C078 derivation: weak-field reduction of drive (A) and identification of $H_s$

1. Drive-(A) Molmer-Sorensen Hamiltonian

From the main text (Eq. near line 115), the bichromatic drive (A) realises a Molmer-Sorensen interaction coupling the collective operator $S_{x,e} = \sigma_{x,\downarrow e}\otimes\mathbf{1} + \mathbf{1}\otimes\sigma_{x,\downarrow e}$ to a single motional mode $\hat a$: $$ H_A = \tfrac{1}{2}\,\hbar\,\eta\Omega\,S_{x,e}\,\big(\hat a\,e^{i\delta t} + \hat a^{\dagger} e^{-i\delta t}\big), $$ with $\eta$ the Lamb-Dicke parameter, $\Omega$ the carrier Rabi frequency and $\delta$ the detuning from the sidebands.

2. Exact closed-form propagator

The paper gives the exact propagator (Magnus expansion, exact for a linear force): $$ U_A(t) = \exp\!\big[\big(\alpha(t)\hat a^{\dagger}-\alpha^{*}(t)\hat a\big)S_{x,e}\big]\, \exp\!\big[i\,\Phi(t)\,S_{x,e}^{2}\big], $$ $$ \alpha(t) = -i\,\frac{\eta\Omega}{\delta}\,e^{-i\delta t/2}\sin\!\frac{\delta t}{2}, \qquad \Phi(t) = \frac{\eta^{2}\Omega^{2}}{4\delta^{2}}\big(\delta t - \sin\delta t\big). $$

3. Weak-field limit $\delta \gg \eta\Omega$

The displacement amplitude is bounded, $|\alpha(t)| = \frac{\eta\Omega}{|\delta|}\, |\sin(\delta t/2)| \le \eta\Omega/|\delta| \ll 1$. Hence in the regime $|\delta|\gg\eta\Omega$ the oscillator is only virtually excited and the spin-dependent displacement factor $\exp[(\alpha\hat a^{\dagger}-\alpha^{*}\hat a)S_{x,e}]$ can be neglected (the motion factorises out, $\hat a$ is never populated). The remaining evolution is generated purely by the $S_{x,e}^{2}$ term. Its instantaneous (secular) rate is the time-average of $\dot\Phi$: $$ \dot\Phi(t) = \frac{\eta^{2}\Omega^{2}}{4\delta^{2}}\big(\delta - \delta\cos\delta t\big) = \frac{\eta^{2}\Omega^{2}}{4\delta}\big(1-\cos\delta t\big) \;\xrightarrow{\text{time avg}}\; \frac{\eta^{2}\Omega^{2}}{4\delta}. $$ Equivalently, second-order adiabatic elimination of the virtually-excited oscillator from $H_A$ gives an effective Hamiltonian $\sim H_A \frac{1}{\delta} H_A$, $$ H^{\mathrm{eff}}A = \hbar\,J\,S, \qquad J = \frac{\eta^{2}\Omega^{2}}{4\delta} = \frac{(\tfrac{1}{2}\eta\Omega)^{2}}{\delta} = \frac{(\eta\Omega/2)^{2}}{\delta}. $$ This matches the claimed $J = (\tfrac{1}{2}\eta\Omega)^{2}/\delta$ exactly, and is identified in the paper as the effective frequency of the collective excitation $\ket{\downarrow\downarrow}\to\ket{ee}$.}^{2

4. Drive-(C) and detuning terms

Drive (C) is a global resonant rotation of the spin qubit $\{\downarrow,\uparrow\}$ with Rabi frequency $\Omega_C$; its generator on two ions is $$ \frac{\hbar\Omega_C}{2}\big(\sigma_x\otimes\mathbf 1 + \mathbf 1\otimes\sigma_x\big), $$ with $\sigma_x = \ket{\uparrow}\bra{\downarrow}+\ket{\downarrow}\bra{\uparrow}$ on the qubit subspace (this is exactly the generator of $U_C$ in the discrete model, P.sx_du). The spin-qubit detuning $\beta$ adds an energy $\hbar\beta$ to each $\ket{\uparrow}$: $$ \hbar\beta\big(\ket{\uparrow}\bra{\uparrow}\otimes\mathbf 1 + \mathbf 1\otimes\ket{\uparrow}\bra{\uparrow}\big). $$

5. Conclusion

Summing the three contributions reproduces Eq. (S9) exactly: $$ H_s = \hbar J S_{x,e}^{2} + \frac{\hbar\Omega_C}{2}\big(\sigma_x\otimes\mathbf 1+\mathbf 1\otimes\sigma_x\big) + \hbar\beta\big(\ket{\uparrow}\bra{\uparrow}\otimes\mathbf 1+\mathbf 1\otimes\ket{\uparrow}\bra{\uparrow}\big), $$ with $J=(\eta\Omega/2)^2/\delta$. The structure (collective $S_{x,e}^2$ coupling, drive-C rotation, spin detuning) and the value of $J$ are all confirmed.

Caveat: the reduction "neglect the oscillator displacement when $|\delta|\gg\eta\Omega$" is the standard adiabatic-elimination assumption stated in the paper (and ref. Kim2010); it is a physical-regime assumption rather than an unconditional identity. The $J=(\eta\Omega/2)^2/\delta$ rate is, however, an exact consequence of the paper's own closed-form $\Phi(t)$. Verdict: verified, with a note on the regime assumption.