Upper Bound for Neutron Emission from Sonoluminescing Bubbles in Deuterated Acetone

An experimental search for nuclear fusion inside imploding bubbles of degassed deuterated acetone at $0°\mathrm{C}$ driven by a 15 atm sound field and seeded with a neutron generator reveals an upper bound that is a factor of 10 000 less than the signal reported by Taleyarkhan et al. The strength of our upper bound is limited by the weakness of sonoluminescence, which we ascribe to the relatively high vapor pressure of acetone.

Figure 1

Block diagram of the apparatus used to study neutron generated cavitation in acetone-$d6$. Length, width, thickness, and curvature of the acoustic resonator and plungers were chosen to match blueprints provided by Taleyarkhan [29]. Accordingly, in early runs the top plunger was suspended with a thin wire so that it touched the free surface. In most runs, however, the top plunger was near to but not in contact with the free surface. This resulted in more stable acoustic modes and avoided excessive breakage. Metal press vacuum seals were used top and bottom. Wax tape and vacuum grease were used to seal the upper and lower glass pieces. Acetone was degassed in situ with the sound field on using a rotary vein vacuum pump. During the last stage of the degas procedure, the PNG generated cavitation in the fluid. Cooling and degassing could lower liquid level by 4 cm. Depending on the gas concentration, neutron seeded cavitation could result in streamers or popping. All of the data presented herein are from the popping mode.

Figure 2

Sequence of events leading to neutron seeded cavitation in a fluid driven by a 15 atm sound field. The top trace labeled PNG is from one of the two neutron detector channels. The starred trace is the pill microphone signal. The solid black line is the square rooted and normalized light scattering signal. The histogram of SL events was obtained with a bias of 1200 V applied to the SL PMT and corresponds to 1000 triggers, at this gain [single photon level $\sim 50\mathrm{mV}$] after pulsing with a 300 ns delay could reach 50% and was rejected out in software. The histogram width is due to the width of the PNG pulse as well as the range of times when the sound field is sufficiently negative to expand a seeded nanobubble. Light coincident with the PNG is due to Cherenkov radiation from electrons which are excited by gammas from nuclear interactions with 14 MeV neutrons. We ascribe the bounce in bubble radius, after the main bubble collapse, as being due to vapor trapped inside the bubble.

Figure 3

Histogram of 58 000 SL coincidences for 2 PMT’s excited above single photoelectron level, for 1500 8 ms sweeps. For a single tube there were 175 000 events above background. The insets show an example of single sweep data for the two SL PMT’s (black and gray) on the scale of an acoustic cycle and within a single sound cycle.

Figure 4

Real time coincidence between neutron count and one SL PMT. The energy of the neutron scintillation corresponds to 582 keV $e.e.$ and occurs $9.6\mu \mathrm{s}$ after the turn-on of the PNG. We interpret this event as a PNG neutron coincident with light from Cherenkov radiation in the acetone due to the same PNG neutron pulse. The neutron trace shows in black the region used to calculate the tail area for the pulse shape discrimination. The inset shows SL events for a cylindrical resonator filled with filtered [$0.2\mu \mathrm{m}$] degassed acetone-$d6$, pressure released to the vapor pressure, run at 15 atm sound, and seeded by an AmBe source.

Figure 5

Neutron counts received during data acquisition for 2300 neutron-induced cavitation events at low PNG setting for 0 °C degassed acetone-$d6$ excited with a 15 atm sound wave. The left inset shows the temporal structure of the PNG pulse. The right inset shows the gamma/PNG discrimination for all of the data in this figure. The upper branch is due to neutrons, and the lower branch is due to gammas. The experimental acceptance window for 2.5 MeV neutrons is indicated on the upper branch. The accuracy of the pulse shape discrimination is affected by overlap of the tail and rise of successive closely spaced neutron-induced scintillation events as well as the dc offsets due to cross talk with the high acoustic drive [30]. Events with an asterisk correspond to background neutrons with energy higher than the 2.45 MeV acceptance window.