Observation of the Deexcitation of the ${}^{229\mathbf{m}}\mathrm{Th}$ Nuclear Isomer

The ${}^{229}\mathrm{Th}$ nucleus possesses the lowest-energy nuclear isomeric state. Two widely accepted indirect measurements of the transition energy place it within reach of existing laser capabilities. Direct searches for the isomer deexcitation have proven elusive despite extensive effort over the past couple of decades. There is now a growing interest in finding this unique transition because of its potential applications in nuclear, atomic, condensed matter, and optical physics, quantum information, metrology, and cosmology, including the development of a new type of clock based on this nuclear transition. In this Letter we report the first direct observation of the deexcitation of the lowest-lying isomeric state in ${}^{229}\mathrm{Th}$. By collecting ${}^{229}\mathrm{Th}$ recoils following the alpha decay of ${}^{233}\mathrm{U}$ into ${\mathrm{MgF}}_2$ plates and measuring the subsequent light emission, we have isolated the isomer deexcitation and measured the transition’s half-life to be $6\pm 1\mathrm{h}$. Through comparison measurements with ${}^{235\mathrm{m}}\mathrm{U}$ isomer, we found that the observed ${}^{229\mathrm{m}}\mathrm{Th}$ deexcitation signal originates from photon emission rather than internal conversion electron emission. This discovery lays the groundwork for optical and laser spectroscopy of ${}^{229\mathrm{m}}\mathrm{Th}$ nuclear isomer and the development of a ${}^{229}\mathrm{Th}$ nuclear clock.

Figure 1

${}^{229}\mathrm{Th}$ nuclear energy levels and experimental details. (a) Ground state and lowest isomeric state of ${}^{229}\mathrm{Th}$ showing the two possible decay channels for the isomer: $\gamma$-decay and internal conversion (IC) electron emission. (b) Schematic of experimental setup used to measure the $\gamma$ decay of the ${}^{229\mathrm{m}}\mathrm{Th}$ isomer: ${}^{229}\mathrm{Th}$ recoils from ${}^{233}\mathrm{U}$ $\alpha$ decay are implanted into ${\mathrm{MgF}}_2$ plates for several hours. The ${}^{233}\mathrm{U}$ sources are then removed and the light emission is measured with a PMT. (c) Quantum efficiencies (as provided by the manufacturer) for Hamamatsu PMTs R8487 and R8486.

Figure 2

Results with R8487 PMT. Emission decay curve measured after an 80 min collection period (red dots) and the fit (blue line) using a ${}^{232}\mathrm{U}$ source. The data fit well to the function: $f(t)=A_0+A_{\mathrm{Pb}}(e^{-t/\tau }\mathrm{Pb}-0.684e^{-t/\tau }\mathrm{Bi})+A_{\mathrm{Ra}}{e}^{-t/\tau }\mathrm{Ra}$, where the lifetimes ($\tau =t_{1/2}/\ln 2$) of all radio nuclides are held fixed, and the amplitude ratio of ${}^{212}\mathrm{Bi}$ to ${}^{212}\mathrm{Pb}$ was calculated from the implantation time $t_i$ using a derived formula [22]: $-\frac{{\tau }_{\mathrm{Bi}}}{{\tau }_{\mathrm{Pb}}}(\frac{1-e^{-t_i/{\tau }_{\mathrm{Bi}}}}{1-e^{-t_i/{\tau }_{\mathrm{Pb}}}})$. Emission measured after 80 min (green dots) collection period using a chemically purified ${}^{233}\mathrm{U}$ source. The data show no ${}^{229}\mathrm{Th}$ isomer emission detected with PMT R8487.

Figure 3

Evidence of ${}^{229\mathrm{m}}\mathrm{Th}$ $\gamma$-ray emission. (a) Emission decay curves measured with PMT R8486 after a 3.5 h collection period using a $4.2\mu \mathrm{Ci}$ ${}^{233}\mathrm{U}$ source (red crosses) and with Mylar foils used as a recoil filter (blue and green dots are phosphorescence data before and after ion implantation, respectively). The PMT signal difference (with the phosphorescence amplitude scaled by 1.15 and then subtracted out) is shown in the inset as black data points. (b) Signal amplitude (red) for 3.5 h implantation period and ${}^{228}\mathrm{Th}$ activity in the ${}^{233}\mathrm{U}$ source (blue line) as a function of source age. (c) Signal half-life as a function of source age; linear fit to zero source age yields a half-life of $6\pm 1\mathrm{h}$. (d) The averaged PMT signal difference for two 3.5 h recoil ion implantation measurements using a $4.9\mu \mathrm{Ci}$ ${}^{239}\mathrm{Pu}$ source (red dots) and the temperature difference for this measurement (black line). The solid red line is an exponential fit with an amplitude of $18\pm 20\mathrm{mHz}$ using a fixed 6 h half-life. For a fixed 26 min half-life, the fit amplitude is $-74\pm 50\mathrm{mHz}$.