betaFeedback

WorksPhysics and Astronomy

229^{229}Th Nuclear Spectroscopy in an Opaque Material: Laser-Based Conversion Electron Mössbauer Spectroscopy of 229^{229}ThO2_2

Abstract

Here, we report the first demonstration of laser-induced conversion electron Mössbauer spectroscopy of the 229^{229}Th nuclear isomeric state, which provides the ability to probe the nuclear transition in a material that is opaque to light resonant with the nuclear transition. Specifically, we excite the nuclear transition in a thin ThO2_2 sample whose band gap (∼\sim 6 eV) is considerably smaller than the nuclear isomeric state energy (8.4 eV). As a result, the excited nucleus can quickly decay by internal conversion, resulting in the ejection of electrons from the surface. By collecting these conversion electrons, nuclear spectroscopy can be recorded. Unlike fluorescence spectroscopy, this technique is compatible with materials whose work function is less than the nuclear transition energy, opening a wider class of systems to study. Further, because ThO2_2 can be made from spinless isotopes and the internal conversion decay process reduces the isomeric state lifetime to only ∼\sim10 μμs, allowing ∼\sim108^8 relative reduction in clock interrogation time, a conversion-electron-based nuclear clock could lead to a ∼\sim104^4 reduction in clock instability.

The recent observation [13, 14] of direct laser excitation of the 229^{229}Th isomeric state in high band gap crystals has, after almost 50 years of work, opened the door to a laser-accessible nuclear transition and driven rapid progress in the development of solid-state optical clocks. Already, the nuclear transition frequency has been compared to an atomic clock and its linewidth studied [17], and work has begun to optimize the clock performance by developing both a theoretical understanding of nuclear quenching mechanisms [12] and using them to shorten the clock interrogation cycle [18, 19]. Similarly, excitation of the nuclear transition in a vacuum ultraviolet (VUV) transmissive thin film has opened the door to integrated-photonic-based nuclear clocks and sensors [15]. Further, new high bandgap materials have been analyzed that could provide dramatic simplification of the clock by, e.g., doping 229^{229}Th into a nonlinear optical crystal [20] and, by using principles of molecular design, provide a system with the potential for performance orders of magnitude beyond any current or planned optical clock [4, 21].

Because the 229^{229}Th nucleus provides a highly-controllable system which can be deployed into various hosts, it is expected, in analogy to Mössbauer spectroscopy, that these same techniques can also be used as new probes of the solid-state chemical and nuclear environment. However, in all studies to date, nuclear excitation is detected by observing the resulting nuclear fluorescence, which requires the material band gap to be larger than the isomeric transition energy (>Eiso> E_{\mathrm{iso}}). This requirement severely limits the material hosts available for study and, therefore, it is highly desirable to extend the spectroscopy of the isomeric transition to low band gap (<Eiso< E_{\mathrm{iso}}) environments.

Conversion electron Mössbauer spectroscopy (CEMS) of 229Th{}^{229}\mathrm{Th} has been proposed as a method to extend the technique of laser-based nuclear spectroscopy to low band gap materials [22, 23]. CEMS utilizes the fact that if the nuclear energy is larger than the material band gap, the internal conversion (IC) relaxation process is possible by promoting an electron across the band gap [16, 24]. If this promoted IC electron originates in a shallow-enough state in the valence band, the electron can overcome the work function barrier and emerge from the material surface. By detecting these conversion electrons, nuclear spectroscopy can be recorded.

In addition to providing a means to study low band gap materials, CEMS could also allow significant improvements in nuclear clock stability as the IC decay rate is roughly 10810^{8} times faster than the radiative decay rate, enabling a much faster clock interrogation rate, leading to a projected clock instability of ∼10−18\sim10^{-18} at 1 s. Further, a CEMS-based nuclear clock could operate by simply reading out the CEMS photocurrent, providing a means to greatly simplify and miniaturize future nuclear clocks [25].

Here, we report the first demonstration of laser-based CEMS of any nucleus. Specifically, a VUV laser system is used to excite 229Th{}^{229}\mathrm{Th} nuclei in a thin sample of thoria (ThO2\mathrm{ThO}_{2}). 229ThO2{}^{229}\mathrm{ThO}_{2} was chosen as the first host as it has a low band gap (∼6\sim6 eV [26]) and is readily available as a stoichiometric Th compound. Conversion electrons from the 229ThO2{}^{229}\mathrm{ThO}_{2} are detected as a function of laser excitation energy providing the first laser-based CEMS nuclear spectrum, which determines the 229Th{}^{229}\mathrm{Th} nuclear transition frequency as 2020407.5(2)stat_{\mathrm{stat}}(30)sys_{\mathrm{sys}} GHz, consistent with previously reported values of the transition [13, 14, 15, 17], and whose width is consistent with the laser linewidth of our system reported in previous studies [14, 15]. Further, the IC lifetime is measured as ∼10\sim10 μ\mus, which is consistent with that measured in 229Th{}^{229}\mathrm{Th} implantation studies [27] and theoretical calculations of the IC lifetime described below.

In what follows we describe the 229ThO2{}^{229}\mathrm{ThO}_{2} target, experimental apparatus, and the spectroscopic protocol, before presenting the results of the CEMS spectroscopy. Next, we present calculations of the isomer shift and IC lifetime in 229ThO2{}^{229}\mathrm{ThO}_{2} using a theoretical approach that combines solid-state density functional theory and relativistic atomic many-body perturbation theory methods. Following these results, we use the measured lifetime of the internal conversion decay of the isomeric state in 229ThO2{}^{229}\mathrm{ThO}_{2} to evaluate the ideal performance of a 229ThO2{}^{229}\mathrm{ThO}_{2} conversion electron clock, and find its projected instability to be ∼2×10−18\sim2 \times10^{-18} at 1 s.

229ThO2{}^{229}\mathrm{ThO}_{2} Target and Apparatus

The CEMS target was constructed by electrodepositing 229ThO2{}^{229}\mathrm{ThO}_{2} onto a stainless steel disk (see Figure 1 for image of target, procured from Eckert & Ziegler Analytics, Inc.). Although other oxide and hydroxide compounds of Th may be present in the target, this paper assumes that the predominant chemical form of 229Th{}^{229}\mathrm{Th} in the target is 229ThO2{}^{229}\mathrm{ThO}_{2}, based on the method of target preparation (see SI for details). The diameter of the 229ThO2{}^{229}\mathrm{ThO}_{2} is ≈5\approx5 mm and its thickness is estimated from the width of the α\alpha-particle spectrum as ∼10\sim10 nm (Figure 1(c)), which is consistent with a measured total activity of ≈6.3\approx6.3 kBq. The VUV absorption of ThO2\mathrm{ThO}_{2} is estimated as α≈0.1\alpha\approx0.1 nm−1^{-1} [28], implying that 229Th{}^{229}\mathrm{Th} nuclei within a depth of ≈10\approx10 nm can be efficiently excited by the laser. In addition, the inelastic mean free path of electrons in solids at the eV energy scale is typically ∼10\sim10 nm [29], meaning that an IC electron generated at greater depths likely scatters before extraction from the surface. As such, further 229ThO2{}^{229}\mathrm{ThO}_{2} thickness primarily contributes to the background in the form of electrons produced via radioactive decays [30], and thus a thickness of ≈10\approx10 nm is likely optimal for CEMS.

Image not extracted

Figure 1. (a) Cut-away rendering of the 229^{229}ThO2_2 target mount. Arrows denote front aperture, window, target, and pyroelectric detector. (b) Rendering of the spectroscopy chamber. (Magenta arrow) Direction of VUV laser propagation. (Yellow arrows) IC electron trajectories from target to detection MCP. (Blue arrows) Background photoelectrons generated from VUV scatter diverted to secondary electrode. (Green arrow) Direction of static B-field used to guide IC electrons. (c) α\alpha-spectrum of the 229^{229}ThO2_2 target. (Inset) Photograph of 229^{229}ThO2_2 target used in this study. The peak labeled to 4845 keV corresponds to the dominant α\alpha-decay mode of 229^{229}Th. The other large peaks correspond to the α\alpha-decays of daughter nuclei. The 229^{229}Th peak had a FWHM of ∼\sim22 keV, consistent with energy loss through a ∼\sim10 nm sample, as estimated with SRIM [31].

CEMS spectroscopy is recorded by directing a tunable VUV laser onto the 229ThO2{}^{229}\mathrm{ThO}_{2} target. Briefly, VUV radiation was produced via resonance-enhanced four-wave mixing of two pulsed dye lasers in Xe gas. The frequency of the first pulsed dye laser, ωu\omega_{u}, was locked to the 5p6 1S0→5p5 (2P3/2∘) 6p 2[1/2]05p^{6}\,{}^{1}S_{0} \rightarrow5p^{5}\,(^{2}P^{\circ}_{3/2})\,6p\,^{2}[1/2]_{0} two-photon transition of Xe at ∼249.63\sim249.63 nm. The frequency of the second pulsed dye laser, ωv\omega_{v}, was scanned to produce VUV radiation in the Xe cell given by the difference mixing relation ω=2ωu−ωv\omega= 2\omega_{u} - \omega_{v}. All three laser beams then impinge off-axis with respect to a MgF2\mathrm{MgF}_{2} lens, whose chromatic dispersion is used with downstream pinholes to spatially filter the VUV beam and pass it towards the 229ThO2{}^{229}\mathrm{ThO}_{2} spectroscopy chamber. The laser system delivers 30 pulses per second to the target with a typical VUV pulse energy of ∼6−8\sim6 - 8 μ\muJ/pulse (see Refs. [14, 32] for details). For these experiments, N2\mathrm{N}_{2} gas is added to the Xe cell to quench a process leading to amplified spontaneous emission at 147 nm (see SI for details), which occurs on a timescale similar to IC decay and otherwise contributes to the photoelectron background.

The 229ThO2{}^{229}\mathrm{ThO}_{2} target is held in a custom mount (see Figure 1(a)) that provides the ability to electrically bias the sample, as well as VUV laser monitoring capabilities. Biasing of the target is necessary as the VUV laser causes a prompt burst of photoelectrons when it impinges on the 229^{229}ThO2_2 target, which will overwhelm the electron detector. These initial photoelectrons are suppressed for the first ∼\sim100 ns after the laser pulse by positively biasing the 229^{229}ThO2_2 target to 135 V, before the bias voltage is changed to -405 V to aid the extraction of IC electrons from the surface; at all times the front aperture of the mount is held at 100 V. The mount also houses a fused silica target that fluoresces under VUV illumination aiding alignment and a bare pyroelectric crystal to provide in situ monitoring of the VUV laser pulse energy. The target chamber is mounted on a combined lever arm-stepper motor system that allows it to be raised and lowered, causing the VUV laser to illuminate either the 229^{229}ThO2_2 target or laser energy monitor. To limit hydrocarbon buildup on the target and to limit scattering of the IC electrons due to background gases, the target chamber was ozone plasma cleaned prior to baking, and operated at a pressure of ∼\sim10−7^{-7} Pa.

In principle, IC electron detection can be accomplished by simply positively biasing, e.g., a multi-channel plate (MCP) detector held near the 229^{229}ThO2_2. However, it was found that scattered light from the VUV laser generated a large afterglow of photoelectrons originating at positions dispersed throughout the chamber which lasted tens of microseconds – presumably due to fluorescence of components within the beamline and chamber. To overcome this background, a combination of electric and magnetic fields were used that focused electrons originating from the 229^{229}ThO2_2 target onto a detection MCP, while diverting electrons generated elsewhere to another region of the chamber (see SI). Finally, as an additional layer of protection for the detection MCP, its front plate voltage was controlled so that it had no gain during the initial burst of photoelectrons from the target.

Spectroscopy

Using this apparatus, CEMS was performed by recording the number of detected electrons in the window between 6 μ\mus and 40 μ\mus after each laser pulse as a function of VUV laser wavelength; this window was chosen based on previous theoretical estimates [24] and observations of IC decay [33] following 233^{233}U decay. The electron signal, normalized by the laser pulse energy, is shown in Figure 2(a) as a function of frequency for a region ∼\sim100 GHz wide and centered on the 229^{229}Th nuclear transition energy. These data were recorded by firstly measuring the pulse energy, then detecting the total number of electrons emitted after 36,000 pulses (20 minutes) before measuring the pulse energy again. This process was repeated to record a laser CEMS spectrum. The data in Figure 2(a) is the average of four spectra with the vertical error bars representing the standard error. Due to differences in system alignment between campaigns, the photoelectron background varied slightly between different spectra. Therefore, a Lorentzian with a background fit was used to determine and subtract the photoelectron background before the data is combined. Further, because the laser frequency is not perfectly controlled these spectra are binned in frequency and the horizontal error bars are the standard deviation of the frequency. Each data point represents the average of at least 140,000 laser pulses.

Image not extracted

Figure 2. (a) The laser-based CEMS spectra of 229ThO2{}^{229}\mathrm{ThO}_2 is shown in black with a fit of a Lorentzian profile. Each point is an average of ≥140,000\geq140,000 laser pulses. A constant baseline of ≈75 e−\approx75\ e^- per laser pulse has been subtracted. The vertical error bars denote the standard error, while the horizontal error bars are the standard deviation of the laser frequency for the points in the bin. For comparison, the radiative decay spectrum of 229Th{}^{229}\mathrm{Th}:LiSrAlF6_6, taken from Ref. [14], is shown in blue. (b) The IC decay rate as a function of time is shown in black alongside a fit of a decaying exponential with a background, leading to an IC lifetime of 12.3(3) μs12.3(3)\ \mu\mathrm{s}.

A nonlinear, least squares fit of a Lorentzian to the spectrum yields a central value of 2020407.5(2)stat(30)sys2020407.5(2)_{\mathrm{stat}}(30)_{\mathrm{sys}} GHz and full-width-half-maximum linewidth of 12.4(4)12.4(4) GHz, both consistent with our observations in LiSrAlF6_6 [14] and ThF4_4 [15]. Here, statistical error is the 68% confidence interval and the systematic uncertainty is predominantly due to the accuracy of the wavemeter and likely conservative. On resonance, an average of 0.41(5) e/μJ0.41(5)\ e/\mu\mathrm{J} per laser pulse is detected. The expected signal can be estimated as ηeηcNe×e−ti/τIC\eta_e\eta_cN_e \times e^{-t_i/\tau_{\mathrm{IC}}}, where ηe\eta_e is the extraction of efficiency of IC electrons from the sample, ηc\eta_c is the collection efficiency of electrons emitted from the sample, NeN_e is the number of excited 229Th{}^{229}\mathrm{Th} nuclei, and tit_i is the time that electron counting begins after the laser pulse. Based on calibrations of the apparatus (see SI), the expected signal is found as 0.15−0.09+0.25 e−/μJ0.15^{+0.25}_{-0.09}\ e^-/\mu\mathrm{J}, in reasonable agreement with the recorded spectrum.

Using the same system, the lifetime of the IC decay is measured by comparing the time binned electron counts collected on resonance in the first 190 μs190\ \mu\mathrm{s} after laser excitation. Specifically, the average number of time binned electron counts per μJ\mu\mathrm{J} of laser energy is obtained for both the on and off (∼100\sim100 GHz detuning) resonant illumination periods. The average off resonance counts are then subtracted from the average on resonance counts to remove any photoelectron background, and the result is plotted in Figure 2(b). A background free nonlinear least-squares fit of a decaying exponential reveals an IC lifetime of 12.3(3) μs12.3(3)\ \mu\mathrm{s}, as shown in Figure 2(b). As discussed later, the IC decay rate is sensitive to the local chemical environment, so a range of decay rates is possible and this lifetime should be interpreted as an estimate.

Calculation of Isomer Shifts and Internal Conversion Lifetime in 229ThO2{}^{229}\mathrm{ThO}_2

The measured isomer energy is in agreement with the earlier spectroscopic experiments in three wide band gap hosts [13, 14, 15]. Theoretically, this is supported by the calculations [34] of isomer shifts for a large variety of 229Th{}^{229}\mathrm{Th} solid-state hosts and by our 229ThO2{}^{229}\mathrm{ThO}_2-specific calculations presented in the SI. For example, we expect that the isomer shift between bulk 229ThO2{}^{229}\mathrm{ThO}_2 and 229Th{}^{229}\mathrm{Th}:LiSrAlF6_6 is on the order of 100 MHz, well below the ∼3\sim3 GHz reported experimental resolution. Our calculations predict the nuclear transition frequency for bulk 229ThO2{}^{229}\mathrm{ThO}_2 to be 2,020,407,338(70) MHz.

In the IC process, the energy of the excited nucleus is transferred to the electrons. In a crystalline solid, that means promoting a valence band electron ∣vκσv⟩\lvert v\kappa\sigma_v\rangle into a conduction band state ∣cκσc⟩\lvert c\kappa\sigma_c\rangle and creating a hole in the valence band (Figure 3(a)). The process is mediated by the hyperfine interaction (HFI) WW which, for periodic lattices, conserves the electronic crystal momentum k\mathbf{k}, but can connect electronic states of different spin projections σv,c\sigma_{v,c}. We derived the rate for the described IC process (see SI) as

Image not extracted

Figure 3. (a) Feynman diagram of the internal conversion process. The interaction is mediated by a virtual photon exchange, corresponding to the hyperfine interaction WW. The crystal momentum k\mathbf{k} is conserved in this process. (b) Conventional unit cell of 229ThO2{}^{229}\mathrm{ThO}_2 with Th atoms in green and O atoms in red. (c) Th projected density of states (PDOS) computed with G0W0G_0W_0. The horizontal dashed line denote the Fermi level. (d) Th projected joint density of states (PJDOS) computed with G0W0G_0W_0. The vertical dashed line denotes the nuclear isomeric energy.

ΓIC=2πℏ12Ie+1∑MgMe∑σcσv∣Wcσcvσvg(k)∣2Gcv(ℏωnuc).(1)\Gamma_{\mathrm{IC}} = \frac{2\pi}{\hbar}\frac{1}{2I_e+1}\sum_{M_gM_e}\sum_{\sigma_c\sigma_v}\left|W^{g}_{c\sigma_c v\sigma_v}(\mathbf{k})\right|^2G_{cv}(\hbar\omega_{\mathrm{nuc}}). \tag*{(1)}

Here ∣e⟩\lvert e\rangle and ∣g⟩\lvert g\rangle are the isomer and the ground nuclear states with spins Ie,gI_{e,g} and magnetic quantum numbers Me,gM_{e,g}. The HFI WW connects the two nuclear states and the valence and conduction band electronic states. Gcv(ℏωnuc)G_{cv}(\hbar\omega_{\mathrm{nuc}}) is the conventional [35] Joint Density of States (JDOS) at the nuclear transition energy ℏωnuc\hbar\omega_{\mathrm{nuc}}. The bar denotes averaging over a surface of constant energy in k\mathbf{k}-space.

The same JDOS appears in the theory of electromagnetic interband absorption [35] and is clearly a highly material-dependent quantity. Therefore, to understand the physics of the IC process, and in particular its rate, we calculated the properties of 229ThO2{}^{229}\mathrm{ThO}_2 with periodic electronic structure theory (GWGW and DFT). The calculations used the G0W0G_0W_0 approximation for the electronic self-energy and the Bethe–Salpeter Equation (BSE) for optical properties (see SI for details). The fundamental band gap computed with G0W0G_0W_0 is 6.2 eV, in good agreement with experimental values [36, 37] (calculation details provided in the SI).

The absorption spectrum, computed with G0W0G_0W_0+BSE, agrees with the measurements in Ref. [28], supporting our estimation of α≈0.1 nm−1\alpha\approx0.1\ \mathrm{nm}^{-1}.

229ThO2{}^{229}\mathrm{ThO}_2 crystallizes in the fluorite structure (space group Fm3‾mFm\overline{3}m, #225), where each Th atom is coordinated by eight oxygen atoms in a cubic arrangement (Figure 3(b)). The Th4+\mathrm{Th}^{4+}–O2−\mathrm{O}^{2-} bonding is predominantly ionic with a modest covalent admixture arising from overlap of Th 6d/5f6d/5f and O 2p2p orbitals. That Th is in the +4+4 oxidation state may be verified from the projected density of states (PDOS) on Th, where the 5f5f and 6d6d components in the valence band are nearly empty. In contrast, the Th 5f5f and 6d6d valence orbitals form large peaks in the conduction band, indicating the transfer of an electron from an O2−\mathrm{O}^{2-} anion to form Th3+\mathrm{Th}^{3+}. Both the valence and conduction bands contain small contributions from Th-projected pp orbitals. As it turns out, the IC process primarily involves transferring electrons between these pp-orbital components.

In computing the HFI matrix element in Eq. (1), it is convenient to project the valence and conduction band states ∣vkσv⟩\lvert v\mathbf{k}\sigma_v\rangle and ∣ckσc⟩\lvert c\mathbf{k}\sigma_c\rangle onto the basis of Th atomic orbitals. The matrix element Wcσcvσvge(k)W^{ge}_{c\sigma_c v\sigma_v}(\mathbf{k}) may then be expressed as a sum over different HFI matrix elements connecting atomic orbitals of definite angular momenta, weighted by the expansion coefficients of ∣vkσv⟩\lvert v\mathbf{k}\sigma_v\rangle and ∣ckσc⟩\lvert c\mathbf{k}\sigma_c\rangle. To the lowest order, cross terms arising when squaring the expansion of Wcσcvσvge(k)W^{ge}_{c\sigma_c v\sigma_v}(\mathbf{k}) can be neglected. This allows the expression of the IC rate in terms of the nuclear ground state HFI constants AA and the “projected joint density of states” (PJDOS) in Th. The PJDOS is a JDOS weighted by the projections of the crystal electronic states onto the thorium atomic orbitals that allows us to describe the character of the delocalized crystal orbitals close to the thorium nucleus (see SI). The HFI AA constants for Th3+\mathrm{Th}^{3+} are known from experiments and high-precision relativistic atomic-structure calculations, see Ref. [12]. The PJDOS of transitions with the largest contributions to the IC rate are shown in Figure 3(c).

Using these values, we arrive at an estimate for the IC rate of ΓIC≈1.3×104 s−1\Gamma_{\mathrm{IC}} \approx1.3 \times10^{4}\,\mathrm{s}^{-1}, corresponding to an IC lifetime of ≈80 μs\approx80\,\mu\mathrm{s}.

Sources of the disparity between the measured and theoretical lifetimes could include known errors in the local projections of the plane wave orbitals done by VASP [38], small changes to the valence and conduction states expected with the addition of spin-orbit coupling in solid-state calculations, cross- and higher-order expansion terms neglected in our evaluation of the rate (1), and deviations of the sample from bulk 229ThO2{}^{229}\mathrm{ThO}_{2} due to both surface and self-radiation damage effects. Compared to the pristine 229ThO2{}^{229}\mathrm{ThO}_{2}, the IC rate for 229Th{}^{229}\mathrm{Th} adjacent to point defects can be enhanced by the electric quadrupole HFI contribution [39] due to symmetry breaking. Nevertheless, the order of magnitude agreement supports the physical picture that observed IC decay results from the nucleus relaxing via transferring an electron from an oxide anion into the Th 6p6p component of the conduction band.

Projected Clock Performance

In addition to its use as a new chemical probe, laser-based CEMS may allow the construction of a new type of nuclear clock with several advantages over crystal-based clocks [40]. Chief among these advantages are a greatly reduced clock interrogation time, as the IC decay rate is about 10810^{8} times faster than radiative decay, and the potential for clock readout by simply monitoring the current leaving the target, which could facilitate dramatic miniaturization.

For such a clock, assuming 229ThO2{}^{229}\mathrm{ThO}_{2} is produced from 16O{}^{16}\mathrm{O}, the largest sources of instability would be broadening due to magnetic dipole interactions between the 229Th{}^{229}\mathrm{Th} nuclei and lifetime broadening. Given that the Th-Th distance in ThO2_2 is 3.96 A˚3.96\,\text{\AA} [41], the expected Zeeman broadening due to to neighboring 229Th{}^{229}\mathrm{Th} nuclei is ∼10\sim10 Hz. Thus, the primary source of instability is the lifetime broadening of Γ≈2π×16\Gamma\approx2\pi\times16 kHz. Assuming a polished face of a ∼10\sim10 nm thick single crystal is prepared and realizes an electron extraction efficiency ηe≈0.5\eta_e \approx0.5 [29], and that the photoelectron background is eliminated by material purification and baffling, a 100 μW100\,\mu\mathrm{W} laser leads to a projected clock instability of ∼2×10−18\sim2 \times10^{-18} at 1 s averaging. Interestingly, it is possible that the electric current from the IC process could also be used as a form of readout, as it would generate a current of ∼300\sim300 nA on the μ\mus timescale, providing a simple means for clock operation and locking [25].

Discussion and outlook

With this first demonstration of laser-based CEMS, an entirely new class of materials is now compatible with laser nuclear spectroscopy. By implanting 229Th{}^{229}\mathrm{Th} into low band gap materials and observing both the lifetime of the internal conversion decay channel and the isomer shift of the transition, detailed information can be gathered about the local phononic, electronic, and nuclear structure. This in turn allows the isomeric transition to serve as a sensor of strain and temperature in the solid [11, 42]. Further, at the energy scale of the IC electrons, their inelastic mean free path is a sensitive function of their energy relative to the Fermi level [29]. In the future, it may be possible to relate the efficiency with which IC electrons emerge from a surface with both the surface quality and the band structure local to the implanted 229Th{}^{229}\mathrm{Th}. Looking further forward, off-resonant excitation of 229Th{}^{229}\mathrm{Th} nuclei through a combined nuclear and phononic transition in the solid could enable laser-based nuclear resonance vibrational spectroscopy [43, 44]. This could enable measurement of the local phonon density of states with sub-μ\mueV precision, considerably beyond what is possible with conventional Mössbauer spectroscopy [44].

Given our previous work in 229ThF4{}^{229}\mathrm{ThF}_4 films [15], it can be concluded that as the 229Th{}^{229}\mathrm{Th} is converted from the oxide to the fluoride state, the 229Th{}^{229}\mathrm{Th} isomer will convert from decaying by IC to decaying by the emission of a VUV photon. Thus, the relative ratio of radiative decay to conversion electron decay following laser excitation could provide a new battery of methods to characterize the chemical state of Th compounds, perhaps aiding studies of nuclear power generation [45].

Additionally, laser-based CEMS provides a new platform to realize a solid-state Th clock that benefits from the inherent ease of a production of stoichiometric Th compounds, a 10810^{8} reduction in clock interrogation cycle, and the potential for a current-based readout allowing simplification and miniaturization of future nuclear clocks.

Acknowledgement

This work was supported by NSF awards PHYS-2013011, PHY-2412869 and PHY-2207546, and ARO award W911NF-11-1-0369. PGT and KS acknowledge support by BaCaTeC (Grant 7 [2029-2]). ERH acknowledges institutional support by the NSF QLCI Award OMA-2016245.

This work used Bridges-2 at Pittsburgh Supercomputing Center through allocation PHY230110 from the Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support (ACCESS) program, which is supported by National Science Foundation grants #2138259, #2138286, #2138307, #2137603, and #2138296. DAR acknowledges support from the Institutional Computing Program at Los Alamos National Laboratory, via the Center for Integrated Nanotechnologies, a DOE BES user facility, for computational resources.

References

Supplemental Information

Expected Internal Conversion Signal

The incident VUV flux is attenuated, so that it follows φ0e−αz\varphi_0 e^{-\alpha z} where α\alpha is the VUV attenuation coefficient. The number of excited thorium atoms in a target of length LL should go as [46]

Ne≈46λ22πnThT∣n~∣2ΓLRe⁡[n~]φ0(1−e−αL)α×11+4(δΓL)2×(teτrad),(2)N_{\mathrm{e}} \approx\frac{4}{6}\frac{\lambda^2}{2\pi}\frac{n_{\mathrm{Th}}T|\tilde{n}|^2}{\Gamma_{\mathrm{L}}\operatorname{Re}[\tilde{n}]}\frac{\varphi_0(1-e^{-\alpha L})}{\alpha}\times\frac{1}{1+4\left(\frac{\delta}{\Gamma_{\mathrm{L}}}\right)^2}\times\left(\frac{t_{\mathrm{e}}}{\tau_{\mathrm{rad}}}\right), \tag*{(2)}

where λ\lambda is the vacuum transition wavelength, nThn_{\mathrm{Th}} is the density of 229Th^{229}\mathrm{Th} in the 229ThO2^{229}\mathrm{ThO}_2 target, Γrad=1/τrad\Gamma_{\mathrm{rad}}=1/\tau_{\mathrm{rad}} is the vacuum radiative decay rate, ΓL\Gamma_{\mathrm{L}} is the VUV laser bandwidth, ΓIC=1/τIC\Gamma_{\mathrm{IC}}=1/\tau_{\mathrm{IC}} is the IC decay rate, δ\delta is the laser detuning, φ0\varphi_0 is incident the laser photon flux, LL is the target thickness, TT is the transmission of the VUV laser into the target, and n~=n−iκ\tilde{n}=n-i\kappa is the complex index of refraction with κ=λα/4π\kappa=\lambda\alpha/4\pi. The target was produced using a 0.75:0.25 229Th:232Th^{229}\mathrm{Th}:^{232}\mathrm{Th} isotope mix from Oak Ridge National Lab, leading to an effective 229Th^{229}\mathrm{Th} density of nTh=0.75×2.28×1022 cm−3n_{\mathrm{Th}}=0.75\times2.28\times10^{22}\ \mathrm{cm}^{-3} [41].

ParameterValueRel. Err.
nThn_{\mathrm{Th}}1.71×1022 cm−31.71\times10^{22}\ \mathrm{cm}^{-3}0 (Ref. [41])
LL10 nm0.7*
ΓL\Gamma_{\mathrm{L}}2π×12 GHz2\pi\times12\ \mathrm{GHz}0.25
α\alpha0.1 nm−10.1\ \mathrm{nm}^{-1}0 (Ref. [28])
nn2.340 (Ref. [28])
tit_i6 μ\mus1e−41\mathrm{e}{-4}
τrad\tau_{\mathrm{rad}}1800 s0.07
τIC\tau_{\mathrm{IC}}10 μ\mus0.1
TηeT\eta_{\mathrm{e}}1×10−41\times10^{-4}0.7*
ηc\eta_{\mathrm{c}}0.70.35*

Table I. Parameter values and estimated relative errors used in calculation of expected IC signal. Relative uncertainties labeled by an asterisk (*) follow log-normal distribution and should be read as ln⁡(X/Xˉ)\ln(X/\bar{X}). 0 error entries correspond to assumed values from literature.

From the number of excited 229Th^{229}\mathrm{Th} nuclei, the number of detected IC electrons is given by

Ndet=ηeηcNe×e−ti/τIC(3)N_{\mathrm{det}}=\eta_{\mathrm{e}}\eta_{\mathrm{c}}N_{\mathrm{e}}\times e^{-t_i/\tau_{\mathrm{IC}}} \tag*{(3)}

where ηe\eta_{\mathrm{e}} is the extraction efficiency from the 229ThO2^{229}\mathrm{ThO}_2 target, ηc\eta_{\mathrm{c}} is the collection efficiency, and tit_i is the start of the acquisition time window. The extraction efficiency represents the probability that an IC electron is able to leave the 229ThO2^{229}\mathrm{ThO}_2 target, and combines many physical processes such as whether the electron is promoted high enough into the conduction band to overcome the work function barrier, whether the electron inelastically scatters, whether surface conditions are favorable, etc. Due to all of these confounding factors, theoretical calculation of ηe\eta_{\mathrm{e}} is difficult, and it must be estimated from experiment. To do this, we make the assumption that the probability for a photoelectron promoted by a VUV photon to leave the target is the same as an IC electron promoted by an excited 229Th^{229}\mathrm{Th} nucleus. By making this assumption we are able to use the efficiency with which photoelectrons are generated by our VUV laser system to obtain an estimate of TηeT\eta_{\mathrm{e}}. The collection efficiency ηc\eta_{\mathrm{c}} is simply the efficiency with which electrons that leave the target are detected on the MCP given the voltage biasing and magnetic field configuration used in the experiment. This can readily be determined by measuring the ratio of the number of photoelectrons collected on the MCP front plate relative to the number of photoelectrons leaving the target.

Using the values listed in Table I, the expected on-resonance IC signal was estimated to be Ndet/(ℏω0φ0te)=0.15−0.09+0.25 e−/μJ/shotN_{\mathrm{det}}/(\hbar\omega_0\varphi_0t_{\mathrm{e}})=0.15^{+0.25}_{-0.09}\ \mathrm{e}^{-}/\mu\mathrm{J}/\mathrm{shot}.

Quenching of ASE in Xe Four-wave Mixing

The four-wave mixing process can produce a beam of amplified spontaneous emission (ASE) along the beam axis, complicating detection. The origin of the ASE is resonant excitation by the 249.6 nm laser of the 5p6 1S0→5p5(2P3/2∘)6p2 [1/2]05p^6\,{}^{1}S_0\rightarrow5p^5\left({}^{2}P^{\circ}_{3/2}\right)6p^2\,[1/2]_0 transition. Any Xe population not participating in the four-wave mixing process will be left in the excited two-photon state. This 5p5(2P3/2∘)6p2 [1/2]05p^5\left({}^{2}P^{\circ}_{3/2}\right)6p^2\,[1/2]_0 (abbreviated as 6p2 [1/2]06p^2\,[1/2]_0) state will then decay to the 5p5(2P3/2∘)6s2[3/2]∘5p^{5}\left({}^{2}P^{\circ}_{3/2}\right)6s^{2}\left[3/2\right]^{\circ} (abbreviated as 6s2[3/2]∘6s^{2}\left[3/2\right]^{\circ}) by 828 nm emission in ≈30\approx30 ns, and the 6s2[3/2]∘6s^{2}\left[3/2\right]^{\circ} state will decay to the 1S0{}^{1}S_{0} ground state by emission of a 147 nm photon in ≈3.7\approx3.7 ns. If the Xe pressure is in the several hundred Pa range, as it is for efficient four-wave mixing, there are enough nearby Xe atoms that they may re-absorb these 828 nm and 147 nm photons. Effectively, this leads to radiation trapping which extends the effective fluorescence lifetime of the Xe spontaneous emission [47]. At the same time the 828/147 nm spontaneous emission will experience gain as it stimulates Xe in the excited 6p[1/2]06p\left[1/2\right]_{0} and 6s[3/2]1∘6s\left[3/2\right]^{\circ}_{1} to emit, yielding amplified spontaneous emission (ASE). This gain will be highly directional, since the excited Xe will essentially lie in a column defined by the propagation of the 249.6 nm pulse of the pump laser. This interplay between radiation trapping and ASE will then yield bi-directional emission from the Xe cell along the pumping axis, which will have a timescale much longer than the spontaneous emission lifetime of the excited states [48].

To mitigate this effect, we introduce N2_2 gas in the Xe, which quenches the excited state Xe population via collision.

Simulation of Electron Trajectories

SIMION simulations were carried out in order to determine a combination of voltage biasing and static magnetic field that would guide IC electrons to our detector while diverting background photoelectrons to a secondary electrode. The voltage biasing chosen during the IC observation period was to have the target at -405 V, the front aperture of the target mount at +135V, the front of the detection MCP at +140 V, and the secondary electrode at +2500V. The magnetic field was set to ≈3\approx3-5 G. As can be seen in Figure 4(a), under the voltage biasing described there is a ”saddle” in the potential through which the IC electrons that have been accelerated by our target mount system can be made to travel through by the magnetic field. Meanwhile, as can be seen in Figure 4(b), electrons that are generated randomly throughout the chamber with ∼8\sim8 eV of kinetic energy either crash into the chamber wall or fall into the sacrificial electrode.

Image not extracted

Figure 4. SIMION simulations of electron trajectories under the voltage biasing and static magnetic field used in the experiment. (a) Trajectories of IC electrons from the target region. The magnetic field bends them towards the detection MCP through a saddle in the electric potential. (b) The trajectories of photoelectrons generated by VUV scattered light at random locations in the chamber. Most either crash into the chamber walls or end up falling into the electrode biased at +2500V.

Preparation of the Target

The target used in this study was prepared in an electrodeposition buffer solution of NaOH and sulfuric acid. 229{}^{229}Th material was deposited onto the stainless steel disk until target activity was reached, whereupon a quench was performed with ammonium hydroxide. The disk was then heated overnight at 200-300 °C in ambient atmosphere.

While the electrodeposited target is likely a mixture of Th hydroxides, oxides, and metal impurities, the use of low impurity (>99.99%>99.99\% grade) chemicals during electrodeposition and heat treatment in air lead us to assume it is predominantly Th in an oxidized state. As the most stable oxide form, we assume the target is predominately 229{}^{229}ThO2_2.

Internal Conversion Rate derivation

We consider the IC processes in an insulator, when the bandgap Δ\Delta is smaller than the nuclear isomer energy ωnuc\omega_{\mathrm{nuc}}. In the IC process, the nuclear excitation is transferred to the electrons, spawning a particle-hole pair, with a valence band electron promoted into the conduction band leaving behind a hole in the valence band, see Figure 5.

Image not extracted

Figure 5. Feynman diagram of the internal conversion process. Single black strokes represent the electron lines, while the double blue ones denote the nuclear lines. The interaction is mediated by a virtual photon exchange, corresponding to the hyperfine interaction WW. The crystal momentum k\mathbf{k} is conserved in this process.

The 229Th^{229}\mathrm{Th} nuclear subsystem is modeled as two distinct energy levels: the ground state ∣g⟩\lvert g\rangle with nuclear spin Ig=5/2I_g = 5/2 and the excited (isomeric) state ∣e⟩\lvert e\rangle with Ie=3/2I_e = 3/2, separated by the energy gap ωnuc\omega_{\mathrm{nuc}}.

The IC process is mediated by the hyperfine interaction (HFI). In a crystal, containing N\mathcal{N} 229Th^{229}\mathrm{Th} nuclei with one 229Th^{229}\mathrm{Th} per unit cell, HFI reads

W(r)=∑ν=1NM(Rν)⋅T(r−Rν),(4)W(\mathbf{r}) = \sum_{\nu=1}^{\mathcal{N}} \mathbf{M}(\mathbf{R}_{\nu}) \cdot\mathcal{T}(\mathbf{r} - \mathbf{R}_{\nu}), \tag*{(4)}

where we sum over unit cells, M(Rν)\mathbf{M}(\mathbf{R}_{\nu}) is a 229Th^{229}\mathrm{Th} nuclear magnetic moment operator and the 229Th^{229}\mathrm{Th}-centered T\mathcal{T} is a rank-1 tensor acting on the electronic degrees of freedom. The nuclear electric-quadrupolar and higher-rank HFI interactions can be added in a similar fashion. For 229ThO2^{229}\mathrm{ThO}_{2}, due to the cubic symmetry, the electric-quadrupole contribution vanishes.

To compute the IC rate, we use the Fermi’s golden rule. The initial state of the electron subsystem is a fully-occupied valence band ∣0~⟩\lvert\tilde{0}\rangle, and the final state is the particle-hole excitation ackσc†avqσv∣0~⟩a^{\dagger}_{c\mathbf{k}\sigma_c}a_{v\mathbf{q}\sigma_v}\lvert\tilde{0}\rangle with energy εck−εvq\varepsilon_{c\mathbf{k}}-\varepsilon_{v\mathbf{q}}, where εck\varepsilon_{c\mathbf{k}} and εvq\varepsilon_{v\mathbf{q}} are the band functions of the conduction (c) and the valence (v) bands. Then, the final state quantum numbers are spanned by crystal momenta k\mathbf{k}, q\mathbf{q} and two electron spin projections. We also sum over the nuclear ground state magnetic quantum numbers MgM_g and average over the nuclear isomer state magnetic quantum numbers MeM_e,

ΓIC(N)=2πℏ12Ie+1∑MgMe∑σcσv∫BZd3k∫BZd3q∣⟨ckσc∣Wge∣vqσv⟩∣2δ(εck−εvq−ωnuc).(5)\Gamma_{\mathrm{IC}}^{(\mathcal{N})} = \frac{2\pi}{\hbar}\frac{1}{2I_e+1}\sum_{M_g M_e}\sum_{\sigma_c \sigma_v}\int_{\mathrm{BZ}} d^3 k \int_{\mathrm{BZ}} d^3 q \left|\langle c\mathbf{k}\sigma_c|W^{ge}|v\mathbf{q}\sigma_v\rangle\right|^2 \delta(\varepsilon_{c\mathbf{k}}-\varepsilon_{v\mathbf{q}}-\omega_{\mathrm{nuc}}). \tag*{(5)}

The integrations are carried over the Brillouin zone (BZ). We use conventions of Ref. [35] in our derivation. The HFI, Eq. (4) is cell-periodic, and its matrix element ⟨ckσc∣Wge∣vqσv⟩\langle c\mathbf{k}\sigma_c|W^{ge}|v\mathbf{q}\sigma_v\rangle between Bloch functions can be reduced from an integration over the entire crystal to an integration over a single unit cell. The result reads

⟨ckσc∣Wge∣vqσv⟩=δ(k−q)Wcσcvσvge(k)(6)\langle c\mathbf{k}\sigma_c|W^{ge}|v\mathbf{q}\sigma_v\rangle= \delta(\mathbf{k}-\mathbf{q})W^{ge}_{c\sigma_c v\sigma_v}(\mathbf{k}) \tag*{(6)}

with

Wcσcvσvge(k)≡(2π)3Ω⟨g∣M∣e⟩⋅∫Ωd3r uck∗(r)χσc†T(r)χσvuvk(r).(7)W^{ge}_{c\sigma_c v\sigma_v}(\mathbf{k}) \equiv\frac{(2\pi)^3}{\Omega}\langle g|\mathbf{M}|e\rangle\cdot\int_{\Omega} d^3r\,u^*_{c\mathbf{k}}(\mathbf{r})\boldsymbol{\chi}^{\dagger}_{\sigma_c}\mathcal{T}(\mathbf{r})\boldsymbol{\chi}_{\sigma_v}u_{v\mathbf{k}}(\mathbf{r}). \tag*{(7)}

Here Ω\Omega is the unit cell volume, u…(r)u_{\ldots}(\mathbf{r}) are cell-periodic envelopes of Bloch functions [35], and χ\chi are the conventional electron spinors. Note that the cell-periodicity imposes conservation of crystal momentum, k=q\mathbf{k}=\mathbf{q}.

Now with this matrix element, we evaluate the IC rate.

ΓIC(N)=2πℏ12Ie+1∑MgMe∑σcσv∫d3k∫d3q∣Wcσcvσvge(k)∣2δ(k−q)δ(k−q)δ(εck−εvq−ωnuc)(8)\Gamma_{\mathrm{IC}}^{(\mathcal{N})} = \frac{2\pi}{\hbar}\frac{1}{2I_e+1}\sum_{M_g M_e}\sum_{\sigma_c \sigma_v}\int d^3 k \int d^3 q \left|W^{ge}_{c\sigma_c v\sigma_v}(\mathbf{k})\right|^2\delta(\mathbf{k}-\mathbf{q})\delta(\mathbf{k}-\mathbf{q})\delta(\varepsilon_{c\mathbf{k}}-\varepsilon_{v\mathbf{q}}-\omega_{\mathrm{nuc}}) \tag*{(8)}

There is a product of two identical delta functions. q\mathbf{q} is replaced by k\mathbf{k} while integrating over q\mathbf{q}, but we encounter δ(0)=lim⁡Δqx→0δ(Δqx)\delta(0)=\lim_{\Delta q_x\to0}\delta(\Delta q_x). Using the following identity for each component of Δq\Delta\mathbf{q}:

lim⁡Δqx→0δ(Δqx)=lim⁡Lx→∞lim⁡Δqx→012π∫−Lx/2Lx/2eiΔqxx dx=Lx2π,\lim_{\Delta q_x\to0}\delta(\Delta q_x)=\lim_{L_x\to\infty}\lim_{\Delta q_x\to0}\frac{1}{2\pi}\int_{-L_x/2}^{L_x/2}e^{i\Delta q_x x}\,dx=\frac{L_x}{2\pi},

where LxL_x is the crystal size in the xx-direction, we can show that

lim⁡Δq→0δ(Δq)=Vxtal(2π)3,\lim_{\Delta\mathbf{q}\to0}\delta(\Delta\mathbf{q})=\frac{V_{\mathrm{xtal}}}{(2\pi)^3},

where the crystal volume Vxtal=LxLyLzV_{\mathrm{xtal}}=L_xL_yL_z. This is similar to the formal time-domain limit in deriving the Fermi’s golden rule, see, e.g., p. 72 of Ref. [49].

Thereby, the IC rate per crystal volume

ΓIC(N)Vxtal=2πℏ1(2π)312Ie+1∑MgMe∑σcσv∫BZd3k∣Wcσcvσvge(k)∣2δ(εck−εvk−ℏωnuc)(9)\frac{\Gamma_{\mathrm{IC}}^{(N)}}{V_{\mathrm{xtal}}}=\frac{2\pi}{\hbar}\frac{1}{(2\pi)^3}\frac{1}{2I_e+1}\sum_{M_gM_e}\sum_{\sigma_c\sigma_v}\int_{\mathrm{BZ}}d^3k\left|W_{\mathrm{c}\sigma_c\mathrm{v}\sigma_v}^{ge}(\mathbf{k})\right|^2\delta(\varepsilon_{c\mathbf{k}}-\varepsilon_{v\mathbf{k}}-\hbar\omega_{\mathrm{nuc}}) \tag*{(9)}

The assumption made during the derivation was that all NN 229Th{}^{229}\mathrm{Th} nuclei were initially in the excited (isomer) state. Consider an ensemble of quantum emitters, with N(t)N(t) being the number of emitters in the excited state at time tt. With a single emitter decay rate γ\gamma and

dNdt=−γN(t)⇒N(t)=N(0)exp⁡{−γt}⇒Ensemble decay rate Γ=−dNdt=γN(0)exp⁡{−γt}.(10)\begin{aligned} \frac{dN}{dt}=-\gamma N(t)\Rightarrow N(t)=N(0)\exp\{-\gamma t\} \\ &\Rightarrow\text{Ensemble decay rate }\Gamma=-\frac{dN}{dt}=\gamma N(0)\exp\{-\gamma t\}. \tag*{(10)} \end{aligned}

We see that the ensemble decay rate Γ\Gamma at t=0t=0 is γN(0)\gamma N(0). The experimentally relevant excited state population, however, decays as exp⁡{−γt}\exp\{-\gamma t\}. Based on this discussion, we define the experimentally measured IC decay rate

ΓIC=ΓIC(N)/N=ΩΓIC(N)/Vxtal.(11)\Gamma_{\mathrm{IC}}=\Gamma_{\mathrm{IC}}^{(N)}/N=\Omega\Gamma_{\mathrm{IC}}^{(N)}/V_{\mathrm{xtal}}. \tag*{(11)}

Following the derivation of interband electromagnetic absorption rates [35], we define a surface SES_E of constant energy in k\mathbf{k}-space through an implicit relation εck−εvk=ℏωnuc\varepsilon_{c\mathbf{k}}-\varepsilon_{v\mathbf{k}}=\hbar\omega_{\mathrm{nuc}}. If the matrix element remains reasonably constant on SES_E, the rate simplifies to

ΓIC=1τIC=2πℏ12Ie+1∑MgMe∑σcσv∣Wcσcvσvge(k)‾∣2Gcv(ℏωnuc),(12)\Gamma_{\mathrm{IC}}=\frac{1}{\tau_{\mathrm{IC}}}=\frac{2\pi}{\hbar}\frac{1}{2I_e+1}\sum_{M_gM_e}\sum_{\sigma_c\sigma_v}\left|\overline{W_{\mathrm{c}\sigma_c\mathrm{v}\sigma_v}^{ge}(\mathbf{k})}\right|^2G_{cv}(\hbar\omega_{\mathrm{nuc}}), \tag*{(12)}

with the joint density of states (JDOS)

Gcv(E)=Ω(2π)3∫BZd3k δ(εck−εvk−E)=Ω(2π)3∫dSE∣∇k(εck−εvk)∣E.(13)G_{cv}(E)=\frac{\Omega}{(2\pi)^3}\int_{\mathrm{BZ}}d^3k\,\delta(\varepsilon_{c\mathbf{k}}-\varepsilon_{v\mathbf{k}}-E)=\frac{\Omega}{(2\pi)^3}\int\frac{dS_E}{\left|\nabla_{\mathbf{k}}(\varepsilon_{c\mathbf{k}}-\varepsilon_{v\mathbf{k}})\right|_E}. \tag*{(13)}

With this simplification, ∣Wcσcvσvge(k)‾∣2\left|\overline{W_{\mathrm{c}\sigma_c\mathrm{v}\sigma_v}^{ge}(\mathbf{k})}\right|^2 has the meaning of ∣Wcσcvσvge(k)∣2\left|W_{\mathrm{c}\sigma_c\mathrm{v}\sigma_v}^{ge}(\mathbf{k})\right|^2 averaged over the surface of constant energy SES_E.

This concludes the derivation of the IC rate, Eq. (1) of the main text.

As a side note, notice that the electromagnetic absorption coefficient at laser frequency ωnuc\omega_{\mathrm{nuc}} is proportional to the same JDOS Gcv(ℏωnuc)G_{cv}(\hbar\omega_{\mathrm{nuc}}). This can be used to back out the JDOS value from the laser absorption measurements at a frequency slightly detuned away from ωnuc\omega_{\mathrm{nuc}}.

To proceed with the computation of the hyperfine matrix element, we expand the functions uck∗(r)u_{c\mathbf{k}}^*(\mathbf{r}) in terms of the Th atomic states. Due to the short-range nature of the hyperfine interaction, relativistic effects play an important role. As a result, we use relativistic Dirac spinors for the Th atomic states. However, Eq. (7) is given in terms of nonrelativistic (or scalar relativistic [50]) two-component spinors. Here, for simplicity, we replace Eq. (7) with

Wcσcvσvge(k)≡(2π)3Ω⟨g∣M∣e⟩⋅∫Ωd3r uckσc†(r)T(r)uvkσv(r),(14)W_{\mathrm{c}\sigma_c\mathrm{v}\sigma_v}^{ge}(\mathbf{k})\equiv\frac{(2\pi)^3}{\Omega}\langle g|\mathbf{M}|e\rangle\cdot\int_{\Omega}d^3r\,u_{c\mathbf{k}\sigma_c}^{\dagger}(\mathbf{r})\mathcal{T}(\mathbf{r})u_{v\mathbf{k}\sigma_v}(\mathbf{r}), \tag*{(14)}

where unkσn(r)u_{n\mathbf{k}\sigma_n}(\mathbf{r}) is now a four-component Dirac spinor with momentum k\mathbf{k} and spin projection σn\sigma_n.

The crystal function uckσc(r)u_{c\mathbf{k}\sigma_c}(\mathbf{r}) (and similarly uvkσv(r)u_{v\mathbf{k}\sigma_v}(\mathbf{r})) may be expanded in terms of the Th atomic states of definite principle and angular momentum quantum numbers φnjlm(r)\varphi_{njlm}(\mathbf{r})

uckσn(r)=Ω(2π)3∑njlmanjlm(ckσn)φnjlm(r)+…,(15)u_{c\mathbf{k}\sigma_n}(\mathbf{r})=\sqrt{\frac{\Omega}{(2\pi)^3}}\sum_{njlm}a_{njlm}(\mathrm{c}\mathbf{k}\sigma_n)\varphi_{njlm}(\mathbf{r})+\ldots, \tag*{(15)}

where the ellipses denote contributions from O atomic states, which, due to their insignificant overlap with the Th nucleus, do not affect the HFI matrix element. Using the expansion (15), we may write

Wcσcvσvge(k)=∑njlm∑n′j′l′m′anjlm∗(ckσc)an′j′l′m′(vkσv)W~njlmn′j′l′m′ge,(16)W^{ge}_{c\sigma_c v\sigma_v}(\mathbf{k}) = \sum_{njlm} \sum_{n'j'l'm'} a^*_{njlm}(c\mathbf{k}\sigma_c)a_{n'j'l'm'}(v\mathbf{k}\sigma_v)\widetilde{W}^{ge}_{njlm n'j'l'm'}, \tag*{(16)}

where W~njlmn′j′l′m′ge=⟨g∣M∣e⟩⋅∫Ωd3r ϕnjlm∗(r)T(r)ϕn′j′l′m′(r)\widetilde{W}^{ge}_{njlm n'j'l'm'} = \langle g|\mathcal{M}|e\rangle\cdot\int_{\Omega} d^3r\,\phi^*_{njlm}(\mathbf{r})T(\mathbf{r})\phi_{n'j'l'm'}(\mathbf{r}).

Given the crystal functions and the atomic wavefunctions, the expansion coefficients anjlm(kσn)a_{njlm}(\mathbf{k}\sigma_n) may be computed and the total HFI matrix element obtained. Here, we make a simplifying assumption that only a few terms contribute significantly to the sum (16). Furthermore, we shall neglect the contribution cross terms in ∣Wcσcvσvge(k)∣2\left|W^{ge}_{c\sigma_c v\sigma_v}(\mathbf{k})\right|^2. With these simplifications, the IC rate may be written as

ΓIC≈πℏ12Ie+1∑MgMe∑njlm∑n′j′l′m′∣W~njlmn′j′l′m′ge∣2×∑σcσv2Ω(2π)3∫BZd3k ∣anjlm(ckσc)∣2∣an′j′l′m′(vkσv)∣2δ(ϵck−ϵvk−ℏωnuc).(17)\Gamma_{\mathrm{IC}} \approx\frac{\pi}{\hbar}\frac{1}{2I_e+1}\sum_{M_g M_e}\sum_{njlm}\sum_{n'j'l'm'}\left|\widetilde{W}^{ge}_{njlm n'j'l'm'}\right|^2 \times\sum_{\sigma_c\sigma_v}\frac{2\Omega}{(2\pi)^3}\int_{\mathrm{BZ}}d^3k\,\left|a_{njlm}(c\mathbf{k}\sigma_c)\right|^2\left|a_{n'j'l'm'}(v\mathbf{k}\sigma_v)\right|^2\delta(\epsilon_{ck}-\epsilon_{vk}-\hbar\omega_{\mathrm{nuc}}). \tag*{(17)}

Let us now consider the matrix element Wcσcvσvge(k)W^{ge}_{c\sigma_c v\sigma_v}(\mathbf{k}) in more details. Using the Wigner-Eckart theorem, we may write

W~njlmn′j′l′m′ge=∑ν=−11(−1)ν+Ig−Mg+j−m(Ig1Ie−MgνMe)(j1j′−m−νm′)⟨g∥M∥e⟩⟨njl∥T∥n′j′l′⟩.(18)\widetilde{W}^{ge}_{njlm n'j'l'm'}=\sum_{\nu=-1}^{1}(-1)^{\nu+I_g-M_g+j-m} \begin{pmatrix} I_g & 1 & I_e \\ -M_g & \nu& M_e \end{pmatrix} \begin{pmatrix} j & 1 & j' \\ -m & -\nu& m' \end{pmatrix} \langle g\|\mathcal{M}\|e\rangle\langle njl\|T\|n'j'l'\rangle. \tag*{(18)}

Here, ⟨g∥M∥e⟩≈0.84μN\langle g\|\mathcal{M}\|e\rangle\approx0.84\mu_N (μN\mu_N is the nuclear magneton) is the reduced matrix element of the nuclear magnetic moment operator (see, e.g. Ref. [12]), and ⟨njl∥T∥n′j′l′⟩\langle njl\|T\|n'j'l'\rangle is the reduced matrix element of the rank-1 electronic HFI tensor. The off-diagonal HFI matrix elements are at least two orders of magnitude smaller than the diagonal ones, so the most significant contributions to the IC rate come from terms with (njl)=(n′j′l′)(njl)=(n'j'l'). The reduced matrix element ⟨njl∥T∥njl⟩\langle njl\|T\|njl\rangle may be related to the ground state HFI constant AnjlA_{njl} via [51]

Anjl=μgIgj(2j)!(2j−1)!(2j+2)!⟨njl∥T∥njl⟩,(19)A_{njl}=\frac{\mu_g}{I_gj}\frac{(2j)!}{\sqrt{(2j-1)!(2j+2)!}}\langle njl\|T\|njl\rangle, \tag*{(19)}

where μg=0.360(7)μN\mu_g=0.360(7)\mu_N is the magnetic moment of the ground nuclear state [52]. With this, the summations over MgM_g and MeM_e in Eq. (17) may be carried out, giving

∑MgMe∣W~njlmn′j′l′m′ge∣2=13[∑ν=−11(j1j′−mνm′)2]⟨g∥M∥e⟩2⟨njl∥T∥n′j′l′⟩2.(20)\sum_{M_gM_e}\left|\widetilde{W}^{ge}_{njlm n'j'l'm'}\right|^2=\frac{1}{3}\left[\sum_{\nu=-1}^{1}\begin{pmatrix} j & 1 & j' \\ -m & \nu& m' \end{pmatrix}^2\right]\langle g\|\mathcal{M}\|e\rangle^2\langle njl\|T\|n'j'l'\rangle^2. \tag*{(20)}

To proceed, we make a further assumption that the quantity in the second line of Eq. (17) does not depend strongly on the magnetic quantum numbers mm and m′m', i.e.,

∑σcσv2Ω(2π)3∫BZd3k ∣anjlm(ckσc)∣2∣an′j′l′m′(vkσv)∣2δ(ϵck−ϵvk−ℏωnuc)≈∑σcσv2Ω(2π)3∫BZd3k ∣anjl(ckσc)∣2∣an′j′l′(vkσv)∣2δ(ϵck−ϵvk−ℏωnuc),(21)\sum_{\sigma_c\sigma_v}\frac{2\Omega}{(2\pi)^3}\int_{\mathrm{BZ}}d^3k\,\left|a_{njlm}(c\mathbf{k}\sigma_c)\right|^2\left|a_{n'j'l'm'}(v\mathbf{k}\sigma_v)\right|^2\delta(\epsilon_{ck}-\epsilon_{vk}-\hbar\omega_{\mathrm{nuc}}) \approx\sum_{\sigma_c\sigma_v}\frac{2\Omega}{(2\pi)^3}\int_{\mathrm{BZ}}d^3k\,\left|a_{njl}(c\mathbf{k}\sigma_c)\right|^2\left|a_{n'j'l'}(v\mathbf{k}\sigma_v)\right|^2\delta(\epsilon_{ck}-\epsilon_{vk}-\hbar\omega_{\mathrm{nuc}}), \tag*{(21)}

then the summation over mm and m′m' may be carried out analytically, giving finally

ΓIC≈π3ℏ12Ie+1⟨g∥M∥e⟩2∑njl∑n′j′l′⟨njl∥T∥n′j′l′⟩2×∑σcσv[2Ω(2π)3∫BZd3k ∣anjl(ckσc)∣2∣an′j′l′(vkσv)∣2δ(ϵck−ϵvk−ℏωnuc)].(22)\Gamma_{\mathrm{IC}}\approx\frac{\pi}{3\hbar}\frac{1}{2I_e+1}\langle g\|\mathcal{M}\|e\rangle^2\sum_{njl}\sum_{n'j'l'}\langle njl\|T\|n'j'l'\rangle^2 \times\sum_{\sigma_c\sigma_v}\left[\frac{2\Omega}{(2\pi)^3}\int_{\mathrm{BZ}}d^3k\,\left|a_{njl}(c\mathbf{k}\sigma_c)\right|^2\left|a_{n'j'l'}(v\mathbf{k}\sigma_v)\right|^2\delta(\epsilon_{ck}-\epsilon_{vk}-\hbar\omega_{\mathrm{nuc}})\right]. \tag*{(22)}

One recognizes the quantity in the square bracket as the (twice) projected joint density of states (PJDOS), which counts the number of allowed transitions at energy ℏωnuc\hbar\omega_{\mathrm{nuc}} between the atomic component ∣n′j′l′m′⟩|n'j'l'm'\rangle of ∣vkσv⟩|v\mathbf{k}\sigma_v\rangle and the atomic component ∣njlm⟩\lvert njlm\rangle of ∣ckσc⟩\lvert c k \sigma_c\rangle. In Figure 3(c) in the main text, we present the most relevant PJDOS. These PJDOS were computed using VASP (see the next section) and are use the scalar relativistic approximation of Koelling & Harmon [50]. To obtain the relativistic PJDOS, we adopt the procedure developed in Ref. [12], i.e., for any l>0l > 0,

∣anl−1/2l(ckσc)∣2=∣anl+1/2l(ckσc)∣2=∣anl∣2/2\left|a_{nl-1/2l}(c k \sigma_c)\right|^2 = \left|a_{nl+1/2l}(c k \sigma_c)\right|^2 = \left|a_{nl}\right|^2 / 2

where ∣anl∣2\left|a_{nl}\right|^2 is the VASP projection. As a result, for l>0l > 0 the relativistic (njl)→(njl)(njl) \to(njl) PJDOS is a quarter of the scalar relativistic value. On the other hand, since the VASP calculation was spin-independent, the double summation over σc\sigma_c and σv\sigma_v in Eq. (22) results in a factor of 2×2=42 \times2 = 4. In Table II, we list the HFI constants AA, the corresponding “diagonal” values of the PJDOS at the nuclear energy for various Th atomic states, and the contributions to the IC rate. Using these values, we arrive at an estimate for the IC rate of

AA (MHz)PJDOS(ℏωnuc\hbar\omega_{\mathrm{nuc}}) (1/eV)Contribution to ΓIC\Gamma_{\mathrm{IC}} (s−1^{-1})
7s1/2−7s1/27s_{1/2} - 7s_{1/2}+62495×10−55 \times10^{-5}431
6p1/2−6p1/26p_{1/2} - 6p_{1/2}+86042.5×10−42.5 \times10^{-4}8182
6p3/2−6p3/26p_{3/2} - 6p_{3/2}+4262.5×10−42.5 \times10^{-4}200
6d3/2−6d3/26d_{3/2} - 6d_{3/2}+155.32.5×10−32.5 \times10^{-3}267
6d5/2−6d5/26d_{5/2} - 6d_{5/2}−12.62.5×10−32.5 \times10^{-3}6.14
5f5/2−5f5/25f_{5/2} - 5f_{5/2}+82.02.5×10−22.5 \times10^{-2}2601
5f7/2−5f7/25f_{7/2} - 5f_{7/2}+31.42.5×10−22.5 \times10^{-2}915

Table II. Magnetic-dipole AA HFI constants for low-lying levels of 229Th3+{}^{229}\mathrm{Th}^{3+}, the corresponding values of PJDOS at the nuclear energy ≈8.35\approx8.35 eV, and the contributions to the IC rate. The AA constants for 7s7s and 6p6p states are computed using our relativistic atomic-structure code with random-phase-approximation (RPA) and perturbative Brueckner-orbital (BO) corrections. The AA constants for 6d6d and 5f5f states are taken from experiment [53].

ΓIC≈1.3×104 s−1,(23)\Gamma_{\mathrm{IC}} \approx1.3 \times10^{4}\ \mathrm{s}^{-1}, \tag*{(23)}

corresponding to an IC lifetime of 80 μs80\ \mu\mathrm{s}.

Computational methods for electronic structure theory

Calculations were performed with VASP [54], version 6.4.2, using the PAW [55] method. The structure of ThO2_2 was optimized with DFT in the conventional unit cell using the PBE [56] functional, 6-6-6 Γ\Gamma-centred kk-mesh and a 500 eV plane wave cutoff. Subsequent electronic structure calculations used the optimized structure in the primitive cell representation.

Parameters for G0W0G_0W_0 calculations [57, 58, 59], specifically the kk-mesh, plane wave cut-off energy, and the number of frequency grid points (NOMEGA tag in VASP), were tested for convergence of the band gap. The number of unoccupied bands was 812 (there are 12 occupied bands). Results of these tests are shown in Table III, Table IV, and Table V. On the basis of these results, further G0W0G_0W_0 calculations were done with an 8-8-8 kk-mesh, a 400 eV plane wave cut-off, and 80 frequency grid points.

meshspacing (Å−1^{-1})band gap (eV)
6-6-60.056.31
7-7-70.0456.21
8-8-80.046.18
9-9-90.0356.22

Table III. Convergence of the G0W0G_0W_0 band gap with respect to kk-mesh.

Cut-off (eV)band gap (eV)
4006.31
5006.30
6006.29

Table IV. Convergence of the G0W0G_0W_0 band gap with respect to plane wave cut-off.

NOMEGAband gap (eV)
87.45
206.36
406.21
806.18

Table V. Convergence of the G0W0G_{0}W_{0} band gap with respect to frequency grid points (NOMEGA).

The Bethe-Salpeter equation (BSE) is a two-particle Green’s-function formalism to explicitly account for electron–hole interactions in electronic excited states [60, 61]. The G0W0G_{0}W_{0}+BSE parameters were converged with respect to the predicted absorption spectrum. The tested parameters were the highest excitation energy (OMEGAMAX) and the number of occupied and unoccupied bands (NBANDSO and NBANDSV) considered in the BSE calculation. Absorption spectra computed with various settings for these methods are shown in Figure 6. On the basis of these tests, further G0W0G_{0}W_{0}+BSE calculations were done with OMEGAMAX = 20 eV and (NBANDSO, NBANDSV) = (8, 16).

Image not extracted

Figure 6. Absorption spectra of ThO₂ computed with G0W0G_{0}W_{0}+BSE using different OMEGAMAX and (NBANDSO, NBANDSV) settings. The meanings of these parameters are given in the text.

We can validate our method against published experimental data by computing the dielectric function with G0W0G_{0}W_{0}+BSE. Our computed data, shown in Figure 7, are an excellent match to the spectroscopic data in Ref. [36]. We also compute the absorption spectrum of ThO₂ with G0W0G_{0}W_{0}+BSE, as shown in Figure 8. The value at the nuclear transition energy is 1×106 cm−1≡0.1 nm−11 \times10^{6}\ \mathrm{cm}^{-1} \equiv0.1\ \mathrm{nm}^{-1}, in agreement with ellipsometric measurements [28].

Image not extracted

Figure 7. Real (ϵ1\epsilon_1) and imaginary (ϵ2\epsilon_2) parts of the frequency-dependent dielectric function for bulk ThO2\mathrm{ThO}_2 computed with G0W0+BSEG_0W_0+\mathrm{BSE}.

Image not extracted

Figure 8. Absorption spectrum of bulk ThO2\mathrm{ThO}_2 computed with G0W0+BSEG_0W_0+\mathrm{BSE}

The absorption spectrum plotted in the main text was computed using a 500 eV plane-wave cut-off and the O_GW pseudopotential, though later testing revealed that these changes did not have a noticeable effect on the calculated absorption spectrum. Absorption spectra and dielectric functions were processed from VASP output with VASPKIT.

Isomer shift

The isomer shift originates from differences in the nuclear charge distribution between the ground and excited nuclear states. The value of the shift depends on the local electronic environment of ²²⁹Th. Here we follow the formalism [34] that combines relativistic many-body atomic structure methods with periodic DFT to evaluate the isomer shifts in ²²⁹ThO₂. The scalar relativistic periodic DFT reproduces valence-band properties while relativistic atomic-structure methods captures the essential core-electron relaxation effects. The valence-band (VB) contribution to the isomer shift in ²²⁹Th solid-state hosts is expressed as [34]

δEisoVB=∑ℓIPDOSℓ δεisoℓ(Th3+)(24)\delta E_{\mathrm{iso}}^{\mathrm{VB}} = \sum_{\ell} \mathrm{IPDOS}_{\ell}\,\delta\varepsilon_{\mathrm{iso}}^{\ell}\left(\mathrm{Th}^{3+}\right) \tag*{(24)}

where IPDOSℓ\mathrm{IPDOS}_{\ell} denotes the integrated projected (on 229Th{}^{229}\mathrm{Th}) valence band density of states for angular-momentum ℓ\ell, and δisoℓ(Th3+)\delta_{\mathrm{iso}}^{\ell}(\mathrm{Th}^{3+}) are the isomer-shift for the lowest-energy valence orbitals of Th3+\mathrm{Th}^{3+} of angular-momentum ℓ\ell. The VB isomer shift is to be added to the isomer shift in 229Th4+{}^{229}\mathrm{Th}^{4+} ion; this contribution remains constant across a wide range of materials. Table VI presents the calculated isomer shifts for 229ThO2{}^{229}\mathrm{ThO}_2 for three different methods. Compared to the PBE and MBJ [62, 63] methods, the G0W0G_0W_0 method includes self-energy correction and, as discussed earlier, we consider it of a higher quality.

ℓ\ellG0_0W0_0 IPDOSℓ_\ellG0_0W0_0 δEiso,ℓVB\delta E_{\mathrm{iso},\ell}^{\mathrm{VB}} (a.u.)MBJ IPDOSℓ_\ellMBJ δEiso,ℓVB\delta E_{\mathrm{iso},\ell}^{\mathrm{VB}} (a.u.)
ss0.119+1.62×10−8+1.62 \times10^{-8}0.078+1.07×10−8+1.07 \times10^{-8}
pp0.435−1.69×10−9-1.69 \times10^{-9}0.408−1.58×10−9-1.58 \times10^{-9}
dd0.759−2.42×10−8-2.42 \times10^{-8}0.688−2.19×10−8-2.19 \times10^{-8}
ff0.534−4.25×10−8-4.25 \times10^{-8}0.432−3.43×10−8-3.43 \times10^{-8}
δEisoVB\delta E_{\mathrm{iso}}^{\mathrm{VB}} (MHz)−343-343−310-310

Table VI. Integrated partial densities of states, IPDOSℓ_\ell, and the products, δEiso,ℓVB=IPDOSℓ δεisoℓ(Th3+)\delta E_{\mathrm{iso},\ell}^{\mathrm{VB}} = \mathrm{IPDOS}_\ell\,\delta\varepsilon_{\mathrm{iso}}^\ell(\mathrm{Th}^{3+}) obtained with the three electronic-structure methods. The isomer shift of the solid-state Th clock, δEisoVB\delta E_{\mathrm{iso}}^{\mathrm{VB}} (bottom row), is the summation over the orbital, ℓ\ell defined in Eq. (24).

Table VII presents the calculated isomer shifts δisoVB\delta_{\mathrm{iso}}^{\mathrm{VB}}, the corresponding nuclear-clock frequencies ν\nu of 229Th{}^{229}\mathrm{Th}, and their offsets Δν\Delta\nu relative to the ThO2\mathrm{ThO}_2 reference for a range of solid-state hosts.

HostδEisoVB\delta E_{\mathrm{iso}}^{\mathrm{VB}} (MHz)ν\nu (MHz)Δν\Delta\nu (MHz)
ThO2_2−310-3102 020 407 338(70)0
CaF2_2 F–90°–F, Ref. [34]−264-2642 020 407 384(70)+46
LiSrAlF6_6, Ref. [34]−234-2342 020 407 414(70)+76

Table VII. Nuclear-clock frequencies ν\nu obtained from the isomer shifts δEisoVB\delta E_{\mathrm{iso}}^{\mathrm{VB}} and their offsets Δν\Delta\nu relative to ThO2_2 computed with MBJ. The absolute frequencies are referenced to the free-ion value ν(Th4+)=2 020 407 648(70)\nu(\mathrm{Th}^{4+}) = 2\,020\,407\,648(70) MHz taken from Ref. [34].

References

  1. [4]H. W. T. Morgan, J. E. S. Terhune, R. Elwell, H. B. T. Tan, U. C. Perera, A. Derevianko, E. R. Hudson, and A. N. Alexandrova, A spinless crystal for a high-performance solid-state ²²⁹Th nuclear clock (2025), arXiv:2408.12309.
  2. [11]J. S. Higgins, T. Ooi, J. F. Doyle, C. Zhang, J. Ye, T. H. Beeks, T. Sikorsky, and T. Schumm, Temperature sensitivity of a thorium-229 solid-state nuclear clock, Phys. Rev. Lett. 134, 113801 (2025).
  3. [12]H. W. T. Morgan, H. B. Tran Tan, R. Elwell, A. N. Alexandrova, E. R. Hudson, and A. Derevianko, Theory of internal conversion of the ²²⁹Th nuclear isomer in solid-state hosts (2024), arXiv:2411.15641.
  4. [13]J. Tiedau, M. V. Okhapkin, K. Zhang, J. Thielking, G. Zitzer, E. Peik, F. Schaten, T. Pronebner, I. Morawetz, L. Toscani De Col, et al., Laser excitation of the Th-229 nucleus, Phys. Rev. Lett. 132, 182501 (2024).
  5. [14]R. Elwell, C. Schneider, J. Jeet, J. Terhune, H. Morgan, A. Alexandrova, H. Tran Tan, A. Derevianko, and E. R. Hudson, Laser excitation of the ²²⁹Th nuclear isomeric transition in a solid-state host, (2024).
  6. [15]C. Zhang, L. von der Wense, J. F. Doyle, J. S. Higgins, T. Ooi, H. U. Friebel, J. Ye, R. Elwell, J. E. S. Terhune, H. W. T. Morgan, A. N. Alexandrova, H. B. Tran Tan, A. Derevianko, and E. R. Hudson, ²²⁹ThF₄ thin films for solid-state nuclear clocks, Nature 636, 603 (2024).
  7. [16]E. V. Tkalya, A. N. Zherikhin, and V. I. Zhudov, Decay of the low-energy nuclear isomer ²²⁹Thᵐ(3/2⁺, 3.5 ± 1.0eV) in solids (dielectrics and metals): A new scheme of experimental research, Phys. Rev. C 61, 064308 (2000).
  8. [17]C. Zhang, T. Ooi, J. S. Higgins, J. F. Doyle, L. von der Wense, K. Beeks, A. Leitner, G. A. Kazakov, P. Li, P. G. Thirolf, T. Schumm, and J. Ye, Frequency ratio of the ²²⁹ᵐTh nuclear isomeric transition and the ⁸⁷Sr atomic clock, Nature 633, 63 (2024), arXiv:2406.18719.
  9. [18]J. E. S. Terhune, R. Elwell, H. B. T. Tan, U. C. Perera, H. W. T. Morgan, A. N. Alexandrovа, A. Derevianko, and E. R. Hudson, Photo-Induced Quenching of the 229Th Isomer in a Solid-State Host (2024), arXiv:arXiv:2412.08998.arxiv.org/abs/2412.08998
  10. [19]F. Schaden, T. Riebner, I. Morawetz, L. T. De Col, G. A. Kazakov, K. Beeks, T. Sikorsky, T. Schumm, K. Zhang, V. Lal, G. Zitzer, J. Tiedau, M. V. Okhapkin, and E. Peik, Laser-Induced Quenching of the Th-229 Nuclear Clock Isomer in Calcium Fluoride (2024), arXiv:2412.12339.arxiv.org/abs/2412.12339
  11. [20]H. W. T. Morgan, R. Elwell, J. E. S. Terhune, H. B. Tran Tan, U. C. Perera, A. Derevianko, A. N. Alexandrovа, and E. R. Hudson, Proposal and theoretical investigation of 229Th-doped nonlinear optical crystals for compact solid-state clocks, Appl. Phys. Lett. 126, 111101 (2025), arXiv:2410.23364.arxiv.org/abs/2410.23364
  12. [21]H. W. T. Morgan, H. B. Tran Tan, A. Derevianko, R. Elwell, J. E. S. Terhune, E. R. Hudson, and A. N. Alexandrovа, Design of new thorium nuclear clock materials based on polyatomic ions 10.26434/chemrxiv-2025-z2bkd (2025).DOI
  13. [22]L. von der Wense et al., A laser excitation scheme for 229mTh, Phys. Rev. Lett. 119, 132503 (2017).DOI
  14. [23]L. C. von der Wense, B. Seiferle, C. Schneider, J. Jeet, I. Amersdorffer, N. Arlt, F. Zacherl, R. Haas, D. Renisch, P. Mosel, P. Mosel, M. Kovacev, U. Morgner, C. E. Düllmann, E. R. Hudson, and P. G. Thirolf, The concept of laser-based conversion electron Mössbauer spectroscopy for a precise energy determination of 229mTh, Hyperfine Interactions 240, 23 (2019).DOI
  15. [24]E. V. Tkalya, C. Schneider, J. Jeet, and E. R. Hudson, Radiative lifetime and energy of the low-energy isomeric level in 229Th, Phys. Rev. C 92, 054324 (2015).DOI
  16. [25]E. R. Hudson, U.S. Provisional Patent Application No. 63/814,873 (2025).
  17. [26]E. T. Rodine and P. L. Land, Electronic defect structure of single-crystal ThO2 by thermoluminescence, Phys. Rev. B 4, 2701 (1971).DOI
  18. [27]B. Seiferle, L. von der Wense, and P. G. Thirolf, Lifetime measurement of the 229Th nuclear isomer, Phys. Rev. Lett. 118, 042501 (2017).DOI
  19. [28]R. Gillis, Determining The Extreme Ultraviolet Constant of Thoria by Spectral Ellipsometry, Bachelor’s thesis, Brigham Young University (2007).
  20. [29]M. P. Seah and W. A. Dench, Quantitative electron spectroscopy of surfaces: A standard data base for electron inelastic mean free paths in solids, Surface and Interface Analysis 1, 2 (1979), https://analyticalsciencejournals.onlinelibrary.wiley.com/doi/pdf/10.1002/sia.740010103.DOI
  21. [30]S. Stellmer, Y. Shigekawa, V. Rosecker, G. A. Kazakov, Y. Kasamatsu, Y. Yasuda, A. Shinohara, and T. Schumm, Toward an energy measurement of the internal conversion electron in the deexcitation of the 229Th isomer, Phys. Rev. C 98, 014317 (2018).DOI
  22. [31]J. F. Ziegler, M. D. Ziegler, and J. P. Biersack, SRIM: The stopping and range of ions in matter (2010), Nucl. Instrum. Methods Phys. Res. B 268, 1818 (2010).
  23. [32]J. Jeet, Search for the low lying transition in the 229Th Nucleus, Ph.D. thesis, University of California, Los Angeles (2018).
  24. [33]L. von der Wense, B. Seiferle, M. Laatiaoui, J. B. Neumayer, H.-J. Maier, H.-F. Wirth, C. Mokry, J. Runke, K. Eberhardt, C. E. Düllmann, N. G. Trautmann, and P. G. Thirolf, Direct detection of the 229Th nuclear clock transition, Nature 533, 47 (2016).DOI
  25. [34]U. C. Perera, H. W. T. Morgan, E. R. Hudson, and A. Derevianko, Host-dependent frequency offsets in 229Th nuclear clockwork (2025), arXiv:2503.20984 [physics.atom-ph].arxiv.org/abs/2503.20984
  26. [35]J. Callaway, Quantum Theory of the Solid State, 2nd ed. (Academic Press, San Diego, CA, 1991).
  27. [36]A. Mock, C. Dugan, S. Knight, R. Korlacki, J. M. Mann, M. M. Kimani, J. C. Petrosky, P. A. Dowben, and M. Schubert, Band-to-band transitions and critical points in the near-infrared to vacuum ultraviolet dielectric functions of single crystal urania and thoria, Applied Physics Letters 114, 10.1063/1.5087059 (2019).DOI
  28. [37]C. Dugan, L. Wang, K. Zhang, J. M. Mann, M. M. Kimani, W.-N. Mei, P. A. Dowben, and J. Petrosky, Interband transitions and critical points of single-crystal thoria compared with urania, physica status solidi (b) 259, 10.1002/pssb.202200238 (2022).DOI
  29. [38]S. Maintz, V. L. Deringer, A. L. Tchougréeff, and R. Dronskowski, Analytic projection from plane-wave and paw wave-functions and application to chemical-bonding analysis in solids, Journal of Computational Chemistry 34, 2557–2567 (2013).DOI
  30. [39]P. V. Bilous, N. Minkov, and A. Pálffy, Electric quadrupole channel of the 7.8 eV Th 229 transition, Phys. Rev. C 97, 2 (2018).DOI
  31. [40]L. von der Wense and C. Zhang, Concepts for direct frequency-comb spectroscopy of 229mth and an internal-conversion-based solid-state nuclear clock, The European Physical Journal D 74, 146 (2020).DOI
  32. [41]J. Belle and R. M. Berman, Thorium dioxide: properties and nuclear applications, Tech. Rep. (USDOE Assistant Secretary for Nuclear Energy, Washington, DC. Office of Naval Reactors, 1984).DOI
  33. [42]W. G. Rellergert, D. DeMille, R. R. Greco, M. P. Hehlen, J. R. Torgerson, and E. R. Hudson, Constraining the evolution of the fundamental constants with a solid-state optical frequency reference based on the 229Th nucleus, Phys. Rev. Lett. 104, 200802 (2010).pubmed.ncbi.nlm.nih.gov/20867019
  34. [43]J.-F. Lin, J. S. Tse, E. E. Alp, J. Zhao, M. Lerche, W. Sturhahn, Y. Xiao, and P. Chow, Phonon density of states of Fe2O3 across high-pressure structural and electronic transitions, Phys. Rev. B 84, 064424 (2011).DOI
  35. [44]H. Wang, A. Braun, S. P. Cramer, L. B. Gee, and Y. Yoda, Nuclear resonance vibrational spectroscopy: A modern tool to pinpoint site-specific cooperative processes, Crystals 11, 10.3390/cryst11080909 (2021).DOI
  36. [45]S. Banerjee, H. P. Gupta, and S. A. Bhardwaj, Nuclear power from thorium: different options, Current Science 111, 1607 (2016).DOI
  37. [46]L. von der Wense, P. V. Bilous, B. Seiferle, S. Stellmer, J. Weitenberg, P. G. Thirolf, A. Pálffy, and G. Kazakov, The theory of direct laser excitation of nuclear transitions, The European Physical Journal A 56, 176 (2020).DOI
  38. [47]V. A. Alekseev and D. W. Setser, A pulsed source for Xe(6s[3/2]1) and Xe(6s′[1/2]1) resonance state atoms using two-photon driven amplified spontaneous emission from the Xe(6p) and Xe(6p′) states, The Journal of Chemical Physics 105, 4613 (1996), https://pubs.aip.org/aip/jcp/article-pdf/105/11/4613/19016841/4613_1_online.pdf.DOI
  39. [48]M. B. Romankin, J. P. Davis, C. O. Giranda, and L. C. Bobb, Two-photon-excited bidirectional emission in xenon, Optics Communications 70, 345 (1989).
  40. [49]H. A. Bethe, S. S. Schweber, and F. de Hoffmann, Mesons and Fields, Volume I: Fields (Row, Peterson and Company, Evanston, IL, 1955).DOI
  41. [50]D. Koelling and B. Harmon, A technique for relativistic spin-polarised calculations, Journal of Physics C: Solid State Physics 10, 3107 (1977).DOI
  42. [51]W. R. Johnson, Atomic structure theory (Springer, 2007).DOI
  43. [52]M. S. Safronova, U. I. Safronova, A. G. Radnaev, C. J. Campbell, and A. Kuzmich, Magnetic dipole and electric quadrupole moments of the 229 Th nucleus, Physical Review A 88, 6 (2013).DOI
  44. [53]C. J. Campbell, A. G. Radnaev, and A. Kuzmich, Wigner crystals of 229Th for optical excitation of the nuclear isomer, Phys. Rev. Lett. 106, 230802 (2011).DOI
  45. [54]G. Kresse and J. Furthmuller, Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set, Physical Review B 54, 11169 (1996).pubmed.ncbi.nlm.nih.gov/9984901
  46. [55]P. E. Blochl, Projector augmented wave method, Physical Review B 50, 17953 (1994).DOI
  47. [56]J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Physical Review Letters 77, 3865 (1996).pubmed.ncbi.nlm.nih.gov/10062328
  48. [57]M. Shishkin and G. Kresse, Implementation and performance of the frequency-dependent GW method within the paw framework, Physical Review B 74, 10.1103/PhysRevB.74.035101 (2006).DOI
  49. [58]M. Shishkin and G. Kresse, Self-consistent GW calculations for semiconductors and insulators, Physical Review B 75, 10.1103/PhysRevB.75.235102 (2007).DOI
  50. [59]P. Liu, M. Kaltak, J. Klimeš, and G. Kresse, Cubic scaling GW: Towards fast quasiparticle calculations, Physical Review B 94, 10.1103/PhysRevB.94.165109 (2016).
  51. [60]S. Albrecht, L. Reining, R. Del Sole, and G. Onida, Ab initio calculation of excitonic effects in the optical spectra of semiconductors, Physical Review Letters 80, 4510–4513 (1998).DOI
  52. [61]M. Rohlfing and S. G. Louie, Electron-hole excitations in semiconductors and insulators, Physical Review Letters 81, 2312–2315 (1998).
  53. [62]F. Tran and P. Blaha, Accurate band gaps of semiconductors and insulators with a semilocal exchange-correlation potential, Physical Review Letters 102, 226401 (2009).pubmed.ncbi.nlm.nih.gov/19658882
  54. [63]A. D. Becke and E. R. Johnson, A simple effective potential for exchange, Journal of Chemical Physics 124, 221101 (2006).pubmed.ncbi.nlm.nih.gov/16784253

Paper details

vlnc:W_zqt5vjsn
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents