---
title: Yb(trensal) Molecular Nanomagnet Quantum Platform
url: https://www.emergentmind.com/topics/yb-trensal-molecular-nanomagnet
type: topic
---

# Yb(trensal) Molecular Nanomagnet Quantum Platform

Searching arXiv for recent and foundational papers on Yb(trensal) molecular nanomagnets.
Yb(trensal) is a lanthanide molecular nanomagnet whose low-energy spectrum supports several complementary quantum-information descriptions: an effective electronic spin-\(\tfrac12\) qubit, a hyperfine-coupled \(^{173}\)Yb nuclear spin-\(\tfrac52\) qudit, a full 12-level electronuclear manifold, and a clock-transition spin system with coherent electric-field tunability. In the literature it is treated as a prototypical platform for molecular quantum information processing because the same molecule combines resolved transitions, long nuclear coherence, fast radio-frequency addressability enabled by electron–nuclear mixing, compatibility with superconducting resonators, and experimentally realistic protocols for higher-dimensional entanglement tests [1909.02374], [2005.01029], [2203.00965], [2507.22768].

## 1. Molecular identity and effective Hilbert spaces

[Yb(trensal)] is a single-ion lanthanide molecular complex containing ytterbium(III) coordinated by the trensal ligand. Within the \(\mathrm{Ln(trensal)}\) family it is described as an isostructural member with Yb\(^{3+}\) at a nearly axial coordination site, and its electronic ground doublet is well isolated from excited crystal-field states. In the low-energy regime the molecule is therefore modeled as an effective \(S=1/2\) electronic degree of freedom. For the isotope \(^{173}\)Yb, the nucleus has \(I=5/2\), so the nuclear sector contains \(2I+1=6\) levels, while the combined electronuclear manifold contains \((2S+1)(2I+1)=12\) unequally spaced spin levels [1909.02374], [2203.00965].

Different experimental and theoretical studies exploit different subspaces of this same molecule. In coherent NMR control, the emphasis is on the six nuclear levels coupled to the electronic doublet. In hybrid cavity experiments, the full 12-level electronuclear structure is treated as a \(d=12\) qudit. In generalized Bell tests, four selected nuclear-spin levels are used as a qudit of dimension 4, i.e. a spin-\(3/2\) subspace, coupled to the electronic qubit [2507.22768].

| Effective description | Dimensionality | Role in the literature |
|---|---:|---|
| Electronic ground doublet | \(2\) | Electronic qubit or ancilla |
| \(^{173}\)Yb nuclear manifold | \(6\) | Nuclear qudit |
| Full electronuclear manifold | \(12\) | \(d=12\) qudit in circuit-QED |
| Selected nuclear subspace | \(4\) | Spin-\(3/2\) qudit for Bell tests |

A recurrent source of ambiguity is that Yb(trensal) can be described either as a qubit or as a qudit system. The published work resolves this by assigning the label according to the active subspace: effective \(S=1/2\) when only the electronic doublet is used, a six-level qudit when the \(^{173}\)Yb nuclear manifold is addressed, and a 12-level qudit when the full electronuclear spectrum is relevant [1909.02374], [2203.00965].

## 2. Spin Hamiltonian, anisotropy, and electron–nuclear mixing

For the coupled electronic and nuclear degrees of freedom, the effective Hamiltonian is
$$
H = A_\parallel S_z I_z + A_\perp (S_x I_x + S_y I_y) + p I_z^2
    + \mu_B \mathbf{S}\cdot \mathbf{g}\cdot \mathbf{B}
    + \mu_N g_I \mathbf{I}\cdot \mathbf{B}.
$$
Here \(S=1/2\) is the effective electronic spin, \(I=5/2\) is the \(^{173}\)Yb nuclear spin, \(A_\parallel\) and \(A_\perp\) are the longitudinal and transverse hyperfine couplings, \(p I_z^2\) is the nuclear quadrupolar term, and the remaining terms are the electronic and nuclear Zeeman interactions. Parameters extracted from NMR and transition frequencies are
$$
g_x=g_y=2.9,\qquad g_z=4.3,
$$
$$
A_\parallel=-0.02993~\mathrm{cm^{-1}},\qquad A_\perp=-0.0205~\mathrm{cm^{-1}},
$$
$$
g_I=-0.2592,\qquad p=-0.0022~\mathrm{cm^{-1}}.
$$
The quadrupolar term \(p\) is identified as essential to reproduce the NMR data [1909.02374].

At fields along the molecular \(C_3\) axis and sufficiently large \(g_z\mu_B B\), diagonalization gives electron–nuclear eigenstates that are close to product states but retain a finite admixture induced by the transverse hyperfine interaction. The paper expresses this as
$$
|\psi_{m_S,m_I}\rangle \propto |m_S,m_I\rangle \pm \frac{\alpha A_\perp\,|m_S\mp1,m_I\pm1\rangle}
     {g_z\mu_B B + (m_I \pm 1/2)A_\parallel}.
$$
The physical consequence is that nominally nuclear transitions acquire an electronic contribution to their matrix elements. For adjacent nuclear levels,
$$
|\psi_{m_S,m_I}\rangle \rightarrow |\psi_{m_S,m_I\pm1}\rangle,
$$
the transition matrix element is approximately proportional to
$$
g_I\mu_N \pm \frac{A_\perp}{2B}.
$$
The mixing-induced term can be about \(10^3\) times larger than the purely nuclear one, and even a modest wavefunction mixing of order \(\sim 0.08\) at \(B=0.15\) T is reported to enhance the RF transition strength dramatically [1909.02374].

This electron-assisted nuclear addressability is central to the molecule’s role as a hybrid quantum element. The nuclear spin is naturally noise-resilient, whereas the electron spin supplies speed and control. The literature explicitly frames the mixing not as a complication to be eliminated, but as a resource for fast qudit manipulation [1909.02374].

## 3. Coherent nuclear-qudit control and embedded logical encoding

Coherent control was demonstrated on a single crystal of \(^{173}\)Yb(trensal) diluted \(2\%\) into \(\mathrm{Lu(trensal)}\), at \(T=1.4\) K. The dilution suppresses unwanted intermolecular dipolar interactions. The experimental toolbox comprised NMR spectroscopy to characterize transition frequencies and fit Hamiltonian parameters, Hahn echo to measure \(T_2\), CPMG sequences for dynamical decoupling, and transient nutation or Rabi-oscillation measurements to establish coherent manipulation. Field orientations \(\theta=0^\circ,\ 81^\circ,\ 90^\circ\) relative to the molecular \(C_3\) axis were studied [1909.02374].

The key dynamical result is that RF \(\pi\) pulses can be executed in only a few hundred ns, which is unusually fast for a nuclear-spin system. Rabi oscillations of the nuclear magnetization are described by
$$
\propto e^{-t/\tau}\sin(2\pi \nu_R t),
$$
with \(\nu_R\) scaling linearly with the RF amplitude \(B_1\). The measured enhancement of \(\nu_R\) agrees with calculations based on the effective Hamiltonian, and the Rabi frequency is about \(10^3\) larger than in the absence of transverse hyperfine mixing. Nuclear coherence times \(T_2\) from Hahn-echo measurements increase with field, a trend attributed to reduced electron–nuclear mixing at higher field, stronger electron polarization, reduced spin-flop processes, and residual nuclear or electronic dipolar interactions in the environment. CPMG further enhances \(T_2\) as the number of refocusing pulses \(n\) increases [1909.02374].

The six nuclear levels are also used to encode a logical qubit protected against basic amplitude-shift errors. With the electron initialized in \(m_S=-1/2\), the logical state is
$$
|\Phi\rangle = \alpha |\psi_{-1/2,-3/2}\rangle + \beta |\psi_{-1/2,3/2}\rangle.
$$
An unwanted \(\Delta m_I=+1\) shift produces
$$
|\Phi_+\rangle = \alpha |\psi_{-1/2,-1/2}\rangle + \beta |\psi_{-1/2,5/2}\rangle,
$$
while a \(\Delta m_I=-1\) shift produces
$$
|\Phi_-\rangle = \alpha |\psi_{-1/2,-5/2}\rangle + \beta |\psi_{-1/2,1/2}\rangle.
$$
Because these error spaces do not overlap with the code space, the error is detectable. Detection is performed by two simultaneous microwave \(\pi\)-pulses resonant with
$$
|\psi_{-1/2,-1/2}\rangle \leftrightarrow |\psi_{1/2,-1/2}\rangle,\qquad
|\psi_{-1/2,5/2}\rangle \leftrightarrow |\psi_{1/2,5/2}\rangle.
$$
If the system occupies the error space, the electron flips; otherwise it does not. The electronic ancilla can then be measured and reset, and the nuclear error corrected by RF pulses. The same paper stresses that \(I=5/2\) is exactly the minimal dimension needed to make the code space and the two error spaces non-overlapping, so the six-level structure realizes a minimal qudit code. It further shows that rotations about logical \(x\) and \(y\) can be implemented by RF pulse sequences on the \(-3/2\leftrightarrow-1/2\), \(-1/2\leftrightarrow1/2\), and \(1/2\leftrightarrow3/2\) gaps, while a logical \(z\)-rotation is obtained by a single microwave \(2\pi\) pulse semiresonant with \( |\psi_{-1/2,3/2}\rangle \leftrightarrow |\psi_{1/2,3/2}\rangle\) [1909.02374].

## 4. Clock transitions and coherent spin-electric control

A distinct line of work uses Yb(trensal) as a molecular nanomagnet with a clock transition, meaning a transition whose energy is to first order independent of magnetic field:
$$
\left.\frac{\partial \omega}{\partial B_0}\right|_{\mathrm{CT}} = 0.
$$
Near such a transition, magnetic dephasing is suppressed, but the transition frequency is dominated by a tetragonal anisotropy term in the spin Hamiltonian. In the reported physical picture, this anisotropy originates from a slight off-centering or structural distortion that is electrically polarizable, so the clock-transition energy becomes electrically tunable [2005.01029].

The molecule is not treated as perfectly symmetric. The actual crystal structure deviates slightly because of counterions and lattice water. Two distortions are emphasized: the skew angle \(\theta\) differs slightly from the ideal \(45^\circ\), and the lanthanide ion is slightly off center by
$$
d=\frac{h-h'}{2}.
$$
This breaks inversion symmetry, generates an electric dipole moment, and is associated with a small continuous-shape deviation \(S<0.1\) around the lanthanide center. The same distortion creates the tetragonal anisotropy term \(B_4^4\) that opens the anticrossing and defines the clock transition. The causal chain is stated explicitly as electric field \(\to\) molecular distortion \(\to\) anisotropy change \(\to\) transition-frequency shift [2005.01029].

The spin-electric coupling is linear,
$$
\delta f \propto E,
$$
with an experimentally extracted coupling constant
$$
\frac{\delta f}{E} = 11.4 \pm 0.3~\mathrm{Hz\,Vm^{-1}}.
$$
The work also relates the effect to a modulation
$$
\delta B_4^4/h = 8.8\pm0.2 \times 10^{-3}~\mathrm{MHz},
$$
equivalent to
$$
5.9\times 10^{-2}~\mathrm{Hz\,Vm^{-1}}.
$$
A modest field of order \(10^5~\mathrm{V/m}\) is reported to shift the transition by more than its natural linewidth. Experimentally, a Hahn-echo ESR sequence,
$$
\pi/2 - \tau - \pi - \tau - \mathrm{echo},
$$
was augmented with an electric-field pulse during one free-evolution interval. The crystal was mounted between parallel metal plates separated by \(2~\mathrm{mm}\), with \(\pi/2 = 32~\mathrm{ns}\) and \(\pi = 64~\mathrm{ns}\) microwave pulses. The induced phase accumulation \(\Delta\phi=\delta f\,t_E\) generates echo oscillations with a period about \(1~\mu\mathrm{s}\) [2005.01029].

The flat quadrature channel is identified as evidence for linear rather than quadratic spin-electric coupling. Because the two molecules in the unit cell are inversion-related, linear coupling shifts their ESR frequencies in opposite directions, so their quadrature contributions cancel. The same symmetry is exploited for selective addressing: when the refocusing frequency satisfies \(f_r=f_0\pm\delta f\), the \(\pi\) pulse refocuses one or the other inversion-related subpopulation, and each echo peak has about half the amplitude of the zero-field echo. The same study also notes an important tradeoff: operating at the clock transition suppresses magnetic dephasing, but increases sensitivity to electric-field fluctuations and structural distortions [2005.01029].

## 5. Coupling to superconducting resonators and circuit-QED relevance

Yb(trensal) has also been integrated into a hybrid circuit-QED architecture by placing magnetically diluted single crystals on the inductors of superconducting lumped-element \(LC\) resonators. The on-chip device contains 8 superconducting LERs with different bare resonance frequencies \(f_r\), all inductively coupled to a common superconducting transmission line used for readout. The external magnetic field is applied parallel to the crystal \(C_3\) axis and to the transmission line. Experiments were performed at \(10\) mK with typical input power \(-95\) dBm, corresponding to about \(10^8\) photons in the resonator, far below the \(\sim 10^{16}\) participating spins. Crystals with Yb concentrations between \(1.9\%\) and \(10\%\), diluted in [Lu(trensal)], were studied [2203.00965].

The frequency layout separates electronic and nuclear transitions. Resonators around \(2.93\) GHz couple to electronic-spin transitions, while resonators around \(403.6\), \(414.3\), \(548.5\), and \(549.5\) MHz couple to \(^{173}\)Yb nuclear transitions. Broadband spectroscopy through the transmission line resolves the Zeeman diagram and shows resonances from \(^{171}\)Yb, \(^{173}\)Yb, and \(I=0\) isotopes, with signal intensities following the natural abundances \(14\%\), \(16\%\), and \(70\%\), respectively. The broadband line geometry yields an electronic transition half-width of about \(\gamma\sim23\) MHz [2203.00965].

The circuit-QED result is the achievement of high cooperative coupling to all electronic and most nuclear spin transitions. At \(2.93\) GHz and \(5\%\) Yb concentration, the reported values are \(G=7.9\) MHz, \(\gamma=15.9\) MHz, cooperativity \(C=129\) for the \(^{171}\)Yb \(0\to3\) transition; \(G=22.0\) MHz, \(\gamma=16.8\) MHz, \(C=946\) for the \(I=0\) \(0\to1\) transition; and \(G=5.0\) MHz, \(\gamma=17.7\) MHz, \(C=46\) for the \(^{173}\)Yb \(0\to11\) transition. For nuclear transitions, the paper explicitly demonstrates coupling to \(0\to1\), \(1\to2\), \(2\to3\), \(7\to8\), and \(8\to9\). With a main-text nuclear resonator linewidth \(\kappa\simeq3.78\) kHz, it reports \(C_{0,1}=2.0\), \(C_{1,2}=9.1\), and \(C_{2,3}=4.9\). The strongest nuclear coupling is obtained at \(f_r=403.3\) MHz and \(8\%\) Yb concentration, where
$$
G_{1,2}=0.43~\mathrm{MHz},\qquad \gamma_{1,2}=2.3~\mathrm{MHz},\qquad C_{1,2}=24.0.
$$
This is described as close to the cooperativities obtained for electronic transitions [2203.00965].

The significance of these results lies in the fact that nuclear spins are usually difficult to integrate with superconducting circuits because of their weak coupling to external stimuli. In Yb(trensal), hyperfine coupling to the electronic spin amplifies the response of the nuclear spin states to microwave fields, making molecular qubits and qudits accessible within a resonator architecture [2203.00965].

## 6. Higher-dimensional entanglement and Bell-inequality protocols

A 2025 study proposes Yb(trensal) as an experimentally realistic testbed for generalized Bell inequalities in a qubit–qudit system. The working model uses the Hamiltonian
$$
H = g_S \mu_B B_z S_{z} + g_I \mu_N B_z I_{z} + A_{\perp} (S_{x} I_{x} + S_{y} I_{y}) + A_{\parallel} S_{z} I_{z} + p I_{z}^2 ,
$$
with \(g_S=(2.9,2.9,4.3)\), \(g_I=-0.2592\), \(A_{\parallel}=-883\) MHz, \(A_{\perp}=-628\) MHz, \(p=-66\) MHz, and \(B_z=0.3\) T along the molecular \(z\)-axis. At this field the eigenstates are almost factorized into electronic and nuclear parts, which the authors identify as ideal for preparing and reading out entangled qubit–qudit states. The proposal relies on measured nuclear coherence \(T_{2n}\gtrsim500~\mu\mathrm{s}\), with \(T_{2n}=560~\mu\mathrm{s}\) used as the shortest value in the simulations, and on electronic coherence \(T_{2e}\sim2.4~\mu\mathrm{s}\) at \(2.7\) K [2507.22768].

The generalized CHSH operator is
$$
\hat{O}_{\rm Bell} = A \otimes (B + B') + A' \otimes (B - B'),
$$
where \(A,A'\) are qubit observables and \(B,B'\) are qudit observables with eigenvalues \(\pm1\). For the target qubit–qudit state in the \((\tfrac12-\tfrac32)\) sector, the numerically optimized maximum is
$$
\langle \psi | \hat{O}_{\rm Bell} | \psi \rangle_{\max} = 2.64575.
$$
The paper rewrites the Bell expectation value in terms of diagonal spin observables after local unitaries and emphasizes that the protocol can be implemented using ensemble expectation values rather than projective single-shot measurements. State preparation from the ground state \(\ket{\downarrow,-\tfrac52}\) uses five resonant pulses, and the total preparation time is about \(0.5\)–\(1~\mu\mathrm{s}\) depending on amplitude [2507.22768].

Under a Lindblad pure-dephasing model, fidelities above \(0.95\) are obtained for a broad range of parameters, and with the measured \(T_{2e}=2.4~\mu\mathrm{s}\) the fidelity remains above \(0.94\) at optimal pulse amplitude. GRAPE optimization is reported as essential: realistic Bell values are around \(2.53\) for \(T_{2e}=10~\mu\mathrm{s}\), \(2.47\) for \(T_{2e}=5~\mu\mathrm{s}\), and \(2.34\) for \(T_{2e}=2.4~\mu\mathrm{s}\), whereas hand-optimized square pulses yield much lower values around \(1.32\)–\(1.77\). The same work proposes a trimer architecture \(\left(\frac32-\frac12-\frac32\right)\) for qudit–qudit entanglement, with a central spin-\(\tfrac12\) ancilla acting as a switchable mediator. For the \(d=4\) CGLMP inequality, the ideal theoretical maximum is \(I_{\max}=2.89624\), and values up to \(I\approx2.61\) are obtained for \(T_2=30~\mu\mathrm{s}\), with entangled-state fidelities typically above \(0.91\) for \(T_2\) in the range \(5\)–\(30~\mu\mathrm{s}\) [2507.22768].

Taken together, these results place Yb(trensal) at the intersection of molecular magnetism, coherent control, hybrid quantum hardware, and high-dimensional quantum information. The published record supports a consistent picture: the molecule’s utility derives from the simultaneous presence of a fast electronic ancilla, a long-lived nuclear register, hyperfine-enabled control pathways, and spectroscopically resolvable multilevel structure [1909.02374], [2005.01029], [2203.00965], [2507.22768].

Source: https://www.emergentmind.com/topics/yb-trensal-molecular-nanomagnet