — undergo high-temperature chemistry, forming NOₓ and other species that dissolve into the surrounding water. Over thousands of cycles the bubble is chemically purged of everything except its ~1% argon, an inert monatomic gas that survives. The stable SBSL bubble is therefore essentially a pure-argon bubble, which is why replacing air with pure noble gases dramatically changes the intensity.

Other levers:

• Drive amplitude P_a: too low and there's no collapse; too high and the bubble becomes shape-unstable (Rayleigh–Taylor and parametric surface modes) and pinches off. There is a narrow ~1.2–1.5 atm window.
• Temperature: cooling the water dramatically brightens SBSL — near 0 °C the light can be ~100× brighter than at room temperature, because lower vapor pressure means less energy-sapping water vapor trapped in the collapsing bubble.
• Diffusive stability: the bubble must sit at a fixed point where gas leaking out during expansion balances gas rectified in — this pins R₀ to a few microns.

History, Manifestations, and Cousin Phenomena

Multi-bubble sonoluminescence was discovered accidentally in 1934 by H. Frenzel and H. Schultes in Cologne, who were experimenting with ultrasonic sonar in a photographic developing tank and found their films fogged by light emitted from cavitating water. For decades this remained a curiosity of chaotic cavitation clouds. The field was transformed in 1990 when Felipe Gaitan (working with Lawrence Crum) isolated and stabilized a single bubble, producing a clock-stable point source that could be studied one flash at a time — turning a messy effect into precision physics.

Related and derived phenomena:

• Cavitation damage: the same supersonic implosions that emit light also erode ship propellers and pump impellers — sonoluminescence is the optical signature of destructive cavitation.
• Sonochemistry: the transient hot spots drive radical chemistry (splitting H₂O into ·OH and ·H), used industrially for cleaning, emulsification, and pollutant breakdown.
• Snapping shrimp: the pistol shrimp collapses a cavitation bubble with its claw fast enough to stun prey — and the collapse emits a faint sonoluminescent flash ("shrimpoluminescence").
• The bubble-fusion controversy: a 2002 claim (Taleyarkhan et al.) that acoustic cavitation in deuterated acetone produced neutrons from D–D fusion was widely investigated and never independently confirmed; mainstream SBSL temperatures (~10⁴ K) are far below the ~10⁷ K needed for fusion.

Subtleties and Common Misconceptions

It is not friction or triboluminescence. The light does not come from the water "rubbing" — it comes from compressive heating of the trapped gas, an adiabatic implosion, followed by thermal/plasma emission. The bubble contents, not the liquid, glow.

The interior is not a simple uniform blackbody. Because water absorbs below ~200 nm, we only see the tail of the spectrum, and fits are model-dependent: a blackbody fit, a thermal-bremsstrahlung fit, and a recombination-continuum fit give different "temperatures." Quoting a single number hides that the emitting plasma has steep gradients and lasts only ~100 ps — the peak core temperature may exceed the spectral-fit value while the outer shell is cooler.

The flash is astonishingly short and stable. SBSL flashes recur with jitter of only tens of nanoseconds, synchronized to the acoustic drive — the bubble is a better clock than most electronic oscillators of its era. The sub-nanosecond width means the emission mechanism switches on and off faster than any chemical luminescence could, which is the strongest evidence for a thermal plasma origin rather than sonochemiluminescence.

It is a room-temperature, tabletop phenomenon. No lasers, no high voltage, no vacuum chamber — just a flask, a transducer, degassed water, and a signal generator. That such extreme conditions (Sun-surface temperatures, supersonic shocks, transient plasma) arise from a gentle sound wave is precisely what makes sonoluminescence a canonical demonstration of nonlinear energy focusing.

Multi-bubble (cavitation cloud) sonoluminescence vs. single-bubble sonoluminescence

Property	Multi-bubble (MBSL)	Single-bubble (SBSL)
First observed	Frenzel & Schultes, 1934	Gaitan & Crum, 1990
Number of bubbles	Transient cloud, thousands	One trapped, stable bubble
Drive amplitude	High (chaotic cavitation)	Moderate (1.2–1.5 atm)
Flash timing	Incoherent, jittery	Clock-stable, synchronous with sound
Spectrum	Broadband + emission lines (OH, etc.)	Smooth, line-free, blackbody-like
Inferred T	~5,000 K	10,000–20,000+ K