Arxiv:2008.08803V1 [Physics.Atom-Ph] 20 Aug 2020

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Caltech Authors - Main

Probing Fundamental Symmetries of Deformed Nuclei in Symmetric Top Molecules

Phelan Yu∗ and Nicholas R. Hutzler† Division of Physics, Mathematics, and Astronomy, California Institute of Technology, Pasadena, California 91125, USA (Dated: August 21, 2020) Precision measurements of Schiff moments in heavy, deformed nuclei are sensitive probes of beyond Standard Model T,P -violation in the hadronic sector. While the most sensitive limits on Schiff moments to date are set with diamagnetic atoms, polar polyatomic molecules can offer higher sensitivities with unique experimental advantages. In particular, symmetric top molecular ions possess K-doublets of opposite parity with especially small splittings, leading to full polarization at low fields, internal co-magnetometer states useful for rejection of systematic effects, and the ability to perform sensitive searches for T,P -violation using a small number of trapped ions containing heavy 225 + exotic nuclei. We consider the symmetric top cation RaOCH3 as a prototypical and candidate platform for performing sensitive nuclear Schiff measurements and characterize in detail its internal structure using relativistic ab initio methods. The combination of enhancements from a deformed nucleus, large polarizability, and unique molecular structure make this molecule a promising platform to search for fundamental symmetry violation even with a single trapped ion.

Searches for permanent electric dipole moments (ThO+, ThF+, AcF, AcO+, AcN, EuO+, EuN, RaO, (EDMs) in atoms and molecules are powerful probes RaF) [26–31] and triatomic molecules (RaOH+, TlOH, of time reversal and parity (T,P-) violating physics ThOH+, TlCN) [27, 32, 33] suitable for Schiff moment posited by beyond Standard Model (BSM) theories [1,2]. measurements. Combining enhancements due to nuclear Unsuppressed T,P-violation, and by extension charge deformation and the polarizability of molecules results 5 conjugation-parity (CP-) violation, are needed to explain in & 10 sensitivity increase relative to atomic Schiff mo- the observed lack of free antimatter in the universe [3]. In ment measurements with spherical nuclei. the Standard Model, however, CP -violation only weakly Molecular ions have proven to be a powerful platform manifests in quark and neutrino mixing phases and is for very sensitive measurements of symmetry violation apparently absent for strong interactions (the strong CP [22] due to long trapping and coherence times [34]. This puzzle) [4]. The hadronic sector thus provides a natu- enables the ability to perform measurements with small ral venue for introducing many new CP-violating BSM quantities of the target molecule, for example those con- interactions to resolve this discrepancy [5–7]. taining scarce or unstable nuclei. However, many BSM- New T, P-violating nuclear effects, including nucleon- sensitive species, including radium, do not have the pre- nucleon interactions mediated by QCD, are understood requisite electronic structure to make diatomic molecular to induce a collective EDM in atomic nuclei of non-zero ions with opposite-parity (Ω) doublets, which are needed spin known as a Schiff moment [8–10]. This effect, which to fully realize the advantages of this approach. Poly- scales with atomic mass Z, is particularly pronounced in atomic molecules, by contrast, possess rovibrational par- heavy, octopole-deformed nuclei, such as 225Ra [11, 12], ity doublets [32], and thus provide a generic approach where low-lying nuclear states couple strongly to the to conducting molecular ion measurements with a broad opposite-parity ground state [13]. The resulting Schiff range of useful, and possibly rare, species. moment, and corresponding sensitivity to BSM physics, In this manuscript, we consider a symmetric top is a factor of 100 larger [14–16] when compared to & molecule (STM), the radium monomethoxide cation heavy spherical nuclei, such as 129Xe [17, 18] and 199Hg (RaOCH+), as a platform to combine nuclear and molec- [19], the latter of which is used in the current most sen- 3 ular enhancements with the advantages of a polyatomic sitive Schiff moment experiment. structure and extended coherence time achievable with Heavy, octopole-deformed isotopes, however, are typ- an ion trap. This molecule, which was recently produced arXiv:2008.08803v1 [physics.atom-ph] 20 Aug 2020 ically short-lived and difficult to produce in large quan- and co-trapped [35] with laser-cooled Ra+ [36], has ax- tities [11, 13, 20]. Maximizing experimental sensitiv- ial symmetry that gives rise to near-degenerate opposite ity and coherence time is thus paramount to overcom- parity K-doublets, thereby enabling full polarization in ing a limited count rate. One demonstrated method small fields and the co-magnetometer states necessary for increasing experimental sensitivity is to use a polar for sensitive measurements in an ion trap. The ground 1 molecule, whose internal fields can be easily oriented to electronic state (X˜ A1) is diamagnetic, suppressing sen- provide an enhancement of & 100 over atoms in EDM sitivity to magnetic noise. Due to this combination of en- + measurements [21–23]. TlF, for instance, is sensitive to hancements and features, even a single trapped RaOCH3 the Schiff moment of 205Tl nuclei [24, 25], and theoret- ion could be used to explore interesting parameter space ical proposals have identified a wide variety of diatomic for new physics. 2

K=+1 K=−1 and is T, P-violating. Similar to the closed-shell alkali monomethyls [40], electron spin terms are omitted. We obtain molecular parameters (see tableI) using a va- K = 0 K × m < 0 + N riety of relativistic ab initio methods. Geometries are Ra (energy > 0) optimized at the level of CCSD(T) with an ANO-RCC- VQZ basis [41–46] via CFOUR [47–49], and scalar rela- O para-NSI E-field tivistic effects are modeled using the one-electron vari- − C ant of the spin-free X2C Hamiltonian [50–52]. Nuclear K × mN > 0 ortho-NSI (energy < 0) spin-rotation and rotational Zeeman parameters are computed via a four-component relativistic linear response approach [53–55] in the DIRAC19 code [56] using the K=−1 K=+1 dyall.v4z basis [57], and electron correlation is treated at mF=−2 mF=+2 the DFT level with a B3LYP functional [58]. Additional 225 + details on the derivation of the Hamiltonian and ab initio Figure 1. Labeled RaOCH3 STMs in stretched states, grouped by energy and total angular momentum projection parameters can be found in the Supplemental Material. mF . The internuclear axis runs from the negatively charged The rotational structure of symmetric tops is parame- end (methyl group) to the positively charged metal (radium). terized by three quantum numbers: the electronic angu- K is the molecule-frame projection of angular momentum lar momentum apart from spin (N), its molecule-frame without spin N, while mN is its projection onto the lab frame. Nuclear spins for each spin-1/2 nucleus (1H, 225Ra) are indi- projection (K), and its lab-frame projection (mN ). For cated. The |K| = 1 states correspond to the mixed para |K| > 0, which corresponds to rotation about the symme- nuclear spin isomer (NSI) of the hydrogens, while the K = 0 try axis, the cylindrical symmetry of the molecule gives state coincides with the stretched ortho-NSI. rise to a pair of degenerate +K and −K states within each |N, |K|i rotational manifold (see fig.1). These degeneracies can be lifted by hyperfine and centrifugal There are several advantages to using a more complex terms that couple states of different K, leading to the STM ion, as opposed to a triatomic analog (e.g. RaOH+) formation of near-degenerate√ opposite parity K-doublets, [27, 32, 37]. First, the increased rovibrational complexity |±i = (|N, +Ki ± |N, −Ki)/ 2. of an STM, which makes laser cooling of neutral species We propose to use the N = |K| = 1 manifold for the more difficult (though indeed possible [38, 39]), does not Schiff moment search. This state is ∼ 160 GHz above pose challenges for the control of STM ions, as trapped the absolute ground state, and accordingly has a much ions do not require photon cycling to achieve high preci- longer lifetime than any vibrationally excited mode (with sion [22, 34]. Furthermore, K-doublets, which arise from frequencies 1 THz) such as those in triatomics. The rotational degrees of freedom, can arise in any vibrational & spontaneous decay rate is further suppressed as the tran- state, such as the ground state considered here. They sition to the lower K = 0 ground state is spin-forbidden, are therefore are low-lying (ν ∼ 100 GHz), have vastly making the radiative lifetime much longer than 1 hour or longer radiative lifetimes than excited vibrational modes, any other relevant experimental timescale. and possess smaller splittings than the `-doublets of tri- Similar to Ω and `-doublets, opposite parity K- atomics. doublets can be mixed in electric fields where the Stark Our theoretical analysis focuses on 225RaOCH+, which 3 energy exceeds the zero-field K-doublet splitting. In this contains the short-lived (τ ≈ 15 d) spin-1/2 radium 1/2 regime, the molecule is polarized and its internal fields isotope. We examine in detail the ground state hyper- are oriented in the lab frame (see. fig2). This gives fine structure, as well as the various contributions to insensitivity to electric field fluctuations by largely satu- the degeneracy-breaking of the K-states. We further- rating the Schiff moment sensitivity, as well as enabling more identify states suitable for measurement of a Schiff co-magnetometer states. In open-shell species, such as moment, including co-magnetometer states, and examine CaOCH , the splitting between the ground state K- the Stark and Zeeman effects in the molecule. 3 doublets is ∼ 0.3 MHz [59], dominated by an anisotropic Internal Structure and K-doubling. The internal struc- hyperfine interaction between the proton spins and the ture of the electronic ground state (X˜ 1A ) is analyzed 1 metal-centered electron. In contrast, the absence of un- with explicit diagonalization of the effective molecular paired electron spin in the 225RaOCH+ ground state im- Hamiltonian 3 plies that the dominant hyperfine contributions to K- doubling are from nuclear spin interactions, which are Htotal = Hrot +Hstark +Hzeeman +Hss +Hnsr +Hsm. (1) suppressed generically by at least an order of magnitude We have included the rotational (rot), Stark, Zeeman, due to the minute size of nuclear magnetic moments com- nuclear spin dipolar (ss), and nuclear spin-rotation terms pared to electronic magnetic moments [60, 61]. (nsr), all of which are generic to STMs. The Schiff mo- For the N = |K| = 1 manifold, we calculate that ment (sm) term arises from the 225Ra(I = 1/2) nucleus, anisotropic nuclear spin-spin and spin-rotation contri- 3

K × mN respond to either of two nuclear spin isomers (NSIs): the −1 “ortho” stretched states (Γ = A, IH = 3/2) and the 2 “para” mixed states (Γ = E,IH = 1/2). 1 In each |N, |K|i manifold, symmetry arguments re- strict the allowed NSIs [63]. In particular, ortho states 0 0 are only allowed with K = 3n rotational states (for inte- −1 ger n), while the para states are associated with K 6= 3n states. All other combinations are forbidden by quantum Dipole Moment (D) Moment Dipole −2 statistics and C symmetries. (See Supplemental Mate- +1 3v rial for details.) Accounting for these restrictions results 0 10 20 30 40 50 Electric Field (mV/cm) in a total of 24 hyperfine states in the N = 1, |K| = 1 manifold, which are all resolvable in high fields. ˜ 1 Figure 2. Lab frame dipole moment of hyperfine states of the Two types of hyperfine terms are present in the X A1 |N = 1, |K| = 1i manifold in the limit where the K doublets state: dipolar couplings between the spins of different nu- are fully mixed yet rotational mixing is negligible. High, low, clei, and couplings between the nuclear spin and molec- and no field seekers correspond to states with negative, pos- ular rotation. The Hamiltonian for dipolar nuclear spin- itive, and zero dipole moment (K × m = +1, 0, and −1). N spin interaction between two spins I1 and I2 can be ex- The jumps indicate avoided crossings. pressed in terms of spherical tensors [64],

√ µ γ γ 2 H = − 6 0 1 2~ T 2(C ) · T 2(I , I ). (2) butions from the hydrogen nuclei generate sub-kHz K- ss 4π dip 1 2 doublings. Combined with the calculated dipole moment of ≈ 5 D, this results in an extremely low threshold for where µ0 is the vacuum permeability, γ1, γ2 are the gyro- reaching the high-field limit and polarizing the molecule. magnetic ratios, and Cdip is a spin-spin coupling tensor. Indeed, we find that states in this manifold are > 90% For the N = |K| = 1 manifold, we only need to consider polarized in external electric fields of 50 mV cm−1 and spin-spin interactions between the ortho-NSI hydrogens 225 > 99.9% polarized in fields of 250 mV cm−1. This thresh- and the Ra atoms, as the inter-hydrogen matrix ele- old is even lower for stretched states with maximal pro- ments vanish between para-NSI states [65, 66]. The dipolar spin coupling tensor T 2(C ), which can be directly jection of total angular momentum (mF ), which reach dip −1 evaluated as a sum of spherical harmonics, gives a diago- full polarization (> 99.9%) in fields as low as . mV cm (see fig.2), small enough that the molecules could be nal shift of ∼ 40 kHz for IH · IR interactions. Anistropic polarized in ∼ mK deep optical traps [62]. Rotational couplings of ∼ 400 Hz mix states differing by ∆K = 2, mixing can be neglected at these small fields, which we which contributes to the K-doubling. assume is the case for the remainder of the manuscript. Nuclear spin-rotation is the interaction between a nu- 1 clear magnetic moment associated with a spin I and the The absence of unpaired electron spin in the X˜ A1 ground state also has implications for the magnetic level magnetic field created by the rotational angular momen- structure, as only the nuclear and rotational moments tum N [67], contribute to the Zeeman energy of the ground state, and 2 1 X are of order ∼ µN , the nuclear magneton. This results H = T k(C ) · T k(N, I) + T k(N, I) · T k(C ), nsr 2 nsr nsr in suppressed sensitivity to magnetic fields and effective k=0 g-factors that are a factor of ∼ 103 smaller than a Bohr (3) 225 magneton µB (see fig3). where Cnsr is a spin-rotation coupling tensor. Both Ra Hyperfine Structure. The large number of internal de- and the hydrogen nuclei in the ortho-NSI contribute to 225 + grees of freedom in RaOCH3 necessitates a detailed the nuclear spin-rotation interaction. The former pro- treatment of the hyperfine structure to both identify reduces a diagonal shift of ∼ 4 kHz for N · IR interactions, solvable states for the Schiff moment measurement as while the latter produces both a diagonal shift of ∼ 15 well as elucidate the sources of K-doubling. We use a kHz for N · IH interactions and ∼ 300 Hz off-diagonal fully decoupled basis, |N, K, mN i|Γ,IH , mIH i|IM , mIM i, couplings between states differing by ∆K = 2, which 1 to describe the hyperfine structure of the X˜ A1 electronic contributes to the K-doubling. state. The spin of the metal 225Ra nucleus is denoted Measurement. Since the energy shift from the Schiff IM , and mIM is its corresponding lab frame projection. moment is proportional to the projection of the radium Similarly, the total nuclear spin of the three hydrogen spin onto the molecular axis (IM · ˆn), different hyperfine atoms and its lab frame projection are denoted IH and states have different Schiff moment sensitivities. In the mIH , while Γ denotes the symmetry character of the hy- fully decoupled limit (≥ 1 V/cm), the value of hIM · ˆni drogen spin wavefunction under C3v transformations. In is K · mN · mIM /2, where the prefactor arises because concrete terms, these hydrogen spin wavefunctions cor- we do not mix rotational states beyond N = 1 and thus 4

methanol [70]. State preparation is possible through K × m HNSR + HSS N state-selective dissociation [34] or non-destructive quan- −0.871 −1.017 +1.017 +0.871 + +0.458 ~6.7 kHz tum logic [71, 72] via co-trapped Ra [36]. −1 −0.312 +0.312 −0.458 Two schemes can be used for performing the EDM +1.2 MHz measurement. In the spin interferometery (SI) method [22], the molecule precesses between states of differ- 0 ~1.8 kHz ent Schiff sensitivity. After a time τ, the phase φ = (ωB + ωTP )τ is extracted via projective measurements, −1.2 MHz where ωB ∝ geff µN |B0|/~ is the Larmor precession fre- +0.871 −1.017 +1.017 −0.871 ~6.7 kHz quency and ωTP ∝ ∆Hsm/ is the contribution from the +0.458 ~ +1 −0.311 +0.311 −0.458 differential Schiff moment between the two states. The −2 0 +1−1 +2 ∆Hsm contribution to the phase can be distinguished from the Larmor precession by repeating the measure- mF ment in different magnetic fields and with different rel- Figure 3. Level structure and Schiff moment sensitivities for ative orientations of the radium nuclear spin and inter- the 24 hyperfine states of the |N = 1, |K| = 1i manifold in nuclear axis, which is enabled in this case by the K- the decoupled regime (1 V/cm), grouped by the projection of doublets. A proposed alternative approach, known as the total angular momentum mF = mN + mIH + mIM and their clock-transition (CT) method [73], uses time-dependent Stark manifold (K × mN ). Gold states have +1/4 effective electromagnetic fields to drive transitions between differ- Schiff sensitivity, while blue states have −1/4 effective Schiff ent hyperfine “clock” states. The T,P -violating inter- sensitivity. Dashed lines denote zero Schiff sensitivity. Labels action is then extracted from phase-dependent shifts to above/below the states indicate the effective g-factor at zero the measured Rabi oscillations. This technique, which magnetic field, in terms of nuclear magnetons (µN ). Table S4 lists the admixtures for each state. is readily adapted to an ion trap, benefits from a sim- pler state preparation scheme and better robustness to electromagnetic noise. the maximum projection of the internuclear axis on the In the polarized limit, the rotational manifold contains lab frame is 1/2. (see fig.3 and table S4). There are many hyperfine states for driving the SI or CT measure- therefore three distinct classes of states, with positive, ment scheme. Each state has different effective g-factors zero, and negative Schiff energy shifts. Even in the in- and Schiff sensitivity (see Fig.3). Performing an EDM termediate regime when a molecule is polarized, but the measurement with pairs of states which have unique dif- spins are not decoupled (∼ 50 mV cm−1), we still find ferential magnetic sensitivities enables one to adjust the stretched states that maximally project the radium spin Larmor precession without changing the applied B-field. onto the molecular axis. In addition, there are multiple pairs of near-degenerate states of opposite m with the same Schiff moment sen- The measurement manifold is naturally populated at F sitivity, but different magnetic moments. These add to cold temperatures, with ∼ 1% of molecules occupying the set of valuable systematic checks. the N = |K| = 1 manifold at 4 K. This yield could BSM Sensitivity. Calculations of the Schiff moment be increased via state-controlled reactions of Ra+ and of 225Ra nuclei have been performed in the framework of T,P -violating pion exchange between nucleons [15], yielding parameterizations of S(225Ra) in terms of the 225 + ¯ 225 ¯ 3 Table I. Molecular parameters for RaOCH3 ground state QCD θ angle given by |S( Ra)| = 1.0θ e fm [74]. 1 (X˜ A1). See Supplemental Material for details. The electrostatic interactions generated by the Schiff moment leads to an effective T,P -violating shift Hsm = Hyperfine Tzz |Txx − Tyy| 225 Ws (I · ˆn) |S|/|I|. The species-dependent coupling con- T (Cnsr( Ra)) 3.67 kHz – 1 stant W , which is an electron-nuclear contact term, has T (Cnsr( H)) 15.3 kHz 0.301 kHz s 225 a αdip · T (Cdip( Ra − H)) −38.0 kHz 0.391 kHz been calculated to be 45, 192 atomic units in RaO [29] Geometry Stark and Zeeman and is estimated to be slightly smaller (∼ 30, 000 a.u.) ˚ + 4 r(Ra-O) 2.1949 A d0 4.969 D for RaOH (a.u. ≡ e/4π0a0), where the larger ligand is 225 [68] r(O-C) 1.4076 A˚ gN ( Ra) −0.7338µN assumed to reduce both electron density around Ra and ˚ [69] r(C-H) 1.0864 A gN (H) 2.7928µN the magnitude of Ws [27]. ◦ ∠(O-C-H) 110.73 gR(k) 6.65µN ◦ To illustrate the power of a Schiff moment mea- ∠(H-C-H) 108.18 gR(⊥) 0.619µN 225 + −1 surement on RaOCH3 , we can combine the A 5.4010 cm 225 −1 QCD parameterization of S( Ra) with the estimate B 0.0673 cm 225 + Ws( RaOCH3 ) ≈ 30, 000 a.u. to calculate the aver- a The scaling constant√ for the dipolar spin-spin interaction is aging time needed to reach a new model-dependent limit 2 + defined as αdip = − 6µ0γHγRa~ /4π ¯ 225 on QCD θ. We assume a single trapped RaOCH3 ion 5 with 5 s coherence time limited by black-body pump- [2] T. Chupp, P. Fierlinger, M. Ramsey-Musolf, and ing at 300 K, which provides√ a frequency sensitivity of J. Singh, Reviews of Modern Physics 91, 015001 (2019). δω = 7.5 mrad s−1/ hour. Spin-precession measure- [3] M. Dine and A. Kusenko, Reviews of Modern Physics ments with this setup would reach a statistical sensitiv- 76, 1 (2003). ¯ −10 [4] H.-Y. Cheng, Physics Reports 158, 1 (1988). ity θ < 10 with two weeks of data taking. Trapping [5] N. Yamanaka, B. K. Sahoo, N. Yoshinaga, T. Sato, multiple ions and improving the coherence time through K. Asahi, and B. P. Das, The European Physical Jour- cryogenic cooling would result in even higher sensitivity. nal A 53, 54 (2017). Conclusion and outlook. We have considered trapped [6] J. Engel, M. J. Ramsey-Musolf, and U. van Kolck, 225 + Progress in Particle and Nuclear Physics 71, 21 (2013). RaOCH3 as a sensitive platform to search for a nuclear Schiff moment in the octopole-deformed Ra nu- [7] J. S. M. Ginges and V. V. Flambaum, Physics Reports cleus. While our theoretical calculations do not replace 397, 63 (2004). [8] L. I. Schiff, Physical Review 132, 2194 (1963). the need for detailed spectroscopic studies on this partic- [9] O. P. Sushkov, V. V. Flambaum, and I. B. Khriplovich, ular species, they do illustrate advantageous structures Zh. Eksp. Teor. Fiz. 87, 1521 (1984). that are quite general for both symmetric and asymmet- [10] V. V. Flambaum and J. S. M. Ginges, Physical Review ric top molecules [66, 75, 76]. This approach can therefore A 65, 032113 (2002). be used to search for a variety of fundamental symme- [11] R. Parker, M. Dietrich, M. Kalita, N. Lemke, K. Bailey, try violations in many different species, including those M. Bishof, J. Greene, R. Holt, W. Korsch, Z.-T. Lu, with exotic nuclei such as Pa [77] and U [78], and many P. Mueller, T. OConnor, and J. Singh, Physical Review Letters 114, 233002 (2015). ligands including chiral species. [12] M. Bishof, R. H. Parker, K. G. Bailey, J. P. Greene, Over the years, a wide variety of metal-monohydroxide R. J. Holt, M. R. Kalita, W. Korsch, N. D. Lemke, Z.- + + (MOH ), metal-monomethoxide (MOCH3 ), as well as T. Lu, P. Mueller, T. P. O’Connor, J. T. Singh, and dual-metal hypermetallic (MOM0+) ions have been cre- M. R. Dietrich, Physical Review C 94, 025501 (2016). ated [79, 80] (and, in some cases, trapped [81, 82]), in- [13] L. P. Gaffney, P. A. Butler, M. Scheck, A. B. cluding species with heavy nuclei such as Ba [81] and Lu Hayes, F. Wenander, M. Albers, B. Bastin, C. Bauer, [80] in addition to Ra [35]. The rich internal complex- A. Blazhev, S. Bnig, N. Bree, J. Cederkll, T. Chupp, D. Cline, T. E. Cocolios, T. Davinson, H. D. Witte, ity of these molecules makes them attractive for a broad J. Diriken, T. Grahn, A. Herzan, M. Huyse, D. G. range of studies not limited to Schiff moment measure- Jenkins, D. T. Joss, N. Kesteloot, J. Konki, M. Kowal- ments [1]. Much of the discussion in this manuscript, for czyk, T. Krll, E. Kwan, R. Lutter, K. Moschner, instance, is directly applicable to searches for the electron P. Napiorkowski, J. Pakarinen, M. Pfeiffer, D. Radeck, EDM or nuclear magnetic quadrupole moments [32, 37]. P. Reiter, K. Reynders, S. V. Rigby, L. M. Robledo, The availability of opposite parity states with diverse, M. Rudigier, S. Sambi, M. Seidlitz, B. Siebeck, T. Stora, tunable splittings is particularly useful for precision mea- P. Thoele, P. V. Duppen, M. J. Vermeulen, M. v. Schmid, D. Voulot, N. Warr, K. Wimmer, K. Wrzosek- surement of electroweak physics, such as nuclear spin- Lipska, C. Y. Wu, and M. Zielinska, Nature 497, 199 dependent parity violation [83] and oscillating symmetry (2013). violations from interactions with axion-like fields [84–86], [14] N. Auerbach, V. V. Flambaum, and V. Spevak, Physi- and the sources that generate the splittings can be sen- cal Review Letters 76, 4316 (1996). sitive to variations of fundamental constants [87, 88]. [15] J. Dobaczewski and J. Engel, Physical Review Letters Acknowledgments. We are grateful for extensive assis- 94, 232502 (2005). tance from Anastasia Borschevsky and Y.A. Chamorro [16] J. Dobaczewski, J. Engel, M. Kortelainen, and P. Becker, Physical Review Letters 121, 232501 (2018). Mena with the ab initio calculations, and to Ben Au- [17] N. Sachdeva, I. Fan, E. Babcock, M. Burghoff, genbraun, Mingyu Fan, Alex Frenett, Arian Jadbabaie, T. Chupp, S. Degenkolb, P. Fierlinger, S. Haude, Andrew Jayich, Ivan Kozyryev, Zack Lasner, and Tim E. Kraegeloh, W. Kilian, S. Knappe-Grneberg, F. Kuch- Steimle for helpful discussions and feedback. This re- ler, T. Liu, M. Marino, J. Meinel, K. Rolfs, Z. Salhi, search was supported by a NIST Precision Measurement A. Schnabel, J. Singh, S. Stuiber, W. Terrano, Grant (60NANB18D253), the Gordon and Betty Moore L. Trahms, and J. Voigt, Physical Review Letters 123, Foundation (7947), and the Alfred P. Sloan Foundation 143003 (2019). [18] F. Allmendinger, I. Engin, W. Heil, S. Karpuk, H.-J. (G-2019-12502). Computations in this manuscript were Krause, B. Niederlnder, A. Offenhusser, M. Repetto, performed on the Caltech High Performance Cluster. U. Schmidt, and S. Zimmer, Physical Review A 100, 022505 (2019). [19] B. Graner, Y. Chen, E. Lindahl, and B. Heckel, Physi- cal Review Letters 116, 161601 (2016). [20] P. A. Butler, L. P. Gaffney, P. Spagnoletti, J. Konki, ∗ [email protected] M. Scheck, J. F. Smith, K. Abrahams, M. Bowry, † [email protected] J. Cederkll, T. Chupp, G. de Angelis, H. De Witte, [1] M. S. Safronova, D. Budker, D. Demille, D. F. Kimball, P. E. Garrett, A. Goldkuhle, C. Henrich, A. Illana, A. Derevianko, and C. W. Clark, Reviews of Modern K. Johnston, D. T. Joss, J. M. Keatings, N. A. Physics 90, 25008 (2018). Kelly, M. Komorowska, T. Krll, M. Lozano, B. S. 6

Nara Singh, D. ODonnell, J. Ojala, R. D. Page, and P.-O. Widmark, The Journal of Physical Chemistry L. G. Pedersen, C. Raison, P. Reiter, J. A. Rodriguez, A 108, 2851 (2004). D. Rosiak, S. Rothe, T. M. Shneidman, B. Siebeck, [42] B. O. Roos, V. Veryazov, and P.-O. Widmark, Theo- M. Seidlitz, J. Sinclair, M. Stryjczyk, P. Van Duppen, retical Chemistry Accounts 111, 345 (2004). S. Vinals, V. Virtanen, N. Warr, K. Wrzosek-Lipska, [43] P.-O. Widmark, P.-A. Malmqvist, and B. O. Roos, The- and M. Zielinska, Nature Communications 10, 2473 oretica Chimica Acta 77, 291 (1990). (2019). [44] B. P. Pritchard, D. Altarawy, B. Didier, T. D. Gibson, [21] J. J. Hudson, D. M. Kara, I. J. Smallman, B. E. Sauer, and T. L. Windus, Journal of Chemical Information and M. R. Tarbutt, and E. A. Hinds, Nature 473, 493 Modeling 59, 4814 (2019). (2011). [45] K. L. Schuchardt, B. T. Didier, T. Elsethagen, L. Sun, [22] W. B. Cairncross, D. N. Gresh, M. Grau, K. C. Cossel, V. Gurumoorthi, J. Chase, J. Li, and T. L. Windus, T. S. Roussy, Y. Ni, Y. Zhou, J. Ye, and E. A. Cornell, Journal of Chemical Information and Modeling 47, 1045 Physical Review Letters 119, 153001 (2017). (2007). [23] V. Andreev, D. G. Ang, D. DeMille, J. M. Doyle, [46] D. Feller, Journal of Computational Chemistry 17, 1571 G. Gabrielse, J. Haefner, N. R. Hutzler, Z. Lasner, (1996). C. Meisenhelder, B. R. O’Leary, C. D. Panda, A. D. [47] J. F. Stanton, J. Gauss, L. Cheng, M. E. Harding, West, E. P. West, and X. Wu, Nature 562, 355 (2018). D. A. Matthews, and P. G. Szalay, “CFOUR, Coupled- [24] D. Cho, K. Sangster, and E. A. Hinds, Physical Review Cluster techniques for Computational Chemistry, a A 44, 2783 (1991). quantum-chemical program package,” With contribu- [25] L. R. Hunter, S. K. Peck, A. S. Greenspon, S. S. Alam, tions from A.A. Auer, R.J. Bartlett, U. Benedikt, C. and D. DeMille, Physical Review A 85, 012511 (2012). Berger, D.E. Bernholdt, Y.J. Bomble, O. Christiansen, [26] V. A. Dzuba, V. V. Flambaum, J. S. M. Ginges, and F. Engel, R. Faber, M. Heckert, O. Heun, M. Hilgen- M. G. Kozlov, Physical Review A 66, 012111 (2002). berg, C. Huber, T.-C. Jagau, D. Jonsson, J. Juslius, [27] V. V. Flambaum, Physical Review C 99, 035501 (2019). T. Kirsch, K. Klein, W.J. Lauderdale, F. Lipparini, [28] L. V.Skripnikov, N. S.Mosyagin, A. V.Titov, and T. Metzroth, L.A. Mck, D.P. O’Neill, D.R. Price, E. V. V.Flambaum, Physical Chemistry Chemical Physics Prochnow, C. Puzzarini, K. Ruud, F. Schiﬀmann, W. , Advance Article (2020). Schwalbach, C. Simmons, S. Stopkowicz, A. Tajti, J. [29] A. D. Kudashov, A. N. Petrov, L. V. Skripnikov, N. S. Vzquez, F. Wang, J.D. Watts and the integral pack- Mosyagin, A. V. Titov, and V. V. Flambaum, Physical ages MOLECULE (J. Almlf and P.R. Taylor), PROPS Review A 87, 020102 (2013). (P.R. Taylor), ABACUS (T. Helgaker, H.J. Aa. Jensen, [30] T. A. Isaev, A. V. Zaitsevskii, and E. Eliav, Journal of P. Jrgensen, and J. Olsen), and ECP routines by A. V. Physics B: Atomic, Molecular and Optical Physics 50, Mitin and C. van Wllen. For the current version, see 225101 (2017). http://www.cfour.de. [31] R. F. Garcia Ruiz, R. Berger, J. Billowes, C. L. Binners- [48] D. A. Matthews, L. Cheng, M. E. Harding, F. Lipparini, ley, M. L. Bissell, A. A. Breier, A. J. Brinson, K. Chrysa- S. Stopkowicz, T.-C. Jagau, P. G. Szalay, J. Gauss, and lidis, T. E. Cocolios, B. S. Cooper, K. T. Flanagan, T. F. J. F. Stanton, The Journal of Chemical Physics 152, Giesen, R. P. deGroote, S. Franchoo, F. P. Gustafsson, 214108 (2020). T. A. Isaev, A. Koszorus, G. Neyens, H. A. Perrett, [49] M. E. Harding, T. Metzroth, J. Gauss, and A. A. Auer, C. M. Ricketts, S. Rothe, L. Schweikhard, A. R. Ver- Journal of Chemical Theory and Computation 4, 64 non, K. D. A. Wendt, F. Wienholtz, S. G. Wilkins, and (2008). X. F. Yang, Nature 581, 396 (2020). [50] K. G. Dyall, The Journal of Chemical Physics 106, 9618 [32] I. Kozyryev and N. R. Hutzler, Physical Review Letters (1997). 119, 133002 (2017). [51] W. Liu and D. Peng, The Journal of Chemical Physics [33] A. V. Kudrin, A. Zaitsevskii, T. A. Isaev, D. E. Maison, 131, 031104 (2009). and L. V. Skripnikov, Atoms 7, 62 (2019). [52] L. Cheng and J. Gauss, The Journal of Chemical [34] Y. Zhou, Y. Shagam, W. B. Cairncross, K. B. Ng, T. S. Physics 135, 084114 (2011). Roussy, T. Grogan, K. Boyce, A. Vigil, M. Pettine, [53] I. A. Aucar, S. S. Gmez, M. C. R. de Aza, and C. G. T. Zelevinsky, J. Ye, and E. A. Cornell, Physical Re- Giribet, The Journal of Chemical Physics 136, 204119 view Letters 124, 053201 (2020). (2012). [35] M. Fan, C. A. Holliman, X. Shi, H. Zhang, M. W. [54] I. A. Aucar, S. S. Gmez, J. I. Melo, C. C. Giribet, and Straus, X. Li, S. W. Buechele, and A. M. Jayich, M. C. Ruiz de Aza, The Journal of Chemical Physics arXiv:2007.11614 (2020). 138, 134107 (2013). [36] M. Fan, C. Holliman, A. Wang, and A. Jayich, Physical [55] I. A. Aucar, S. S. Gomez, C. G. Giribet, and M. C. Review Letters 122, 223001 (2019). Ruiz de Aza, The Journal of Chemical Physics 141, [37] D. E. Maison, L. V. Skripnikov, V. V. Flambaum, and 194103 (2014). M. Grau, arXiv:2006.03848 (2020). [56] A. S. P. Gomes, T. Saue, L. Visscher, H. J. Aa. Jensen, [38] D. Mitra, N. B. Vilas, C. Hallas, L. Anderegg, B. L. and R. Bast, “DIRAC, a relativistic ab initio electronic Augenbraun, L. Baum, C. Miller, S. Raval, and J. M. structure program, Release DIRAC19 (2019),” With Doyle, arXiv:2004.02848 (2020). contributions from I. A. Aucar, V. Bakken, K. G. Dyall, [39] I. Kozyryev, T. C. Steimle, P. Yu, D.-T. Nguyen, and S. Dubillard, U. Ekstrm, E. Eliav, T. Enevoldsen, E. Fa- J. M. Doyle, New Journal of Physics 21, 052002 (2019). hauer, T. Fleig, O. Fossgaard, L. Halbert, E. D. Hede- [40] B. Z. Li, J. Xin, and L. M. Ziurys, Chemical Physics grd, B. HeimlichParis, T. Helgaker, J. Henriksson, M. Letters 280, 513 (1997). Ilia, Ch. R. Jacob, S. Knecht, S. Komorovsk, O. Kullie, [41] B. O. Roos, R. Lindh, P.-A. Malmqvist, V. Veryazov, J. K. Lrdahl, C. V. Larsen, Y. S. Lee, H. S. Nataraj, M. 7

K. Nayak, P. Norman, G. Olejniczak, J. Olsen, J. M. Physical Chemistry Letters 9, 3555 (2018). H. Olsen, Y. C. Park, J. K. Pedersen, M. Pernpointner, [83] E. B. Norrgard, D. S. Barker, S. Eckel, J. A. Fedchak, R. di Remigio, K. Ruud, P. Saek, B. Schimmelpfennig, N. N. Klimov, and J. Scherschligt, Communications B. Senjean, A. Shee, J. Sikkema, A. J. Thorvaldsen, Physics 2, 77 (2019). J. Thyssen, J. van Stralen, M. L. Vidal, S. Villaume, [84] P. W. Graham and S. Rajendran, Physical Review D O. Visser, T. Winther, and S. Yamamoto (available 84, 055013 (2011). at http://dx.doi.org/10.5281/zenodo.3572669, see also [85] Y. Stadnik and V. Flambaum, Physical Review D 89, http://www.diracprogram.org). 043522 (2014). [57] K. G. Dyall, The Journal of Physical Chemistry A 113, [86] V. Flambaum and H. T. Tan, Physical Review D 100, 12638 (2009). 111301 (2019). [58] P. J. Stephens, F. J. Devlin, C. F. Chabalowski, and [87] P. Jansen, H. L. Bethlem, and W. Ubachs, The Journal M. J. Frisch, The Journal of Physical Chemistry 98, of Chemical Physics 140, 010901 (2014). 11623 (1994). [88] I. Kozyryev, Z. Lasner, and J. M. Doyle, [59] K.-i. C. Namiki, J. S. Robinson, and T. C. Steimle, J. arXiv:1805.08185 (2018). Chem. Phys. 109, 5283 (1998). [89] Y. Endo, C. Yamada, S. Saito, and E. Hirota, The [60] W. Klemperer, K. K. Lehmann, J. K. G. Watson, and Journal of Chemical Physics 77, 3376 (1982). S. C. Wofsy, The Journal of Physical Chemistry 97, [90] K. Raghavachari, G. W. Trucks, J. A. Pople, and 2413 (1993). M. Head-Gordon, Chemical Physics Letters 157, 479 [61] R. J. Butcher, C. Chardonnet, and C. J. Bord, Physical (1989). Review Letters 70, 2698 (1993). [91] R. J. Bartlett, J. D. Watts, S. A. Kucharski, and [62] C. Schneider, M. Enderlein, T. Huber, and T. Schaetz, J. Noga, Chemical Physics Letters 165, 513 (1990). Nature Photonics 4, 772 (2010). [92] D. Bykov, T. Petrenko, R. Izsk, S. Kossmann, [63] J. T. Hougen, Journal of Molecular Spectroscopy 81, 73 U. Becker, E. Valeev, and F. Neese, Molecular Physics (1980). 113, 1961 (2015). [64] J. M. Brown and A. Carrington, Rotational Spectroscopy [93] F. Neese, WIREs Computational Molecular Science 2, of Diatomic Molecules (Cambridge University Press, 73 (2012). Cambridge, UK, 2003). [94] T. H. Dunning, The Journal of Chemical Physics 90, [65] G. R. Gunther-Mohr, C. H. Townes, and J. H. 1007 (1989). Van Vleck, Physical Review 94, 1191 (1954). [95] J. G. Hill and K. A. Peterson, The Journal of Chemical [66] M. L. Wall, K. Maeda, and L. D. Carr, Annalen der Physics 147, 244106 (2017). Physik 525, 845 (2013). [96] I. S. Lim, H. Stoll, and P. Schwerdtfeger, The Journal [67] E. Hirota, High-Resolution Spectroscopy of Transient of Chemical Physics 124, 034107 (2006). Molecules, Vol. 40 (Springer-Verlag, Berlin, 1985). [97] S. S. Kondov, C.-H. Lee, K. H. Leung, C. Liedl, [68] E. Arnold, W. Borchers, M. Carre, H. T. Duong, P. Jun- I. Majewska, R. Moszynski, and T. Zelevinsky, Nature car, J. Lerme, S. Liberman, W. Neu, R. Neugart, E. W. Physics 15, 1118 (2019). Otten, M. Pellarin, J. Pinard, G. Ulm, J. L. Vialle, and [98] I. Manai, R. Horchani, H. Lignier, P. Pillet, D. Com- K. Wendt, Physical Review Letters 59, 771 (1987). parat, A. Fioretti, and M. Allegrini, Physical Review [69] B. Singh, Nuclear Data Sheets 106, 601 (2005). Letters 109, 183001 (2012). [70] M.-G. Hu, Y. Liu, M. A. Nichols, L. Zhu, G. Qumner, [99] T. Shimasaki, M. Bellos, C. D. Bruzewicz, Z. Lasner, O. Dulieu, and K.-K. Ni, arXiv:2005.10820 (2020). and D. DeMille, Physical Review A 91, 021401 (2015). [71] D. Patterson, Physical Review A 97, 033403 (2018). [100] P. F. Staanum, K. Hjbjerre, P. S. Skyt, A. K. Hansen, [72] C.-w. Chou, C. Kurz, D. B. Hume, P. N. Plessow, D. R. and M. Drewsen, Nature Physics 6, 271 (2010). Leibrandt, and D. Leibfried, Nature 545, 203 (2017). [101] N. B. Khanyile, G. Shu, and K. R. Brown, Nature Com- [73] M. Verma, A. M. Jayich, and A. C. Vutha, munications 6, 7825 (2015). arXiv:1909.02650 (2020). [102] A. E. Leanhardt, J. L. Bohn, H. Loh, P. Maletinsky, [74] V. V. Flambaum and H. Feldmeier, Physical Review C E. R. Meyer, L. C. Sinclair, R. P. Stutz, and E. A. Cor- 101, 015502 (2020). nell, Journal of Molecular Spectroscopy 270, 1 (2011). [75] B. L. Augenbraun, J. M. Doyle, T. Zelevinsky, and [103] S. C. Wofsy, J. S. Muenter, and W. Klemperer, The I. Kozyryev, Physical Review X 10, 031022 (2020). Journal of Chemical Physics 53, 4005 (1970). [76] P. Yu, L. W. Cheuk, I. Kozyryev, and J. M. Doyle, New [104] M. R. Aliev and J. K. G. Watson, Journal of molecular Journal of Physics 21, 093049 (2019). spectroscopy 52, 29 (1976). [77] V. V. Flambaum, Physical Review A 77, 024501 (2008). [105] K. Kawaguchi, Canadian Journal of Physics 79, 449 [78] M. Wormit, M. Olejniczak, A.-L. Deppenmeier, (2001). A. Borschevsky, T. Saue, and P. Schwerdtfeger, Physi- [106] Y. Endo, S. Saito, and E. Hirota, The Journal of Chem- cal Chemistry Chemical Physics 16, 17043 (2014). ical Physics 81, 122 (2003). [79] W. Lu and S. Yang, The Journal of Physical Chemistry [107] Note that we have used the canonical spherical tensor A 102, 1954 (1998). √formulations (e.g. Table 5.2 in [64]) and do not adopt the [80] W. Lu and S. Yang, The Journal of Physical Chemistry 6 scaling convention nor the tr(T1) = 0 assumption for A 102, 825 (1998). describing the dipolar hyperﬁne interactions in [89]. [81] P. Puri, M. Mills, C. Schneider, I. Simbotin, J. A. Mont- [108] C. Adamo and V. Barone, The Journal of Chemical gomery, R. Ct, A. G. Suits, and E. R. Hudson, Science Physics 110, 6158 (1999). 357, 1370 (2017). [109] T. Chupp and M. Ramsey-Musolf, Physical Review C [82] T. Yang, A. Li, G. K. Chen, C. Xie, A. G. Suits, W. C. 91, 035502 (2015). Campbell, H. Guo, and E. R. Hudson, The Journal of 8

Supplemental Material Vibration Energy [cm−1] Character Ra-O-C bend 164.96/168.68 e Ra-O stretch 390.78 a1 MOLECULAR STRUCTURE CH3 symmetric bend 1086.84 a1 CH3 rock 1170.41/1173.36 e As discussed in the main text, we numerically diago- C-O stretch 1481.50 a1 CH asymmetric bend 1497.61/1498.56 e nalize the eﬀective molecular Hamiltonian for the ground 3 CH symmetric stretch 2993.19 a electronic state (X˜ 1A ) of 225RaOCH+ 3 1 1 3 CH3 asymmetric stretch 3048.22/3059.15 e

Htotal = Hrot +Hstark +Hzeeman +Hss +Hnsr +Hsm (S1) Table S1. Vibrational energies, computed from B3LYP Hes- where we have included the rotational (rot), Stark, Zee- sians [92] at DFT optimized geometry, where each mode is man, nuclear spin dipolar (ss), nuclear spin-rotation classified with respect to its transformations under C3v sym- terms (nsr), and Schiff moment (sm) terms. No electron metry. In total, there are four symmetric a1 states and four spin terms are included, as the molecule has a closed doubly-degenerate e states. The pair of frequencies for the degenerate vibrations is due to slight symmetry-breaking in shell. For generality, however, the matrix elements are the computed geometry. written in the fully decoupled basis including the electron spin S: |N, K, S, J, mJ i|IM , mIM i|Γ,IH , mIH i. atom. The 78 core electrons of radium are modeled using the SK-MCDHF-RSC effective core potential (ECP) Rovibrational Structure [96]. Energies and symmetry characters are listed in Ta- ble S1. While the proposed Schiff moment search would + RaOCH3 is a prolate symmetric top with point group take place in the ground vibrational state, excited ro- C3v, corresponding to its three-fold cylindrical symmetry vibrational states could be a valuable resource for state about the principal molecular axis (Z). The Hamiltonian preparation or readout [97–101]. The energy difference that corresponds to the rotational energy for a prolate between the nominally degenerate states of e character top is indicate that the accuracy of the energies is on the few percent level. H = BN2 + (A − B)N2 , (S2) rot Z The vibrational states are also relevant as they will which has eigenenergies BN(N+1)+(A−B)K2. A and B likely present a limitation on the coherence time [102]. Black-body excitation of the Ra-O stretch mode (with are rotational constants, while hNi = N and hNZ i = K are the canonical rotational quantum numbers. A corre- the transition dipole moment calculated to be µ ∼ 0.26 sponds to the rotation of the molecule about the symme- D) is estimated to occur with a ∼5 second time scale in try axis, while B corresponds to end-over-end rotation. a 300 K environment, though that can be reduced to 20 For |K| 6= 0, each |N, |K|i-level in the rotational Hamil- minutes in a 77 K environment. The radiative lifetime tonian has 2 × (2N + 1) degeneracies, which are given of the N = |K| = 1 states are much longer than one day due to their small energy spacing (a state with one by the 2N + 1 lab-frame projections mN and the two projections of |K| onto the molecular axis. atomic unit of transition dipole moment at this frequency The ab initio geometries, with corresponding rota- would last around one hour) and the fact that they are + spin-forbidden to decay to the ground rotational state, so tional constants, for RaOCH3 (see tableI in main text) were optimized using the coupled cluster method black-body effects are likely to dominate over radiative with single, doubles, and perturbative triples [CCSD(T)] decay at room temperature. [90, 91] in CFOUR [47–49]. Relativistically contracted atomic natural orbital basis sets of polarized double, triple, and quadruple-zeta quality [ANO-RCC-VnZP Molecular Symmetries (n=D,T,Q)] are used [41–46]. Scalar relativistic effects are included via the one-electron variant of the spin-free Classification of Rovibronic and Nuclear Spin Wavefunctions exact two component theory (SFX2C-1e) [50–52]. There are eight vibrational modes for symmetric top The C3v point group (which is isomorphic to S6) has molecules of the MOCH3 form, four symmetric modes three irreducible representations: A1, A2, and E. A1 de- of a1 character and four asymmetric modes of e charac- notes the fully symmetric representation, and A2 is the ter. The vibrational energies and intensities of each mode anti-symmetric representation. E denotes the degener- are calculated via analytic B3LYP Hessians [58, 92] with ate representation, which splits into positive (E+) and ORCA [93]. For this calculation, we employ correlation- negative (E−) components under the action of the cyclic consistent basis sets at the quadruple-zeta level [94], with group C3 ⊂ C3v. The character tables for C3v and C3 core-valence [95] and pseudopotential sets for the radium are written in Table S2 for reference. 9

Eˆ 2Cˆ (z) 3σ 3 v Eˆ Cˆ Cˆ2 with the splitting of E± or A1 ⊕ A2 into distinct repre- A +1 +1 +1 3 3 1 A +1 +1 +1 sentations. A +1 +1 −1 2 E +1 e∓2πi/3 e±2πi/3 We discuss the specific mechanisms lifting the degen- E +2 −1 0 ± eracy below, but there is a simple and intuitive picture how this arises due to hyperfine couplings [65]. The K- Table S2. Character Tables for C3v (left) and C3 (right) doubled states |±i have a different distribution of proton spin about the azimuthal angle relative to the symmetry We start by classifying the rovibronic states. In the axis and therefore the anisotropic nature of the dipolar electronic ground state and a vibrationally relaxed man- or nuclear spin-rotation interaction splits the two states. ifold (|Λ = 0, ` = 0i), the symmetry classification of the For |K| = 1 states, the hyperfine anisotropies directly |N, |K|i rovibronic states is dependent only on K. The couple states differing by ∆K = 2 to produce a first-order |N,K = 0i state transforms as A1 or A2 (depending K-splitting, but have progressively suppressed effects on on N), while |N, |K| = 3ni states, where n is an inte- the splitting for higher |K| > 1 [65, 103]. ger and K 6= 0, transform as A1 ⊕ A2. The remaining For states where |K| = 3n, the leading order source |N, |K|= 6 3ni states transform as the doubly degener- of K-doubling arises from a sextic centrifugal distor- ate E± character, where the positive and negative com- tion term, which couples ∆K = 6. This term appears ponents correspond to the positive and negative projec- as a correction to the rotational Hamiltonian Hsextic = 6 6 tions of |K|. (A generalized classification for non-zero or- q3(J+ + J−)/2, where q3 is the distortion constant [104]. bital and vibrational angular momentum (|Λ 6= 0, l 6= 0i) In the CH3 radical, the q3 distortion constant has been can be found in [63].) The dependence of the molecular measured to be 370 Hz [105]. At higher multiples of three, symmetries on K arises from the fact that the hydrogen the K-splitting generated by q3 is again suppressed. Us- atoms in the methyl group are indistinguishable and must ing higher K states is therefore a general way to obtain obey the Pauli principle, as discussed in detail later. even smaller K-doublets. All states where |K| > 0 are doubly degenerate and The composite nuclear spin states |Γ,IH , mIH i of the thus correspond to either the mixed E± or A1 ⊕ A2 sym- three hydrogen atoms, where Γ denotes the symmetry, metry characters. The splitting of K-doublets (and thus are also classified into four states that transform as A1, the breaking of C3v symmetry) is naturally associated and four states that transform as E±:

3 3E 1 1 1E A, , ± = ± , ± , ± (S3) 2 2 2 2 2 3 1E 1 1 1 1E 1 1 1E 1 1 1E A, , ± = √ ± , ± , ∓ + ± , ∓ , ± + ∓ , ± , ± (S4) 2 2 3 2 2 2 2 2 2 2 2 2 1 1E 1 1 1 1E +2πi/3 1 1 1E −2πi/3 1 1 1E E+, , ± = √ ∓ , ± , ± + e ± , ∓ , ± + e ± , ± , ∓ (S5) 2 2 3 2 2 2 2 2 2 2 2 2 1 1E 1 1 1 1E −2πi/3 1 1 1E +2πi/3 1 1 1E E−, , ± = √ ∓ , ± , ± + e ± , ∓ , ± + e ± , ± , ∓ (S6) 2 2 3 2 2 2 2 2 2 2 2 2

IH refers to the total composite spin of the three protons states correspond to two possibilities (IH = IH1 + IH2 + IH3) and mIH is the corresponding evsr n evsr n lab frame projection. Γ denotes the symmetry of the | A1i| Ai or | A2i| Ai. (S8) composite spin state. To obey Fermi-Dirac statistics, the total nuclear- From eq. (S7), we can observe that any hyperﬁne interaction which couples |nE i and |nE i states also couples rovibronic wavefunction must transform as either A1 + − the |evsrE i and |evsrE i states, and therefore splits any (symmetric under inversion) or A2 (antisymmetric un- + − der inversion) [63]. Therefore, the K = 1 and K = 2 (in residual K-degeneracy into the doublets, the ground state) cases correspond to combined nuclear n evsr 1 evsr n evsr n ( Γ) and rovibronic ( Γ) wavefunctions that have the √ (| E+i| E−i ± | E−i| E+i). (S9) symmetries 2

evsr n evsr n This is the case for the K-doubling that originates from | E+i| E−i or | E−i| E+i, (S7) + the nuclear dipolar spin couplings of RaOCH3 , for in- whereas the fully symmetric K = 0 or K = 3 rotational stance. 10

The symmetry assignments in eqs. (S7) and (S8) also spin basis: lead to additional selection rules for electric dipole tran- 1 p sitions. For instance, let us consider transitions between hΓ,I0|T (I0)|Γ,I0i = I0(I0 + 1)(2I0 + 1), (S17) different |N,Ki manifolds in the electronic-vibrational 1 p ground state. Starting from the K = 1 manifold, the hE±,I0|T (I±)|E∓,I0i = −2 I0(I0 + 1)(2I0 + 1), (S18) electric dipole operator only couples to ∆K = 0, ±1 √ 1 states. |N,K = 2i has the same nuclear symmetry Γ hE±,I0 = 1/2|T (I±)|A1,I0 = 3/2i = 6. (S19) as |N,K = 1i and the transition is allowed. |N,K = 0i does not, and the transition is symmetry-forbidden (as Meanwhile, the parameterized T operators, in cartesian well as spin-forbidden), which is why the radiative life- form, are: time of |N,K = 1i is so long.   (Txx + Tyy)/2 0 0 Hyperfine Structure T0 =  0 (Txx + Tyy)/2 0  , (S20) 0 0 Tzz   Parameterization of the Tensor Operators (Txx − Tyy)/4 ∓i(Txx − Tyy)/4 Txz/2 T± = ∓i(Txx − Tyy)/4 −(Txx − Tyy)/4 ±iTxz/2 , In order to evaluate hyperfine structure of the hydro- Tzx/2 ±iTzx/2 0 gen spins, it is necessary to sum over the interactions in- (S21) volving each individual individual hydrogen nucleus. A spin-spin interaction between the hydrogen spins Ii and which have the corresponding non-zero rank-2 compo- 2 an arbitrary spin S, for instance, has the form nents in spherical tensor form (Tα)q [107]: 3 √ X 2 S· Ti · Ii, (S10) (T0)0 = (2Tzz − Txx − Tyy)/ 6, (S22) i=1 2 (T±)∓ = ∓(Txz + Tzx)/2, (S23) 2 where Ti is the interaction tensor between the ith hydro- (T±)±2 = (Txx − Tyy)/2. (S24) gen and S. In Cartesian form, T1 can be written as   By applying a Wigner rotation to the molecule-frame (q) Txx 0 Txz spherical components, we obtain T1 =  0 Tyy 0  , (S11) T 0 T zx zz 0 0 2 2 X N 0−K0 hN K |Dpq(ω)(Tα)q|NKi = (−1) while T2 and T3 can be obtained with the appropriate q rotations about the z axis [89]. N 0 2 N ×p(2N 0 + 1)(2N + 1) (T )2. (S25) The form of eq. (S10), however, is unwieldy for evalu- −K0 q K α q ating matrix elements over the symmetrized nuclear spin states derived in the previous section. A more intuitive We can see that the anisotropic tensor component in form can be obtained by parameterizing the T and I op- eq. (S24) will couple states with ∆K = 2. In the erators, as first formulated by Hougen in cartesian form case of |K| = 1, this term would couple K = +1 and [63] and Endo et al. in spherical tensor form [89, 106]. K = −1 terms, providing a first-order contribution to the K-doubling. Most of the sources of K-splitting in T = (T + T + T )/3, (S12) 0 1 2 3 the hyperfine structure of symmetric tops arise in this ±2πi/3 ∓2πi/3 T± = (T1 + e T2 + e T3)/3, (S13) manner.

I0 = I1 + I2 + I3, (S14) ±2πi/3 ∓2πi/3 I± = I1 + e I2 + e I3. (S15) Stark and Zeeman Matrix Elements This allows us to rewrite eq. (S10) as

1 1 X The Stark Hamiltonian is HˆS = −T (d) · T (E), S · Tα · I−α. (S16) α=0,± where d is the molecule frame dipole moment and E is the electric ﬁeld. Without loss of generality, we Direct evaluation via the Wigner-Eckart Theorem yields set the electric ﬁeld along thez ˆ direction in the lab the following reduced matrix elements for the parameter- frame. The matrix elements in the decoupled basis ized nuclear spin operators over the symmetrized nuclear |N, K, S, J, mJ i|IM , mIM i|Γ,IH , mIH i are given by 11

30 20 50 10

0 0 -10

( kHz ) Shift Stark -20 -50 ( kHz ) Shift Zeeman -30 0 5 10 15 0 2 4 6 8 10 Electric Field(mV/cm) Magnetic Field(G)

Figure S1. (left) Stark plot of hyperfine states in the |N = 1, |K| = 1i manifold, up to 15 mV cm−1 electric field. (right) Zeeman plot of hyperfine states in the |N = 1, |K| = 1i manifold, up to 10 G magnetic field.

0 0 0 0 0 ˆ hN K S J mJ ; IM mIM ;ΓIH mIH |HS|NKSJmJ ; IM mIM ;ΓIH mIH i 0 0 0 0 0 1 = −EhN K S J mJ |T0 (d)|NKSJmJ i 0 0 0 0 J 1 J 0 0 NJS J −mJ J+N +1+S p 0 = −Ed0(−1) 0 (−1) (2J + 1)(2J + 1) 0 0 −mJ 0 mJ J N 1 0 0 0 N 1 N × (−1)N −K p(2N 0 + 1)(2N + 1) . (S26) −K0 0 K

Note that the dipole moment is along the molecular axis ment for a single nucleus is, generically, 1 such that T0 (d) = d0 and that the matrix element is diagonal with respect to the nuclear spins. The Zeeman Hamiltonian is written as Hˆz = Hˆz,I + 0 ˆ 0 0 I 1 I Hz,R, with two terms corresponding to the coupling of 0 0 1 I −mI hI , mI |Tp (I)|I, mI i = (−1) 0 the nuclear spin and molecular rotation to the external −mI p mI magnetic ﬁeld. p × δI0,I I(I + 1)(2I + 1). (S29) The nuclear spin term Hˆz,I is given by

X 1 1 Hˆz,I = −gN µN T (Ii) · T (B) i For the coupling to hydrogen spins, we evaluate the pa- X p 1 1 1 = −gN µN (−1) Tp (Ii)T−p(B) (S27) rameterized T (I0) matrix elements derived in the previ- i,p ous section, which results in a near-identical expression to eq. (S29), where we note that we need to sum over all the hydrogen and radium nuclear spins. Evaluating the matrix element in the decoupled basis for the radium spin, which is diagonal in |N, K, S, J, mJ i|Γ,IH , mIH i: ˆ hIM mIM |Hz,IM |IM mIM i = X p 0 0 0 1 − gN µN (−1) hΓ ,IH , mIH |Tp (I0)|Γ,IH , mIH i 0 0 ˆ p hIM mIM |Hz,IM |IM mIM i = X p 0 0 1 1 × hT 1 (B)i. (S30) − gN µN (−1) hIM , mIM |Tp (IM )|IM , mIM ihT−p(B)i, −p p (S28) where the nuclear spin angular momentum matrix ele- 12

The rotational Zeeman term [67] is: Nuclear Spin-Rotation Coupling

The nuclear spin-rotation coupling is expressed similarly to the electron spin-rotation elements:

1 X H = C (N I + I N ) (S33) nsr 2 α,β α β β α α,β ˆ Hz,R = gRµN N · B (S31) 2 r 1 X k k k k X 2k + 1 = T (C) · T (N, I) + T (N, I) · T (C) = µ B (−1)k T 1(gk, N) 2 N 0 3 0 r k=0 k (S34) √ X 2k+1 = µN B0 (−1) 2k + 1 2 k 1 X X k k k k k = T (C)T (N, I) + T (N, I)T (C) , 2 p −p p −p X 1 k=0 p × T k(g )T 1 (N) + (−1)kT 1(N)T k (g ) 2 p r −p p −p r (S35) p k 1 1 where the product is decomposed as × . (S32) p −p 0 √ k p X 1 1 1 1 k Tp (N, I) = (−1) 2k + 1 Tp (N)Tp (I) . 1 2 p1 p2 −p p1,p2 (S36)

225 + In RaOCH3 , the nuclear spin rotation interaction is divided into two components: the interaction between k 225 Relativistic ab initio values for the rotational g-tensor gr rotation and the spin-1/2 Ra nucleus, and the inter- have been computed using the four-component linear re- action between rotation and the hydrogen spins in the sponse within elimination of small component (LRESC) methyl group. approach of Aucar et al. [55] as implemented in the Let us consider the spin-rotation coupling for the single DIRAC19 code [56]. Electron correlation is treated at radium spin IM . We write out eq. (S35), in the decoupled the level of DFT with a B3LYP functional [58] and a basis |N, K, S, J, mJ i|IM , mIM i|Γ,IH , mIH i, taking care dyall.v4z basis is used [57]. See tableI in main text for to sum over a complete set of states between T (C) and speciﬁc values. T (N),

Note that eq. (S37) is diagonal with respect to the |Γ,IH , mIH i spin states. 0 0 0 0 0 1 hN K S J mJ |Tp (N)|NKSJmJ i = 0 0 0 J 1 J J −mJ × (−1) 0 Evaluating the nuclear spin-rotation and rotational an- −mJ p mJ 0 0 gular momentum matrix elements yields J+N +1+S p 0 × δS0,S(−1) (2J + 1)(2J + 1) NJS p × δ 0 (N(N + 1)(2N + 1), (S38) J 0 N 0 1 N ,N 13

1 The lab-frame matrix element Tp (IM ) can be evaluated as eq. (S29) for a single spin. 0 0 0 0 0 k hN K S J mJ |Tp (C)|NKSJmJ i = 0 Now, we consider the coupling to rotation for the com- 0 0 J k J J −mJ (−1) 0 posite hydrogen spins (IH ). We can utilize the param- −m p mJ J eterized nuclear spin operators and coupling tensors de- 0 0 J+N +k+S p 0 NJS rived earlier. This adds an additional summation over × δS0,S(−1) (2J + 1)(2J + 1) 0 0 J N k the parameter α, X 0 0 × (−1)N −K p(2N 0 + 1)(2N + 1) q 0 N k N 2 k × T k(C), (S39) 1 X X X −K0 q K q H = T k(C )T k (N, I ) nsr-H 2 p α −p −α α k=0 p where, for an on-axis spin, the non-zero spherical tensor + T k(N, I )T k (C ). (S43) elements for the nuclear spin-rotation coupling are: p −α −p α √ 0 T0 (C) = −(Cxx + Cyy + Czz)/ 3, (S40) √ 2 T (C) = (2Czz − Cxx − Cyy)/ 6, (S41) 0 We can thus generalize eq. (S37) to the case of multiple 2 T±2(C) = (Cxx − Cyy)/2. (S42) spins,

0 0 0 1 where hΓ I mI |Tp (Iα)|ΓImI i is evaluated using Wigner- The contributions from nuclear spin rotation couplings Eckart and the reduced matrix elements listed in eqs. to the K-doubling can be extracted from the anisotropy k (S17)–(S19). The Tp (Cα) tensors can be obtained with in the T (C) tensors. As the radium spin-rotation cou- the appropriate parameterization listed in eqs. (S17)– pling is diagonal in |Γ,IH , mIH i, the Hnsr-M Hamiltonian (S18). does not couple oﬀ-diagonally between |E+i and |E−i hydrogen spin states, and therefore does not contribute to Relativistic ab initio values for the T (C) nuclear spin- the breaking of K-degeneracy. By contrast, the hydrogen rotation tensors are also obtained with a four-component spin-rotation coupling does couple |E i and |E i hydro- LRESC approach [53, 54] in the DIRAC19 code [56] using + − gen spin states. The computed anisotropic contribution a dyall.v4z basis set [57]. Correlation eﬀects are treated to K-doubling is |T −T | ∼ 0.3 kHz, as noted in Table at the level of DFT with the local density approximation xx yy I of the main text. and popular hybrid functions (B3LYP [58], PBE0 [108]). The full results are listed in Table S3. While validating these methods is challenging, due to the absence of experimental data (particularly for atoms heavier than group Nuclear Spin Dipolar Coupling 4), benchmarks with diatomic species in refs. [54, 55] suggests that these predictions are accurate within 10 For coupling two generic spins I1 and I2, the matrix percent. element is 14

(θ, φ) parameterize the vector from I1 to I2. Meanwhile, 2 T (I1, I2) can be decomposed as µ γ γ 2 I · I 3(I · r)(I · r) H = 0 1 2~ 1 2 − 1 2 (S45) II 4π r3 r5 √ X 2 1 1 √ 2 T 2(I , I ) = (−1)p 5 T 1 (I )T 1 (I ) . µ0γ1γ2~ 2 2 p 1 2 p1 1 p2 2 = − 6 T (C) · T (I1, I2) (S46) −p p1 p2 4π p1,p2 2 (S48) √ µ0γ1γ2 X = − 6 ~ T 2(C)T 2 (I , I ), (S47) 4π p −p 1 2 p Let us consider the spin-spin interaction between the 225Ra nucleus and the hydrogen nuclei. We where µ0 is the vacuum permeability and γ1,2 are the write the matrix elements in the decoupled basis 2 nuclear gyromagnetic ratios. Tq (C) is the spherical |N, K, S, J, mJ i|IM , mIM i|Γ,IH , mIH i and then evalu- 2 −3 harmonic tensor hCq (θ, φ)r i, where the polar angles ate the spherical harmonic and nuclear spin-spin tensors

0 0 0 0 0 2 hN K S J mJ |Tp (Cα)|NKSJmJ i 0 0 0 J 2 J 0 0 NJS J −mJ J+N +S +2p 0 =(−1) 0 δS0,S(−1) (2J + 1)(2J + 1) 0 0 −mJ p mJ J N 2 0 0 X 2 2 × hN K | Dpq(ω) · Tq (Cα)|NKi (S50) q 0 0 0 J 2 J 0 0 NJS J −mJ J+N +S +2p 0 =(−1) 0 δS0,S(−1) (2J + 1)(2J + 1) 0 0 −mJ p mJ J N 2 0 X 0 0 N 2 N × (−1)N −K p(2N 0 + 1)(2N + 1) T 2(C ) (S51) −K0 q K q α q 0 0 0 0 0 2 hIM mIM ;Γ IH mIH |Tp (IM , Iα)|IM mIM ;ΓIH mIH i √ p X 0 0 1 0 0 0 1 2 1 1 =(−1) 5 hIM mIM |Tp (IM )|IM mIM ihΓ IH mIH |Tp (Iα)|ΓIH mIH i . (S52) 1 2 −p p1 p2 p1,p2

2 The molecule-frame Tq (Cα) reduced matrix element can be evaluated using the appropriate sums of spherical 3 r X 4π harmonics that parameterize the vector between the ith T 2(C ) = Y 2(θ , φ )r−3, (S53) q 0 5 q i i hydrogen and the radium atom, i=1 r4π T 2(C ) = Y 2(θ , φ ) + e±2πi/3Y 2(θ , φ ) q ± 5 q 1 1 q 2 2 ∓2πi/3 2 −3 + e Yq (θ3, φ3) r . (S54) By convention [89], the Cartesian form of the single- hydrogen dipolar tensor T1 (of the form eq. (S11)) is taken to be traceless tr[T ] = 0 and symmetric, which 15 yields the following simplifications from eqs. (S22)– spin-spin interaction between the hydrogen spins them- (S23), which are invoked in tableI in the main text: selves are only non-zero between the “ortho” stretched √ spin states of A character. This term can therefore be 2 αdip · T0 (C0) = 3Tzz/ 6, (S55) ignored in the |N, |K| = 1i manifold. 2 αdip · T±(C∓) = ∓Txz, (S56) α · T 2 (C ) = (T − T )/2. (S57) dip ±2 ± xx yy SCHIFF MOMENT where√ we define a scaling parameter αdip = 2 Effective molecular sensitivity − 6µ0γHγRa~ /4π. The radium-hydrogen spin-spin interaction couples the opposite symmetry states |E+i and |E−i. This The effective Schiff moment sensitivity is proportional leads to |Txx − Tyy| ∼ 0.4 kHz K-doubling terms to hIM · ˆni, where IM is the metal spin andn ˆ is the inter- which couple the |N, K, S, J, mJ i|IM , mIM i|IH , mIH i nuclear axis [27]. In the decoupled basis, this is written and |N, −K, S, J, mJ i|IM , mIM i|IH , mIH i states. The as:

0 0 0 J+N +S +1p 0 NJS 0 × δ 0 (−1) (2J + 1)(2J + 1) hN |ˆn|Ni S ,S J 0 N 0 1 0 I0 −m0 IM 1 IM p = δ 0 δ 0 (−1) M IM δ 0 I (I + 1)(2I + 1) IH ,IH mIH ,mIH 0 IM ,IM M M M −mIM 0 mIM 0 0 0 0 J 1 J 0 0 NJS J −mJ J+N +S +1p 0 × (−1) δS0,S(−1) (2J + 1)(2J + 1) 0 0 −mJ 0 mJ J N 1 0 0 0 N 1 N × (−1)N −K p(2N 0 + 1)(2N + 1) (S58) −K0 0 K

In zero field, there is no defined orientation of the which was expressed in terms of πNN vertices, was re- molecule, and the Schiff moment sensitivity is zero. In parameterized by ref. [27] in terms of QCD θ¯ and the ˜ ˜ the decoupled limit (& 1 V/cm), the value of hIM · ˆni is quark-chromo EDMs du and dd: K · m · m · 1/2 for the N = 1 and |K| = 1 manifold. N IM 225 ¯ 3 The 1/2 factor arises from the K/N(N + 1) pre-factor to |S( Ra)| = 1.0θ e fm (S59) 4 ˜ ˜ 2 the Stark energy. The states with “stretched” sensitivity = 10 (0.50du − 0.54dd) e fm (S60) therefore take on the values hIM · ˆni = ±1/4. The energy shift resulting from a Schiff moment S and a coupling constant Ws is Hsm = Ws (IM · ˆn) |S|/|I|. For BSM Sensitivity a differential measurement performed between states of opposite effective sensitivity (hIM · ˆni = ±1/4), we arrive The frequency sensitivity of a spin-precession measure- at ment on a molecule or atom with coherence time τ and √ ¯ 2~δω repetition n is expressed as δω = [τ n]−1. In the main δθ = ¯ (S61) WS(S/θ) 225 + text, we assume a single trapped RaOCH3 ion with 5 ¯ 3 sec coherence time, which√ provides a frequency sensitiv- where S/θ = 1.0 e fm and the factor of two arises from −1 + ity of δω = 7.5 mrad s / hour. ∆[hIM ·ˆni] = 1/2. Given the estimate of Ws(RaOCH3 ) ≈ 225 4 Dobaczewski and Engel calculated |S( Ra)| in the 30, 000 a.u. [27] (where a.u. ≡ e/4π0a0), we have framework of T,P -violating interactions between nu- δθ/δω¯ = 24 × 10−8 s. Two weeks of data taking would cleons mediated by a pion using a variety of Skyrme thus result in the model-dependent sensitivity of δθ¯ < energy functionals [15]. The average of their results, ×10−10. 16

This discussion only considers in detail the limit on 2 θ¯ as an example, though the other quantities to which Schiﬀ moments are sensitive are also of interest. Due to 1 the many possible hadronic sources, measurements in dif- 0 ferent systems are needed to obtain robust bounds [109] -1

( D ) Moment Dipole -2

0 10 20 30 40 50 Electric Field(mV/cm) 30

20 SPIN ZERO ISOTOPOLOGUES 10

0 We brieﬂy consider the hyperﬁne structure for 226 + -10 RaOCH3 in the |N = 1, |K| = 1i manifold, which yields 12 unique levels. This is the isotopologue recently

( kHz ) Shift Stark -20 produced in an ion trap [35], and could serve as a valuable platform for spectroscopy and development. As the -30 0 5 10 15 226Ra nucleus is spin-0, the only hyperfine term to con- Electric Field(mV/cm) sider is the spin-rotation interaction between the ortho- hydrogen state and the molecular rotation, and the hy- Figure S2. Lab frame dipole moment (top) and Stark shifts perfine basis reduces to |N, K, S, J, mJ i|Γ,IH , mIH i with (bottom) of hyperfine states of the |N = 1, |K| = 1i manifold the radium spin terms omitted. Fig. S2 shows the Stark 226 + in the fully polarized limit for RaOCH3 . High, low, and shifts and dipole moments for the molecule up to the fully no field seekers correspond to states with negative, positive, polarized limit, and fig. S3 shows the level structure and zero dipole moment (K × mN = +1, 0, and −1). The grouped by the projection of total angular momentum jumps indicate avoided crossings between the Stark states. (mF = mN + mIH ) and their Stark manifold (K × mN ). 17

K × mN HNSR

−1 11 12 ~4.5 KHz 9 10 +1.2 MHz

7 8 ~10 Hz 0 6 5

−1.2 MHz

+1 3 4 ~4.5 KHz 1 2 –³⁄₂ +½–½ +³⁄₂

Figure S3. Level structure for the 12 hyperﬁne states of 226 + RaOCH3 in the |N = 1, |K| = 1i manifold in the decoupled regime (1 V/cm), grouped by the projection of total angular momentum mF = mN + mIH and their Stark manifold (K × mN ). The states are numbered in order of energy.

Method: LRESC(B3LYP) LRESC(PBE0) LRESC(LDA) LRESC(DHF) Basis: dyall.v4z dyall.v4z dyall.v4z dyall.v4z NSR (H) T (Cnsr)zz 15.27503794 15.23988653 15.05936414 15.54351836 T (Cnsr)yy 0.29172044 0.29285025 0.27717575 0.31026183 T (Cnsr)xx -0.00937583 -0.00725048 -0.02242052 0.01481825 T (Cnsr)zy -0.00000000 -0.00000000 -0.00000000 -0.00000000 T (Cnsr)yz -0.00000012 -0.00000012 -0.00000010 -0.00000014 T (Cnsr)yx -0.00000005 -0.00000005 -0.00000005 -0.00000005 T (Cnsr)xy 0.00000005 0.00000005 0.00000005 0.00000005 T (Cnsr)xz 253.81827070 253.90232417 254.10917668 253.71558497 T (Cnsr)zx -3.38051883 -3.37994662 -3.37277269 -3.38404825 NSR (Ra) T (Cnsr)xx 3.67255479 3.52587436 4.86603582 1.71006450 T (Cnsr)yy 2.41881565 2.49385127 3.24531289 1.35943566 T (Cnsr)xx 2.41881768 2.49385648 3.24531347 1.35945538 T (Cnsr)zy -0.00000003 -0.00000004 -0.00000005 -0.00000002 T (Cnsr)yz -0.00000275 -0.00000284 -0.00000369 -0.00000155 T (Cnsr)yx 0.00000000 -0.00000000 -0.00000000 0.00000000 T (Cnsr)xy 0.00000000 -0.00000000 0.00000000 -0.00000000 T (Cnsr)xz -0.00000158 -0.00002529 0.00000641 -0.00013274 T (Cnsr)zx 0.00000009 -0.00000060 -0.00000024 0.00000726

Table S3. Nuclear spin-rotation constants computed with optimized geometries from tableI. Units in kHz. 18 4 4 4 4 4 4 4 4 4 4 4 4 / / / / / / / / / / / / 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 − +1 +1 − +1 − +1 − − +1 +1 − Schiff Sensitivity 2 2 2 2 / / / / i i i i 2 2 2 2 / / / / +1 +1 +1 +1 , , , , − − − − E E E E i| i| i| i| 2 2 2 2 / / / / 1 1 1 1 i| − i| − i| − i| − 0 0 0 0 , , , , 2 2 2 2 + 1 + 1 + 1 + 1 2 2 2 2 √ √ √ √ | | / / / / 2 are omitted for brevity. √ √ √ √ + + / / / / / i i i i i − | i i i − | 2 2 2 2 i i i i 2 2 2 2 / / / / 2 2 2 2 / / / / = 1 1 1 / / / / manifold in the high-field regime (1 V/cm). The mixings 1 1 1 1 +1 +1 − − 1 1 H i − − − − , , , , I − − +1 +1 , , , , − − − − , , , , − − − − E E E E − − − − = 1 E E E E | i| i| i| i| E E E E i| i| i| i| 2 2 2 2 i| i| i| i| 2 2 2 2 K / / / / i i i i i i i i i i i i | 2, and 2 2 2 2 / / / / 1 1 2 2 2 2 2 2 2 2 2 2 2 2 , / / / / / / / / / / / / / / / / / 1 1 + 1 + 1 1 1 1 1 1 1 + 1 + 1 + 1 + 1 = 1 = 1 + 1 + 1 +1 +1 − − +1 − +1 − +1 − − +1 i| i| i| − i| − , , , , , , i| i| i| i| , , , , , , 1 1 i| i| − i| − i| 0 0 0 0 N M + + + + + + − − − − − − +1 0 0 0 − − +1 0 | , , , , I , , , , , , , , E E E E E E E E E E E E i| i| i| i| i| i| i| i| i| i| i| i| + 1 + 1 + 1 + 1 2 2 2 2 2 2 2 2 2 2 2 2 + 1 + 1 + 1 + 1 + 1 + 1 + 1 + 1 | | / / / / / / / / / / / / = 1, | | | | 1 1 1 1 1 1 in the + + + + + + N i − | i − | i i Mixing + 1 + 1 + 1 + 1 + 1 + 1 + 3 i i − | i i − | i − | i i i − | 2 2 2 2 i| − i| i| − i| i| − i| 2 2 2 2 2 2 2 2 i| i| − i| i| − i| i| − / / / / 1 1 1 1 / / / / / / / / 1 1 1 1 1 1 1 − − − +1 − +1 +1 +1 +1 − − − +1 +1 +1 +1 , , , , , , , , , , , , +1 +1 +1 − +1 − − − , , , , 1 1 1 1 1 1 , , , , , , , , + + + + RaOCH + + + + + + + + E E E E + 1 + 1 + 1 + 1 + 1 + 1 E E E E E E E E | − | − | − | − | − | − | | | | | | i| i| i| i| 225 i| i| i| i| i| i| i| i| 2 2 2 2 2 2 2 2 2 2 2 2 / / / / / / / / / / / / 1 1 1 1 1 1 1 1 + 1 + 1 + 1 + 1 i| − i| − i| − i| − i| i| i| i| − i| i| − i| − i| − 0 0 0 0 1 1 0 0 0 0 , , , , , , , , 1 1 1 1 − − +1 +1 1 1 1 1 , , , , 1 1 1 1 . The quantum numbers | − | − i | − | − | − | − + + | − | − | − | − IH i i − | − i i − | − 2 2 2 2 / / / / , m 1 1 1 1 Γ − − − − i| , , , , + + + + IM E E E E m i| i| i| i| i| 2 2 2 2 / / / / N + 1 + 1 + 1 + 1 K, m i| i| i| i| | 0 0 0 0 , , , , 1 1 1 1 | − | − | − | − 31 Hz 287 Hz 612 Hz 584 Hz . . . . 367 Hz 367 Hz 466 Hz 466 Hz . . . . 2545 MHz 2545 MHz 2515 MHz 2515 MHz Energy 24926 MHz 24926 MHz 24781 MHz 24781 MHz . . . . 932 . . . . 932 619 619 24705 MHz 24705 MHz 25004 MHz 25004 MHz 25227 MHz 25227 MHz 25371 MHz 25371 MHz 1 1 1 1 ...... 619 619 932 932 1 1 1 1 − − − − 1 1 1 1 1 1 1 1 − − − − − − − − 2 3 1 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 State are written in the decoupled basis Table S4. Energy, mixings, and Schiff sensitivity of hyperfine states of