{"id":"8f8a3c64-5b21-4330-a8e8-13d843c56782","arxiv_id":"2608.07929","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A single Er 4f spin is coherently controlled with electrical pulses via an anisotropic exchange-coupled Ti partner, reaching 190 MHz Rabi oscillations.","lead":"This paper reports all-electrical coherent control of a single erbium spin on a surface, using a nearby titanium atom to convert radio-frequency electric fields into an effective magnetic drive. The result reaches Rabi frequencies around 190 MHz, about ten times the previous rare-earth record, and points to anisotropic exchange as a general resource for electrically controlling shielded 4f spins.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 190 MHz Rabi demonstration is solid, but the claim that anisotropic-exchange rotation is the drive mechanism is underdetermined: δα0 is a free parameter and alternative electrical channels are not ruled out.","rationale":"The experimental core—Rabi oscillations on the Er-like transition f3 at 190 MHz, linear scaling with V_RF, conductance independence, and observation of a coherent DQT—constitutes strong evidence for all-electrical coherent control of an Er 4f spin. The single-spin character of the Er drive is supported by the order-of-magnitude enhancement over Ti and by the DQT's two-spin nature, which a direct single-spin g_Er modulation cannot drive at first order. However, the paper's second major claim—that the enabling mechanism is the rotational modulation of the anisotropic Er-Ti exchange—is not established at the same level. The model's key parameter δα0 is undetermined by any independent measurement or calculation; it is chosen to reproduce the absolute Rabi scale. The angular maps that validate the model compare it only against selected diagonal J-modulation terms, not against the full space of electrical driving channels. In particular, a Ti-only rotational trajectory is the natural interpretation of the text's own 'wobbling of μ_Ti' picture and yields a different driving operator, -δα S_Ti G_B J S_Er rather than δα S_Ti [J, G] S_Er; whether it reproduces the observed θ,φ dependencies is untested. Direct Er g-tensor modulation, while constrained by the absence of ESR on isolated Er(B), is not excluded in the pair because readout there is mediated by the Ti sensor. The proposed numerical comparison directly tests these alternatives and would settle whether the equal-rotation exchange mechanism is genuinely selected by the data. If it is not selected, the general-resource claim reduces to a special property of the Er(B) site, and the paper's framing should be scaled back accordingly. The reader's verdict of CONDITIONAL remains appropriate; the condition should be that the authors rule out at least the Ti-only rotation alternative and provide a physical or experimental bound on δα0, for example, from multiplet calculations of the orbital response to the RF electric field.","tokens_in":23620,"tokens_out":15564,"duration_ms":179718,"concrete_test":"Re-compute the ESR intensity and normalized Rabi angular maps (θ and φ) for the (3,0) and (2.5,-0.5) pairs using the same fitted Hamiltonian and the same δα0, but with two alternative drive terms: (a) Ti-only exchange rotation, H_drive = -δα0 cos(ωt) S_Ti G_B J S_Er (set A_Er = 0 in Eq. S13), and (b) direct Er g-tensor modulation, H_drive = -μ_B δα0 cos(ωt) S_Er [G_B, g_Er] B_ext. Compare the resulting f3 and f5 intensity and Rabi-frequency maps with the experimental data in Figs. 5b-d, S13, and S16. If either alternative reproduces the observed angular dependence within the same tolerance as the equal-rotation model, the uniqueness of the proposed mechanism collapses; if neither fits, the equal-rotation exchange-rotation model is strongly supported.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's central mechanistic claim rests on identifying the RF drive as a common rotational wobble of the Er and Ti magnetic-moment frames around B_ext, with amplitude δα0 = 0.1π set by hand in every simulation (Fig. S13, S14). This is load-bearing because the fitted g_Er and J_exc are constrained partly by normalized Rabi frequencies, which are independent of δα0; the absolute 190 MHz rate is not predicted, only matched by choosing δα0. More importantly, the comparison against diagonal J modulations (Figs. S9-S11) does not rule out two physically plausible alternative channels: (i) direct modulation of the Er g-tensor by the RF electric field, a single-spin operator H_drive = -μ_B S_Er · δg_Er(t) · B_ext that would also produce Er Rabi signals with angular anisotropy through g_zz = 9.59, and (ii) a Ti-only rotation of the exchange frame, i.e., A_Ti = δα G_B and A_Er = 0 in Eq. S13, which is arguably more natural given the RF field is localized under the tip on Ti. The equal-rotation assumption R_Ti = R_Er = R_B(δα) is asserted in Eq. S20 without physical justification, even though the text's picture in Fig. 5a attributes the drive to the wobbling of μ_Ti alone. Because the paper's general-resource claim depends on anisotropic exchange being the dominant electrical transduction channel, the mechanism should be regarded as plausible but not uniquely established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports ESR-STM experiments on individual Er-Ti pairs on MgO/Ag(100) and claims all-electrical coherent control of a single rare-earth 4f electron spin. The authors observe five ESR transitions of the exchange-coupled Er(B)-Ti pair, identify the Er-dominated transition (f3), and drive coherent Rabi oscillations at up to 190 MHz, an order of magnitude above previous rare-earth spin qubit results. The Rabi frequency scales linearly with RF voltage and, for f3, is independent of tunnel conductance, which supports a non-tip-field driving mechanism. Three pair geometries are compared, showing a non-monotonic dependence of the Er Rabi rate on pair separation. From angular ESR maps under three-dimensional field control, the authors extract the Er g-tensor (g_xx = 1.05, g_yy = 2.22, g_zz = 9.59) and an anisotropic exchange tensor (J_xx = 79.6, J_yy = 543.7, J_zz = 1272.5 MHz). The proposed driving mechanism is an RF-induced rotational wobble of both magnetic-moment frames around the external field, which converts the anisotropic exchange into an effective transverse field; the model is implemented with a commutator term δα μ_Ti[J_M, G_B] μ_Er, and its angular predictions are compared with the experimental maps. Multiplet calculations (4f11 5d2 configuration) with three fitted parameters reproduce the principal g-values, and spin-echo and recovery measurements yield T2 ≈ 113 ns and T1 ≈ 57 ns.","tokens_in":24005,"tokens_out":20509,"duration_ms":220373,"significance":"The experimental core is strong and clearly presented: clean Rabi oscillations at 190 MHz, linear V_RF scaling, conductance independence of the Er Rabi rate (Fig. 2e), a three-pair geometry series, and full (θ, φ) angular maps. If the proposed mechanism is correct, the anisotropic-exchange transducer is a genuinely transferable design resource for electrically driving shielded rare-earth spins, and the demonstrated rate is a real ten-fold advance. The theoretical framework is also elegant: the commutator form vanishes for isotropic J (Eq. S28), and the model yields falsifiable structural predictions, notably the vanishing of the drive at φ = 0° and 90° and a predicted ~30-fold suppression for the near-isotropic Er(O)-Ti pair. The principal weakness is that the validation of the specific drive hypothesis is partly circular and underdetermined: normalized Rabi data enter the spin-Hamiltonian fit that the model later reproduces; the absolute Rabi scale is set by the unconstrained amplitude δα0 = 0.1π; and the equal-rotation ansatz of Eq. (S20) is not distinguished from Ti-only rotation or direct g-tensor modulation.","major_comments":[{"comment":"Supplementary Section III states that the fit of g_Er, J_exc, and B_tip includes 'both the transition energies and the normalized Rabi frequencies,' with Fig. S13(c) identified as the normalization curve. Because the normalized Rabi frequencies entering the fit are computed from a driving operator, and the only driving operator developed in the supplement is the rotational-modulation Hamiltonian of Section IX (Eq. S27), the model's subsequent reproduction of the normalized Rabi angular dependence in Fig. 5(c,d) and Fig. S13(c) is, on the natural reading, partly guaranteed by the fitting procedure rather than an independent test. The authors should either refit the spin Hamiltonian using only the transition energies of Figs. 4(a,b) and report the predicted (not fitted) normalized Rabi curves, or explicitly identify which data were withheld from the fit. Without this, Section V's statement that off-diagonal rotational modulation 'successfully reproduces' the experimental trends overstates the independent validation content.","section":"Supp. III; Figs. 5(c,d), S13(c)"},{"comment":"In the drive term of Eq. (2) and Eq. (S21), δα(t) = δα0 cos(ωRF t) with δα0 = 0.1π adopted in all simulations (Figs. S13, S14, S19, S20). The angular maps used for model validation are normalized (Fig. S13(c) divides by the f3 Rabi frequency at θ = 90°), so they are invariant to δα0; the absolute 190 MHz rate is matched by hand rather than predicted. The paper should state this limitation explicitly and give an order-of-magnitude estimate of δα0 from the RF field at the junction and the calculated orbital response (e.g., the 5d occupation computed in Supp. VIII), or identify which future measurement would calibrate the transduction coefficient. As written, the Discussion's claim that the mechanism 'transduces electrical excitations into an effective coherent rotation' attributes to the model quantitative content it does not yet have.","section":"Sec. V, Eq. (2); Supp. IX Eq. (S21)"},{"comment":"Equation (S20) assumes R_Ti(t) = R_Er(t) = R_B[δα(t)], i.e., both local spin frames rotate rigidly about B_ext by the same angle, while the physically plausible alternatives are not computed: a Ti-only rotation (A_Er = 0 in Eq. S13), an Er-only rotation, and direct modulation of the Er g-tensor through H_drive = -μ_B S_Er · δg_Er(t) · B_ext. Given that the RF field is localized under the tip positioned on Ti, and that the paper's own schematic (Fig. 5a) describes the drive as the wobbling of the Ti moment alone, the equal-rotation ansatz is not the obvious default. The authors should justify it from the field geometry or estimate the relative RF amplitudes at the two sites, and should compare the angular predictions of equal rotation, Ti-only rotation, and the δg_Er channel against the datasets of Figs. 4 and 5 and Fig. S13 to show which mechanism the data can actually discriminate. As it stands, 'common rotational wobble' is one plausible choice among several, and the general-resource claim of the abstract rests on that untested choice.","section":"Supp. IX Eq. (S20); Fig. 5(a)"},{"comment":"Supplementary Section VIII fits three free parameters (α = 1.155, q_red = 0.5932, Δ = 6.6467 eV), together with an imposed 10-eV 5d crystal-field splitting, to reproduce the three principal values of the Er g-tensor, so the agreement in Fig. 4(c) is a fit rather than a prediction. The main text's phrasing in Section IV ('the calculated principal values of g-tensor ... qualitatively agree with the experimental results') should be revised to state that the multiplet calculation was tuned to the experimental g-values. The non-fitted outputs — in particular the ground-doublet decomposition with 79.8% |m_J = ±1/2⟩ weight in Table S1 and the hard-axis direction — are the legitimate predictive content of the calculation and should be presented as such, with the caveat that they are not yet directly measured.","section":"Supp. VIII; Fig. 4(c)"}],"minor_comments":[{"comment":"The measured Rabi frequency is 190 MHz (0.19 GHz), but the abstract and Discussion describe the result as 'near-gigahertz'; this overstates the demonstrated rate by roughly a factor of five and should be rephrased (e.g., 'hundreds of MHz').","section":"Abstract; Sec. VI"},{"comment":"Section V cites 'the fits in Section II.B' and 'Rabi frequencies shown in Section II.A'; these internal references do not match the manuscript structure (the relevant content is in Section III and in Methods), suggesting leftover cross-references from an earlier draft.","section":"Sec. V"},{"comment":"The fitted value of B_tip is never reported, and the tip-drive amplitude 0.17 S_Ti,x introduced in Supp. X is an additional adjustable parameter; both should be stated explicitly wherever the Fig. 5(b,d) agreements are presented.","section":"Supp. III; Supp. X"},{"comment":"No data availability statement is included; given the number of fitted parameters and simulations, depositing the raw angular ESR and Rabi datasets and the simulation scripts would substantially improve reproducibility.","section":"Methods"},{"comment":"The abstract's 'ten-fold improvement over the present record for rare-earth spin qubits' is broader than the main-text comparison to 'the previously reported record for Er-based qubits in solid-state platforms' (ref. 41); the abstract claim should be checked against the actual scope of the prior record.","section":"Abstract; Sec. III"},{"comment":"Typos: 'Chirstopher P. Lutz' in the Acknowledgment and 'A WG 5400' in Methods; the text would also benefit from stating that S_Ti,x in Supp. X is defined in the effective spin-1/2 basis.","section":"Acknowledgment; Methods"},{"comment":"The φ-dependence of the Rabi maps in Fig. 5(c) is simulated; the experimental Rabi data are all at φ = 52° (Fig. S13(c)) plus the vanishing points at φ = 0° and 90° (Fig. S16). The text should mark clearly which parts of the angular maps are experimental and which are predictions.","section":"Sec. V; Fig. 5(c)"}],"recommendation":"major_revision","confidential_remarks":"The experimental advance is solid and appropriate for a high-profile venue; the mechanism section is the main risk. I would ask the authors to (1) reframe the model validation honestly given that normalized Rabi data were used in the spin-Hamiltonian fit, ideally by refitting with those data withheld; (2) address the δα0 and equal-rotation assumptions quantitatively, or soften the 'general resource' claim to a leading interpretation; and (3) tone down 'near-gigahertz' for a measured 190 MHz. If these points are handled, the paper is publishable; I do not see a need for new experiments to reach that point."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe one thing to know: the experiment is real. They show Rabi oscillations for a single Er 4f spin exchange-coupled to Ti, driven electrically, with 190 MHz at the working point—ten times the prior rare-earth record. The central evidence is strong: coherent oscillations, linear scaling with RF voltage, and independence of the Er Rabi rate from tip conductance all point to a genuine electric drive rather than tip-field artifacts. That alone is a publishable result.\n\nWhat's new: prior work read out rare-earth spins through a sensor but did not coherently drive them. Here they drive the Er spin through the anisotropic Er-Ti exchange, and propose a general mechanism: RF modulation of the orbital frame wobbles the coupled moments around the external field, converting an electric field into an effective transverse field.\n\nWhere it gets softer: the mechanism is underdetermined. The spin-Hamiltonian parameters (g_Er, J_exc) are fitted to the same angular ESR data the model then reproduces, so part of the agreement is enforced by construction. More importantly, the wobbling amplitude δα0 = 0.1π is set by hand, and the equal-rotation assumption R_Ti = R_Er = R_B is asserted, not derived. The paper rules out diagonal exchange modulation, but it does not rule out two plausible alternatives: direct RF modulation of the Er g-tensor, or a Ti-only rotation of the exchange frame. Given the RF field is localized under the tip on Ti, a Ti-only wobble is arguably more natural than equal rotations. That does not kill the demonstration of coherent control, but it does mean the \"general resource\" claim is plausible rather than proven.\n\nMinor: the abstract's \"near-gigahertz\" overstates the measured 190 MHz, and no code or raw data are provided.\n\nIs it serious? Yes. The experimental result deserves referee time, and the mechanism can be sharpened with additional measurements—e.g., g-tensor drive estimates or asymmetric rotation fits. I would send it to review with the expectation that the mechanism section needs work. I would cite the 190 MHz demonstration, but not the mechanism as established.\n\nRecommendation: engage, but push for a revised mechanism section that either derives δα0 or tests alternative drive channels.","headline":"Solid experimental demonstration of all-electrical coherent control of a single surface Er spin at 190 MHz, with a plausible but underdetermined exchange-wobbling mechanism.","tokens_in":24515,"tokens_out":1604,"would_cite":true,"duration_ms":18379,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"By coupling a single erbium 4f spin to a titanium atom and driving the pair with a radio-frequency electric field, this paper demonstrates all-electrical coherent control of a shielded rare-earth spin at Rabi frequencies up to 190 MHz, an…","keywords":["rare-earth spin qubit","all-electrical spin control","anisotropic exchange","electron spin resonance STM","erbium","Rabi oscillations","surface spin qubit","scanned probe quantum control"],"falsifier":"Measure the Er Rabi frequency as a function of RF power and magnetic-field direction for an Er-Ti pair engineered to have a nearly isotropic exchange tensor (for example the Er(O)-Ti geometry), since the model predicts a Rabi rate about 30 times smaller than in Er(B)-Ti; additionally, if Rabi oscillations appear when the in-plane field is aligned along $\\phi = 0^\\circ$ or $90^\\circ$, where the rotational mechanism predicts zero drive, the wobbling picture would be refuted.","tokens_in":23436,"feed_emoji":"⚛️","tokens_out":6904,"duration_ms":71548,"temperature":0.7,"pith_summary":"This paper reports all-electrical coherent control of a single erbium 4f electron spin by coupling it to a titanium atom on a MgO film and driving the pair with a radio-frequency electric field. The authors claim that the RF field does not act directly on the shielded 4f electrons; instead it modulates the anisotropic exchange interaction between Er and Ti, effectively wobbling the coupled magnetic moments and producing a transverse field strong enough to drive the Er spin at up to 190 MHz. That is roughly an order of magnitude faster than previous rare-earth spin qubit control. If correct, the result establishes anisotropic exchange as a general, site-engineerable resource for electrically controlling otherwise well-shielded rare-earth spins, a step toward ultrafast local quantum gates and spin-photon interfaces.","feed_headline":"A single erbium spin is driven electrically at 190 MHz","feed_subtitle":"Anisotropic Er-Ti exchange turns shielding into a lever: tenfold faster than prior rare-earth spin control.","key_machinery":"The load-bearing object is the anisotropic exchange tensor $\\mathbf{J}$ and its RF-induced rotational modulation. The paper writes the driving term as $H_1 = \\delta\\alpha\\, \\mathbf{S}_{\\mathrm{Ti}}[\\mathbf{J}, G]\\mathbf{S}_{\\mathrm{Er}}$, where $\\delta\\alpha(t) = \\delta\\alpha_0 \\cos(\\omega_{\\mathrm{RF}} t)$ is an infinitesimal wobbling angle of both magnetic moments about the external field and $G$ is the generator of rotations about $\\mathbf{B}_{\\mathrm{ext}}$. Because the commutator vanishes for isotropic $\\mathbf{J}$, the drive exists only because the Er-Ti exchange is anisotropic, and the resulting transverse field is further amplified by the large out-of-plane Er g-factor $g_{zz}^{\\mathrm{Er}} = 9.59$. This rotational mode is chosen as the energetically favorable trajectory because it preserves the dominant Zeeman energy, and it is formalized as a Lie-derivative change of $\\mathbf{J}$ along the RF trajectory.","core_discovery":"The central claim is that a nearby Ti mediator spin lets an RF electric field drive a single Er 4f spin through rotational modulation of the Er-Ti exchange tensor. With three-dimensional magnetic-field control, the authors map the resonance and Rabi frequencies and fit a spin Hamiltonian with a strongly anisotropic Er g-tensor ($g^{\\mathrm{Er}}_{xx} = 1.05$, $g^{\\mathrm{Er}}_{yy} = 2.22$, $g^{\\mathrm{Er}}_{zz} = 9.59$) and exchange tensor ($J^{\\mathrm{exc}}_{xx} = 79.6$ MHz, $J^{\\mathrm{exc}}_{yy} = 543.7$ MHz, $J^{\\mathrm{exc}}_{zz} = 1272.5$ MHz). They show that diagonal modulations of the exchange coupling cannot reproduce the measured ESR angular maps, whereas a rotational wobbling trajectory, in which the RF field rotates the magnetic-moment frames about the external field while leaving the Zeeman energy unchanged, reproduces both the spectra and the angular dependence of the Rabi frequencies. The Er transition shows Rabi oscillations at 190 MHz, about ten times the previous rare-earth record, while the Ti transition and the double-quantum transition stay near 19 MHz. The authors conclude that anisotropic exchange is the general mechanism enabling electrical access to shielded 4f spins.","pith_inferences":["Inference: if the wobbling mechanism is correct, the orientation of the spin-pair bond becomes a design parameter, so rotating the Er-Ti axis relative to the field should tune the Rabi rate; this could be tested with the (3,0), (3.5,0.5), and (2.5,-0.5) pairs already studied.","Inference: the model fixes the wobbling amplitude to $\\delta\\alpha_0 = 0.1\\pi$ by hand, so a decisive independent check would be to extract this amplitude from the RF-power dependence of Rabi frequencies rather than from a global fit.","Inference: replacing Ti with a different transition-metal partner such as Ho or Sm could both improve coherence and create a tunable series of anisotropic-exchange mediators, extending the paper's central strategy beyond the single Er-Ti pair.","Inference: the same rotational-modulation picture implies that near-isotropic exchange pairs, like the previously studied Er(O)-Ti system, should show almost no electrical Rabi drive, which the paper predicts to be about 30 times weaker than in Er(B)-Ti."],"forward_implications":["The demonstrated 190 MHz Rabi rate is an order of magnitude above the previous rare-earth record and puts coherent Er rotations on a nanosecond timescale, well below the measured coherence time of about 113 ns.","The mechanism is general: any coupled spin pair with anisotropic exchange and contrasting g-tensor anisotropies should exhibit electrically driven coherent rotation, not just Er-Ti on MgO.","Because the drive is electric rather than magnetic, it can be applied locally and at high speed without requiring a resonant magnetic structure.","The angular maps identify field orientations where the Er transition is driven strongly and orientations where it vanishes, providing a practical operating recipe for future devices.","Coherent control of the double-quantum transition, demonstrated here for the first time for surface atoms, offers a route to two-spin rotations relevant for entanglement protocols."],"supporting_citations":[{"why":"Supplies the prior Er(O)-Ti pair that showed no coherent oscillations and provides the baseline and readout scheme this work extends.","marker":"[12]"},{"why":"Introduces the mediator-spin concept for electrically addressing a rare-earth spin, which the Er-Ti pair realizes.","marker":"[28]"},{"why":"Provides the known tip-induced electric-field driving behavior for surface spins that the authors use to separate tip effects from exchange-mediated driving.","marker":"[29]"},{"why":"Establishes the ESR-STM capability for coherent manipulation and readout of single spins with atomic precision.","marker":"[33]"},{"why":"Supplies the antiferromagnetic spin-spin coupling model and the four-level manifold used to assign the ESR transitions.","marker":"[40]"},{"why":"Reports the previous rare-earth spin qubit Rabi record that this paper's 190 MHz result is compared against.","marker":"[41]"},{"why":"Provides the known Ti g-tensor used as a fixed input in the fitting of the Er g-tensor and exchange tensor.","marker":"[46]"},{"why":"Supplies the multiplet-calculation methodology and Slater-integral rescaling used to model the Er(B) ground-state doublet and g-tensor.","marker":"[47]"}],"fun_headline_variants":["Er spin driven electrically at 190 MHz via Ti exchange","Tenfold faster rare-earth spin control via anisotropic exchange","Electrical drive of shielded Er spin hits 190 MHz"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the RF electric field acts mainly as a rigid rotation of the Er and Ti magnetic-moment frames about the external field, with the wobble amplitude set by hand to $\\delta\\alpha_0 = 0.1\\pi$ in all simulations, rather than by measured bond-length or orbital-deformation terms.","fun_headline_variants_meta":{"raw":{"variants":["Er spin driven electrically at 190 MHz via Ti exchange","Tenfold faster rare-earth spin control via anisotropic exchange","Electrical drive of shielded Er spin hits 190 MHz"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000916,"raw_usage":{"total_tokens":3958,"prompt_tokens":998,"completion_tokens":2960,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":614,"completion_tokens_details":{"reasoning_tokens":2909}},"tokens_in":614,"tokens_out":2960,"duration_ms":21824,"temperature":1.0,"reasoning_tokens":2909,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T00:39:09.212776+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the Er Rabi frequency as a function of RF power and magnetic-field direction for an Er-Ti pair engineered to have a nearly isotropic exchange tensor (for example the Er(O)-Ti geometry), since the model predicts a Rabi rate about 30 times smaller than in Er(B)-Ti; additionally, if Rabi oscillations appear when the in-plane field is aligned along $\\phi = 0^\\circ$ or $90^\\circ$, where the rotational mechanism predicts zero drive, the wobbling picture would be refuted.","supporting_citations":[{"cited_title":"Nano Letters , volume =","cited_arxiv_id":null,"evidence_quote":"Establishes the ESR-STM capability for coherent manipulation and readout of single spins with atomic precision."}],"review_version":1}