betaFeedback

WorksPhysics and Astronomy

Wide-band Unambiguous Quantum Sensing via Geodesic Evolution

Abstract

We present a quantum sensing technique that utilizes a sequence of ππ pulses to cyclically drive the qubit dynamics along a geodesic path of adiabatic evolution. This approach effectively suppresses the effects of both decoherence noise and control errors while simultaneously removing unwanted resonance terms, such as higher harmonics and spurious responses commonly encountered in dynamical decoupling control. As a result, our technique offers robust, wide-band, unambiguous, and high-resolution quantum sensing capabilities for signal detection and individual addressing of quantum systems, including spins. To demonstrate its versatility, we showcase successful applications of our method in both low-frequency and high-frequency sensing scenarios. The significance of this quantum sensing technique extends to the detection of complex signals and the control of intricate quantum environments. By enhancing detection accuracy and enabling precise manipulation of quantum systems, our method holds considerable promise for a variety of practical applications.

Introduction

Accurate characterization of the qubit environment holds significant importance across a range of applications, spanning from quantum information processing to quantum sensing [1, 2, 3, 4]. A widely utilized technique for achieving this is the implementation of dynamical decoupling (DD) pulse sequences [5, 6]. These sequences serve to filter out environmental noise, thereby extending the quantum coherence time, as well as to extract and amplify signals of specific frequencies [3, 4, 7]. Consequently, qubits under such sequences become highly sensitive quantum sensors, presenting diverse applications. For instance, nitrogen-vacancy (NV) centers [8, 9, 10] subjected to DD pulse sequences enable nanoscale nuclear magnetic resonance (NMR) [7, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26], spin label detection [27, 28, 29], spin cluster imaging [30, 31, 32, 33, 34, 35, 36, 37, 38], and AC field sensing [39, 40, 41, 42, 43, 44, 45, 46]. Furthermore, they can be employed for controlling nearby single nuclear spins [47, 48, 49, 50, 51, 52] in the context of quantum information processing [53, 54], quantum simulations [55], and quantum networks [56, 57, 58, 59].

However, DD pulse sequences used for quantum sensing, such as the commonly employed Carr-Purcell-Meiboom-Gill (CPMG) [60, 61] and XY8 [62] sequences encounter an issue known as spectral leakage and spurious resonance. These complications make the interpretation of the sensor’s recorded signal challenging and can result in ambiguous signal identification [63, 64]. DD pulses introduce abrupt temporal state flipping of the qubit sensor, causing a pronounced frequency modulation at a specific frequency [indicated by the gray line in Figure 1(d)]. This modulation leads to a strong resonance at the flipping frequency, enabling frequency-selective sensing. However, it also generates resonances at other frequencies, as evident in its Fourier transform [represented by the gray squares in Figure 1(d)]. Consequently, signals outside the target sensing frequency band can contribute to the sensor’s response and each individual frequency signal produces multiple signal peaks in the measured spectrum. Complex signals with a wide frequency range often exhibit significant signal overlap, posing challenges for analysis. Additionally, quantum heterodyne methods employed to down-convert high-frequency signals for sensing can exacerbate signal overlap [43, 44, 45]. Furthermore, the limited power of control pulses introduces spurious signals at unexpected frequencies [63]. These factors collectively present obstacles to reliably measuring environmental signals, particularly when various background noise sources are not fully characterized. With this goal in mind sequence timing has been optimised [65, 66] to eliminate higher-harmonics and sequence randomization has been explored to mitigates spurious signals [67, 68]. However, both approaches can address only specific their aspects and can solve these only partially.

Quantum sensing via geodesic jumping

Figure 1. Quantum sensing via geodesic jumping. (a) Upper panel: Repeated application of a sequence of NN π\pi pulses realizes cyclic quantum adiabatic evolution along the geodesic in (b) or (c). Lower panel: Combined with synchronized readout techniques for arbitrary frequency resolution. (b) GDyGD_y, where the closed geodesic is sampled by NN π\pi pulses. The red solid line illustrates the trajectory of the state evolution starting at ∣1⟩|1\rangle for N=12N=12. When ωscan\omega_{\rm scan} matches the frequency νn\nu_n of the target, a resonance occurs. (c) As (b) but for GDzGD_z which uses a horizontal geodesic, e.g., for robust heterodyne sensing of high-frequency signal. The resonant condition is accurately tuned by TscanT_{\rm scan} and the frequency ωctr\omega_{\rm ctr} of control field. (d) The resulting modulation function F(t)F(t) (red line for N=20N=20) and its Fourier components (red cycles) where fk=0f_k=0 for all 1<k<N−11<k<N-1. The gray line and squares are the corresponding ones for DD pulse sequences.

In this Letter, we present a novel approach called cyclic geodesic driving, which enables unambiguous sensing of signals across a broad frequency range on qubit sensors. Our method involves the application of a sequence of π\pi-pulses to implement accelerated high-fidelity quantum adiabatic driving, effectively inducing periodic evolution of the quantum sensor along the geodesic path on the Bloch sphere. As a result, the resonance frequency of the quantum sensor aligns with the frequency of the periodic evolution, while simultaneously mitigating the impact of environmental background noise. By employing this technique, each individual frequency signal generates a single, distinct signal response within the wide frequency band. This eliminates undesirable signal overlap, enabling precise characterization of complex external environments and signals. For achieving arbitrary high frequency resolution, our method can be combined with synchronized readout techniques [39, 40]. Furthermore, our approach exhibits remarkable resilience against control errors, as it is the counterpart of quantum adiabatic control. Overall, our proposed method of cyclic geodesic driving offers a robust and effective solution for wide-band, unambiguous signal sensing, enabling accurate analysis of intricate quantum systems and external environments.

Adiabatic shortcut by jumping

Our sensing scheme utilizes a sequence of periodic π\pi pulses [Figure 1(a)] which achieves cyclic quantum adiabatic evolution along a geodesic [69, 70, 71, 72] defined by the control Hamiltonian for the qubit sensor

Hc(t)=E(t)2(∣+ϕ⟩⟨+ϕ∣−∣−ϕ⟩⟨−ϕ∣),(1)H_c(t)=\frac{E(t)}{2}\left(|+\phi\rangle\langle+\phi|-|-\phi\rangle\langle-\phi|\right), \tag*{(1)}

where the instantaneous eigenstates ∣±ϕ⟩|\pm\phi\rangle are varied by a parameter ϕ=ϕ(t)\phi=\phi(t) starting with ϕ(0)=0\phi(0)=0 at the initial time t=0t=0. For adiabatic evolution of the parameter ϕ(t)\phi(t), the evolution operator for the quantum state driven by Hc(t)H_c(t) reads [69, 73, 74]

Uc(ϕ)=e−iϕ+(t)∣+ϕ⟩⟨+0∣+e−iϕ−(t)∣−ϕ⟩⟨−0∣,(2)U_c(\phi)=e^{-i\phi_+(t)}|+\phi\rangle\langle+0|+e^{-i\phi_-(t)}|-\phi\rangle\langle-0|, \tag*{(2)}

which transfers the initial state ∣±0⟩|\pm0\rangle to the instantaneous eigenstate ∣±ϕ⟩|\pm\phi\rangle at a later time. We use the Born-Fock gauge ⟨±ϕ∣ddϕ∣±ϕ⟩=0\langle\pm\phi|\frac{d}{d\phi}|\pm\phi\rangle=0 such that ϕ±(t)=±12∫0tE(t′)dt′\phi_\pm(t)=\pm\frac{1}{2}\int_0^t E(t')dt' are the dynamic phases. To realize the adiabatic evolution Uc(ϕ)U_c(\phi) by Hc(t)H_c(t) in finite time, we choose

∣+ϕ⟩=cos⁡(ϕ2)∣+0⟩+sin⁡(ϕ2)∣−0⟩,|+\phi\rangle=\cos\left(\frac{\phi}{2}\right)|+0\rangle+\sin\left(\frac{\phi}{2}\right)|-0\rangle,
∣−ϕ⟩=−sin⁡(ϕ2)∣+0⟩+cos⁡(ϕ2)∣−0⟩,(3)|-\phi\rangle=-\sin\left(\frac{\phi}{2}\right)|+0\rangle+\cos\left(\frac{\phi}{2}\right)|-0\rangle, \tag*{(3)}

which connect two orthonormal states ∣±0⟩|\pm0\rangle via a geodesic curve [75]. In our scheme we apply a sequence of control π\pi pulses via Eq. (1) with ϕ=ϕj\phi=\phi_j at the moments Tj=Tscan2j−12NT_j=T_{\rm scan}\frac{2j-1}{2N} (j=1,2,…j=1,2,\ldots), where NN is the pulse number in one periodic of evolution, see Figure 1(a). We use the linear relation ϕj=ωscanTj\phi_j=\omega_{\rm scan}T_j, where frequency ωscan\omega_{\rm scan} can be negative depending on the change of ϕj\phi_j in time. Each control π\pi pulse has a time duration tjt_j such that the pulse area ∫Tj−tj/2Tj+tj/2E(t)dt=π\int_{T_j-t_j/2}^{T_j+t_j/2}E(t)dt=\pi. Between the π\pi pulses, there is no control, i.e., E(t)=0E(t)=0 if ϕ∉{ϕj}\phi\notin\{\phi_j\}. We remove the dynamic phases at the final time of the evolution, we introduce a π\pi phase shift to the pulses in the second-half of the sequence. According to Refs. [69, 70, 71], the sequence realizes Uc(ϕ)U_c(\phi) with unit-fidelity at the middle of any successive path points ϕ=ϕˉj≡(ϕj+ϕj−1)/2\phi=\bar{\phi}_j\equiv(\phi_j+\phi_{j-1})/2. For other values of ϕ∉{ϕj}\phi\notin\{\phi_j\} the difference between Uc(ϕ)U_c(\phi) and the actual evolution implemented by Hc(t)H_c(t) is negligible when NN is sufficiently large. In contrast to conventional shortcuts to adiabaticity [74] to accelerate adiabatic process, in our method the instantaneous eigenstates of the control Hamiltonian Hc(t)H_c(t) are the same as the evolution path ∣±ϕ⟩|\pm\phi\rangle in Eq. (2). This avoids the use of counter-diabatic fields and retains the intrinsic robustness of traditional adiabatic process [71].

Unambiguous wide-bandwidth robust sensing

To demonstrate the concept of quantum sensing through the aforementioned geometric control, we examine a qubit coupled to its environment via the interaction Hamiltonian

Hint(t)=12σzB(t),(4)H_{\rm int}(t)=\frac{1}{2}\sigma_zB(t), \tag*{(4)}

where the Pauli operator σz=∣1⟩⟨1∣−∣0⟩⟨0∣\sigma_z=|1\rangle\langle1|-|0\rangle\langle0| and B(t)B(t) could be a classical field or a quantum operator in a rotating frame which includes possible dephasing noise. We use the control Hamiltonian Hc(t)=Δ(t)σz2+Ω(t)σx2H_c(t)=\Delta(t)\frac{\sigma_z}{2}+\Omega(t)\frac{\sigma_x}{2} and the states ∣+0⟩=∣1⟩|+0\rangle=|1\rangle and ∣−0⟩=∣0⟩|-0\rangle=|0\rangle for the geodesic in Eq. (3) for sensing. This geodesic driving around the yy-axis (GDyGD_y) is sketched in Figure 1(b). In the rotating frame of Hc(t)H_c(t), Hint(t)H_{\rm int}(t) becomes H~int(t)≈12Uc†σzUcB(t)\widetilde{H}_{\rm int}(t)\approx\frac{1}{2}U_c^\dagger\sigma_zU_cB(t). For the evolution Eq. (2), we find the approximated transformation [76]

Uc†σzUc≈cos⁡ϕσz;Uc†σxUc≈sin⁡ϕσz;Uc†σyUc≈0,(5)U_c^\dagger\sigma_z U_c \approx\cos\phi\sigma_z; \quad U_c^\dagger\sigma_x U_c \approx\sin\phi\sigma_z; \quad U_c^\dagger\sigma_y U_c \approx0, \tag*{(5)}

which have the nice property that they only depend on the geometric parameter ϕ(t)\phi(t). We obtain [76]

H~int(t)≈F(t)σz2B(t),(6)\widetilde{H}_{\mathrm{int}}(t) \approx F(t)\frac{\sigma_z}{2}B(t), \tag*{(6)}

where when NN is sufficiently large the modulation function F(t)≈cos⁡(ωscant)F(t) \approx\cos(\omega_{\mathrm{scan}}t) has only one Fourier component over a large frequency band, see Figure 1(d). Conventional

DD sequences [3, 4] also induce modulation factors F(t)F(t) to the σz\sigma_z operator with F(t)∈{±1}F(t) \in\{\pm1\} for ideal sequences [3, 4, 77]. However, those modulation factors have multiple Fourier components that complicate the interpretation of the sensor’s signal [63, 64, 65], see Figure 1(d) for equally spaced DD sequences [3, 4, 60, 61, 62, 78, 79, 80].

In Figure 2(a)–(Figure 2d) we simulate the measured spectrum of a classical AC field with B(t)=∑j=13bjcos⁡(νjt+θj)B(t) = \sum_{j=1}^{3} b_j \cos(\nu_jt+\theta_j), where νj\nu_j are the frequencies of different components. For the result of Figure 2(a) obtained by the widely-used robust XY8 sequence with an interpulse duration τ\tau [62], all the frequencies ({νj/2π}={500,1500.05,2499.88}\{\nu_j/2\pi\} = \{500, 1500.05, 2499.88\} kHz) cause transitions of the sensor states at 1/(2τ)≈5001/(2\tau) \approx500 kHz via the 1st, 3rd, and 5th harmonics. This problem of ambiguous spectral overlap is not solved even when we improve the frequency resolution sufficiently high via the synchronized readout technique (Qdyne) [39, 40], see Figure 2(c), because all the frequency components contribute to the readout signal in Qdyne. However, using GDy_y, only the frequency ν1/2π=500\nu_1/2\pi= 500 kHz contributes to the dip at the resonance ωscan=ν1\omega_{\mathrm{scan}} = \nu_1, because we have the effective Hamiltonian H~int≈14b1cos⁡θ1σz\widetilde{H}_{\mathrm{int}} \approx\frac{1}{4}b_1\cos\theta_1\sigma_z from Eq. (6) after rotating wave approximation [76]. The phase dependence on the effective Hamiltonian (and hence the signal) allows for arbitrary frequency resolution with synchronized readout, see Figure 2(d) and Figure 1(a). The results also show that our protocol is more resistant to control errors and dephasing noise. The already strong robustness of GDy_y is enhanced further with increasing number NN of pulses, see Figure 4.

Figure 2. Quantum spectroscopy. (a) Population signal of XY8 sequence (blue solid line) for spectroscopy of AC fields with the frequencies {νj/2π}={500,1500.05,2499.88}\{\nu_j/2\pi\} = \{500, 1500.05, 2499.88\} kHz by varying the pulse interval τ\tau. The 1500.05 kHz and 2499.88 kHz AC fields distort the signal centered at 500 kHz via the 3rd and 5th harmonics, respectively. Yellow dashed line is the result when there is dephasing noise (with a control-free decoherence time T2∗≈2 μT_2^* \approx2\ \mus) and control errors (with about 20% drift on the amplitude of control field). (b) As (a) but by using GDy_y. The 500 kHz signal dip is not distorted by the 1500.05 and 2499.88 kHz AC fields. (c) Power spectrum for the AC signal fields in (c) by using the synchronized readout technique in Ref. [39, 40]. The resonances due to higher harmonics make the signal unidentifiable even through the expected spectral resolution is about 1 Hz. (d) As (c) but by using GDy_y. (e) Signal of XY8 sequence for the detection of 1H{}^{1}\mathrm{H} spin with its frequency indicated by a red arrow. The spurious resonance (centered around the pink vertical line) produced by a 13C{}^{13}\mathrm{C} spin in diamond distorts the 1H{}^{1}\mathrm{H} spin signal. Yellow dashed line is the result when there are 2π×22\pi\times2 MHz detuning error and 30% of amplitude drift in the control field. (f) As (e) but by using GDy_y, where the spurious signals disappear. See [76] for details of simulation.

Control fidelity with respect to amplitude and detuning errors for (a) XY8 sequence, (b) GD_y with N = 20, (c) GD_y with N = 80, (d) CPMG sequence with 40 pulses for heterodyne sensing, (e) GD_z with N = 10, and (f) GD_z with N = 10 but each pulse is replaced by a Knill pulse

Figure 4. Control fidelity with respect to amplitude and detuning errors for (a) XY8 sequence, (b) GDy_y with N=20N = 20, (c) GDy_y with N=80N = 80, (d) CPMG sequence with 40 pulses for heterodyne sensing, (e) GDz_z with N=10N = 10, (f) GDz_z with N=10N = 10 but each pulse is replaced by a Knill pulse. All the protocols have the same sequence time length, and the ideal Rabi frequency of the control is 2π×502\pi\times50 MHz.

It is interesting that cyclic geodesic driving also fully solve the problems of spurious response due to finite-width pulses [63], as detailed in [76]. In Figure 2(e) and (Figure 2f), we simulate the quantum sensing of a single proton spin (1H{}^1\mathrm{H}) by an NV center. A 13C{}^{13}\mathrm{C} spin in diamond is also coupled to the NV center as a noise source. For this case B(t)B(t) is a quantum operator [48, 49, 76]. For the result Figure 2(e) obtained by XY8 sequences, the spurious response from the 13C{}^{13}\mathrm{C} spin disturb the target proton spin signal and lead to misidentification of 13C{}^{13}\mathrm{C} nuclei for proton. On the contrary, GDy_y provides a clean signal dip in the spectrum [Figure 2(f)], because when ωscan\omega_{\mathrm{scan}} matches the frequency ν1\nu_1 of the target 1H{}^1\mathrm{H}, H~int≈12a1σzI1z\widetilde{H}_{\mathrm{int}} \approx\frac{1}{2}a_1\sigma_z I_{1}^{z} (where I1,zI_{1,z} is the 1H{}^1\mathrm{H} spin operator and a1xa_1^x is the hyperfine strength [76]) and the effect of the 13C{}^{13}\mathrm{C} spin is removed.

Unambiguous heterodyne sensing. — The idea can be generalized to other settings. Consider the sensing of a multi-frequency signal field B⃗(t)=(Bx,By,Bz)\vec{B}(t) = (B_x, B_y, B_z) with frequencies νj\nu_j much larger than the Rabi frequency of the control field. We use GDz_z [see Figure 1(c)] with ∣±0⟩=(∣0⟩±∣1⟩)/2\lvert\pm0\rangle= (\lvert0\rangle\pm\lvert1\rangle)/\sqrt{2} for the geodesic in Eq. (3) to sense the traverse part B⊥(t)≡Bx+iBy=∑jbjeiαjcos⁡(νjt+θj)B_\perp(t) \equiv B_x + \mathrm{i}B_y = \sum_j b_j e^{\mathrm{i}\alpha_j}\cos(\nu_j t + \theta_j).

The relevant Hamiltonian reads

H=12(ωq+δt)σz+(12B⊥(t)σ++h.c.)+Hctr,H = \frac{1}{2}(\omega_q + \delta_t)\sigma_z + \left(\frac{1}{2}B_\perp(t)\sigma_+ + \mathrm{h.c.}\right) + H_{\mathrm{ctr}},

where ωq\omega_q is the frequency of the qubit and δt\delta_t is an unknown fluctuation. For NV qubits, ωq\omega_q (e.g., ωq=D+γeBz\omega_q = D + \gamma_e B_z with D=2π×2.87D = 2\pi\times2.87 GHz and γe≈2π×2.8\gamma_e \approx2\pi\times2.8 MHz/G) can be adjusted by changing the magnetic field BzB_z. The control Hamiltonian Hctr=Ω(t)cos⁡(ωctrt+ϕ)σxH_{\mathrm{ctr}} = \Omega(t)\cos(\omega_{\mathrm{ctr}}t+\phi)\sigma_x has a controllable detuning Δscan=ωctr−ωq\Delta_{\mathrm{scan}} = \omega_{\mathrm{ctr}} - \omega_q. In the rotating frame with respect to 12(ωq+δt)σz+Hctr\frac{1}{2}(\omega_q+\delta_t)\sigma_z + H_{\mathrm{ctr}}, we obtain the effective interaction [76]

H~int(t)≈14∑jbjei(αj−θj)+i(ωctr+ωscan−νj)tσ++h.c.(7)\widetilde{H}_{\mathrm{int}}(t) \approx\frac{1}{4}\sum_j b_j e^{\mathrm{i}(\alpha_j-\theta_j)+\mathrm{i}(\omega_{\mathrm{ctr}}+\omega_{\mathrm{scan}}-\nu_j)t}\sigma_+ + \mathrm{h.c.} \tag*{(7)}

Under the resonance condition for a signal frequency ωctr+ωscan=νn\omega_{\mathrm{ctr}}+\omega_{\mathrm{scan}} = \nu_n and when bj≪∣νn−νj∣b_j \ll\lvert\nu_n-\nu_j\rvert for j≠nj \ne n, H~int(t)≈14bnei(αn−θn)σ++h.c.\widetilde{H}_{\mathrm{int}}(t) \approx\frac{1}{4}b_n e^{\mathrm{i}(\alpha_n-\theta_n)}\sigma_+ + \mathrm{h.c.} picks up the signal only at the frequency νn\nu_n.

As exemplified in Figure 3(a), the heterodyne sensing using the CPMG sequences produce multiple dips at 1/(2τ)=±(νj−ωq)/k1/(2\tau) = \pm(\nu_j-\omega_q)/k (k=1,3,5,…k = 1, 3, 5, \ldots), which implies ambiguity sensor responses especially when {νj}\{\nu_j\} happen to be close to the qubit frequency ωq\omega_q. In contrast, our geodesic driving gives clear signal dips for accurate determination of all the signal frequencies, see Figure 3(b). Our method is also more resilient to dephasing noise (Figure 3) and is more robust against control errors (Figure 4). The robustness can be further enhanced by combining composite pulse techniques [see Figure 4(e) for the result where each pulse is replaced by a Knill pulse [66, 78, 81]].

Quantum heterodyne spectroscopy

Figure 3. (a) Quantum heterodyne spectroscopy of a signal field with frequencies {νj}=ωq+2π×{−84,−68,−56,−50,−42,−2,72,90}\{\nu_j\} = \omega_q + 2\pi\times\{-84, -68, -56, -50, -42, -2, 72, 90\} kHz by using the protocol in Refs. [43, 44] with the CPMG sequences. ωq\omega_q could be at the GHz range. The blue solid line (the line with yellow filling) is the simulation with (without) dephasing noise that induces a control-free decoherence time T2∗≈2 μT_2^* \approx2\ \mus. The pulse interval τ\tau is varied to measure the spectrum. Because for each AC signal frequency νj\nu_j resonance dips occur whenever 1/(2τ)=(νj−ωctr)/(2πk)1/(2\tau) = (\nu_j-\omega_{\mathrm{ctr}})/(2\pi k), (k=±1,±3,±5,…k = \pm1, \pm3, \pm5,\ldots), the dips at 1/(2τ)=(νj−ωq)/(2π)1/(2\tau) = (\nu_j-\omega_q)/(2\pi) indicated by red vertical arrows are obscured by other resonance dips (vertical lines). Different number N′N' of π\pi pulses are used for different range of τ\tau to insure that the sequence times are smaller than 1 ms. (b) [(c)]: As (a) but by using Protocol with fixed Δscan=0\Delta_{\mathrm{scan}} = 0 (ωscan=2π×80\omega_{\mathrm{scan}} = 2\pi\times80 kHz). All the dips only appear at the right frequencies νj\nu_j. See [76] for details of simulation.

Conclusion. — We propose a quantum sensing scheme based on a quantum adiabatic shortcut along a geodesic path. This scheme offers the capability to resolve complex broadband signals and addresses the issues of spectral overlap and spurious signals that arise in existing DD-based quantum sensing methods. Notably, our approach allows for arbitrary frequency resolution through the utilization of synchronized readout techniques. Moreover, it exhibits robustness against control errors and effectively suppresses unwanted decoherence noise. The versatility of our method extends beyond signal detection; it can also be employed for the detection and control of various quantum objects, including single nuclear spins, spin clusters, and mechanical oscillators. Furthermore, our scheme holds promise for applications aimed at mitigating crosstalk in qubit arrays. In summary, our proposed quantum sensing scheme based on a quantum adiabatic shortcut along a geodesic path provides a universal solution for accurate signal detection, offering improved performance over existing methods. Its potential applications encompass a wide range of quantum systems and address key challenges in the field.

Acknowledgments

This work was supported by National Natural Science Foundation of China (Grant No. 12074131), the Natural Science Foundation of Guangdong Province (Grant No. 2021A1515012030), the ERC Synergy grant HyperQ (grant no. 856432) and the BMBF Zukunftscluster QSense: Quantensenoren für die biomedizinische Diagnostik (QMED) (grant no 03ZU1110FF).

References

Spectral responses

Higher harmonic and spurious responses in DD sequences.

Here we provide a background on DD based quantum sensing. Consider the qubit-environment Hamiltonian

Hint(t)=12σzB(t).(8)H_{\mathrm{int}}(t)=\frac{1}{2}\sigma_z B(t). \tag*{(8)}

A sequence of DD π\pi pulses is applied to the qubit, where every π\pi pulse aims to flip the qubit and to induce a transformation σz→−σz\sigma_z \to-\sigma_z in the interaction picture with respective to the DD control. Let UDD(t)U_{\mathrm{DD}}(t) be the evolution operator driven by the DD control. In the rotating frame of the DD control, Hint(t)H_{\mathrm{int}(t)} becomes

H~int(t)=12UDD†(t)σzUDD(t)B(t).(9)\widetilde{H}_{\mathrm{int}}(t)=\frac{1}{2}U_{\mathrm{DD}}^{\dagger}(t)\sigma_z U_{\mathrm{DD}}(t)B(t). \tag*{(9)}

For the idealized situation where the DD sequences use infinitely sharp δ\delta-shape π\pi pulses (where the pulse duration tpt_p equals 0), H~int(t)=12F(t)σzB(t)\widetilde{H}_{\mathrm{int}}(t)=\frac{1}{2}F(t)\sigma_z B(t) with F(t)∈{+1,−1}F(t)\in\{+1,-1\}. In realistic experiments the π\pi pulses have a nonzero duration tpt_p because the control has a finite Rabi frequency Ω\Omega. In general,

H~int(t)=12∑α=x,y,zFα(t)σαB(t).(10)\widetilde{H}_{\mathrm{int}}(t)=\frac{1}{2}\sum_{\alpha=x,y,z}F_{\alpha}(t)\sigma_{\alpha}B(t). \tag*{(10)}

When the π\pi pulses have a constant high Ω\Omega, tp=π/Ωt_p=\pi/\Omega. For the control Hamiltonian Hc(t)=Ω2σxH_c(t)=\frac{\Omega}{2}\sigma_x, we have during the first π\pi pulse H~int(t)=12[σzcos⁡(Ωt)+σysin⁡(Ωt)]B(t)\widetilde{H}_{\mathrm{int}}(t)=\frac{1}{2}[\sigma_z\cos(\Omega t)+\sigma_y\sin(\Omega t)]B(t). A plot of the Fα(t)F_{\alpha}(t) for a XY8 sequence is shown in Figure S1, and the corresponding Fourier transforms of Fα(t)F_{\alpha}(t) are shown by the blue lines in Figure S2. Under the ideal situation that there is no control error and that the pulse duration tp=0t_p=0, the modulation factor

Figure S1. XY8 sequence and its induced modulation functions Fα(t)F_\alpha(t). The π\pi pulses of the XY8 sequence have a constant high Ω=2π×10\Omega= 2\pi\times10 MHz and a time interval, τ\tau, between pulses. The value of τ\tau determines the frequency 1/(2τ)=11/(2\tau) = 1 MHz for sensing.

Figure S2. The power spectrum of Fα(t)F_\alpha(t) (α=x,y,z\alpha= x, y, z) for the XY8 sequence in Figure S1 (blue lines) and for the geodesic evolution with N=16N = 16 in Figure S3 (orange line). fα(ω)=1Tseq∫0TseqFα(t)eiωt dtf_\alpha(\omega) = \frac{1}{T_{\mathrm{seq}}}\int_{0}^{T_{\mathrm{seq}}} F_\alpha(t)e^{i\omega t}\,\mathrm{d}t. We choose the same total sequence time Tseq=16 μsT_{\mathrm{seq}} = 16\,\mu\mathrm{s} for both methods. Both methods have the strongest spectral response for fz(ω)f_z(\omega) at ω/(2π)=1/(2τ)=ωscan/(2π)=1 MHz\omega/(2\pi) = 1/(2\tau) = \omega_{\mathrm{scan}}/(2\pi) = 1\,\mathrm{MHz}. For the XY8 sequences, there are also resonances at k/(2τ)k/(2\tau) (k=3,5,7,…k = 3, 5, 7, \ldots) due to higher harmonics. The nonzero pulse width (π/Ω)(\pi/\Omega) also results in complicated spurious peaks appeared in fx(ω)f_x(\omega) and fy(ω)f_y(\omega). For the method of geodesic evolution, only one resonance response exists in a wide range of spectrum. For fz(ω)f_z(\omega), we have unambiguous frequency response at ω=ωscan\omega= \omega_{\mathrm{scan}}, and only tiny higher harmonics appear at ∣ω∣≥(N−1)∣ωscan∣\lvert\omega\rvert\ge(N - 1)\lvert\omega_{\mathrm{scan}}\rvert, c.f. Figure 1 in the main text. For fx(ω)f_x(\omega) and fy(ω)f_y(\omega), weak resonances appear at (N±2)ωscan(N \pm2)\omega_{\mathrm{scan}}, (N±1)ωscan(N \pm1)\omega_{\mathrm{scan}}, and NωscanN\omega_{\mathrm{scan}}.

F(t)=Fz(t)=∑k=1∞4sin⁡(kπ2)kπcos⁡(kπτt),(11)F(t)=F_z(t)=\sum_{k=1}^{\infty}\frac{4\sin\left(k\frac{\pi}{2}\right)}{k\pi}\cos\left(k\frac{\pi}{\tau}t\right), \tag*{(11)}

has harmonics of the frequency π/τ\pi/\tau for an interpulse spacing τ\tau of equally spaced DD sequences. For tp>0t_p>0, Fx(t)F_x(t) and Fy(t)F_y(t) can have non-zero values during the π\pi pulses and their Fourier transforms have many peaks, which cause spurious signals in quantum sensing.

Quantum sensing by geometric evolution

Here we provide details on the operator dynamics in the rotating frame of the control Hc(t)=E(t)2(∣+ϕ⟩⟨+ϕ∣−∣−ϕ⟩⟨−ϕ∣)H_c(t)=\frac{E(t)}{2}(\lvert+\phi\rangle\langle+\phi\rvert-\lvert-\phi\rangle\langle-\phi\rvert) [Eq. (1)] in the main text with ϕ=ϕ(t)\phi=\phi(t) and

∣+ϕ⟩=cos⁡(ϕ2)∣+0⟩+sin⁡(ϕ2)∣−0⟩,(12)\lvert+\phi\rangle=\cos\left(\frac{\phi}{2}\right)\lvert+0\rangle+\sin\left(\frac{\phi}{2}\right)\lvert-0\rangle, \tag*{(12)}
∣−ϕ⟩=−sin⁡(ϕ2)∣+0⟩+cos⁡(ϕ2)∣−0⟩.(13)\lvert-\phi\rangle=-\sin\left(\frac{\phi}{2}\right)\lvert+0\rangle+\cos\left(\frac{\phi}{2}\right)\lvert-0\rangle. \tag*{(13)}

Without lost of generality, we assume that σz=∣1⟩⟨1∣−∣0⟩⟨0∣\sigma_z=\lvert1\rangle\langle1\rvert-\lvert0\rangle\langle0\rvert, σx=∣1⟩⟨0∣+∣0⟩⟨1∣\sigma_x=\lvert1\rangle\langle0\rvert+\lvert0\rangle\langle1\rvert, and σy=−i∣1⟩⟨0∣+i∣0⟩⟨1∣\sigma_y=-i\lvert1\rangle\langle0\rvert+i\lvert0\rangle\langle1\rvert with ∣+0⟩=∣1⟩\lvert+0\rangle=\lvert1\rangle and ∣−0⟩=∣0⟩\lvert-0\rangle=\lvert0\rangle. In the rotating frame of the evolution Uc(t)=Te−i∫0tHc(t′)dt′U_c(t)=\mathcal{T}e^{-i\int_0^t H_c(t')\mathrm{d}t'}, we write

Uc†σzUc=∑α=x,y,zFα(t)σα.(14)U_c^{\dagger}\sigma_z U_c=\sum_{\alpha=x,y,z}F_{\alpha}(t)\sigma_{\alpha}. \tag*{(14)}

A plot of the Fα(t)F_{\alpha}(t) for GDy_y is shown in Figure S3. The corresponding Fourier transforms are shown by the yellow lines in Figure S2. One can see that the geodesic evolution removes higher harmonics and spurious responses over a wide frequency band. For the Fourier transform of Fz(t)F_z(t), we have unambiguous frequency response at ω=ωscan\omega=\omega_{\mathrm{scan}} given that

Figure S3. The modulation functions Fα(t)F_\alpha(t) induced by the geodesic evolution using GDy_y. The plots show a complete period of Fα(t)F_\alpha(t). The control field has a Rabi frequency Ω=2π×10\Omega= 2\pi\times10 MHz. The control parameters N=20N = 20 and ωscan/(2π)=1\omega_{\mathrm{scan}}/(2\pi) = 1 MHz.

ϕ(Tj)=ωscanTj\phi(T_j)=\omega_{\mathrm{scan}}T_j. Tiny higher harmonics only appear at the frequency ω>(N−1)ωscan\omega> (N-1)\omega_{\mathrm{scan}}. For the Fourier transforms of Fx(t)F_x(t) and Fy(t)F_y(t), weak resonances only appear at ω>(N−2)ωscan\omega> (N-2)\omega_{\mathrm{scan}}. Therefore, one can arbitary remove unwanted spectral respoances via a sufficently large NN.

When the control Hc(t)H_c(t) realizes the ideal adiabatic evolution Uc(ϕ)=e−iϕ+(t)∣+ϕ⟩⟨+0∣+e−iϕ−(t)∣−ϕ⟩⟨−0∣U_c(\phi)=e^{-i\phi_+(t)}|+_\phi\rangle\langle+_0|+e^{-i\phi_-(t)}|-_\phi\rangle\langle-_0| [S6–S8], we have

Uc†σzUc=cos⁡ϕσz−sin⁡ϕ[eiφ(t)∣1⟩⟨0∣+h.c.],Uc†σxUc=sin⁡ϕσz+cos⁡ϕ[eiφ(t)∣1⟩⟨0∣+h.c.],Uc†σyUc=ieiφ(t)∣1⟩⟨0∣+h.c..(15)\begin{aligned} U_c^\dagger\sigma_zU_c &= \cos\phi\sigma_z-\sin\phi[e^{i\varphi(t)}|1\rangle\langle0|+\mathrm{h.c.}],\\ U_c^\dagger\sigma_xU_c &= \sin\phi\sigma_z+\cos\phi[e^{i\varphi(t)}|1\rangle\langle0|+\mathrm{h.c.}],\\ U_c^\dagger\sigma_yU_c &= ie^{i\varphi(t)}|1\rangle\langle0|+\mathrm{h.c.}. \tag*{(15)} \end{aligned}

where the difference of dynamic phases φ(t)=φ+(t)−φ−(t)=∫0tE(s)ds\varphi(t)=\varphi_+(t)-\varphi_-(t)=\int_0^t E(s)\mathrm{d}s increases with time and eiφ(t)e^{i\varphi(t)} is a fast oscillating factor with a zero mean. Neglecting terms involving eiφ(t)e^{i\varphi(t)} in a way similar to the rotating wave approximation, we obtain Eq. (5) in the main text, i.e.,

Uc†σzUc≈cos⁡ϕσz,Uc†σxUc≈sin⁡ϕσz,Uc†σyUc≈0,(16)\begin{aligned} U_c^\dagger\sigma_zU_c &\approx\cos\phi\sigma_z,\\ U_c^\dagger\sigma_xU_c &\approx\sin\phi\sigma_z,\\ U_c^\dagger\sigma_yU_c &\approx0, \tag*{(16)} \end{aligned}

and

Uc†(σz+iσx)Uc=σzeiϕ.(17)U_c^\dagger(\sigma_z+i\sigma_x)U_c=\sigma_z e^{i\phi}. \tag*{(17)}

The operator dynamics can also be analyzed as follows. The control Hc(t)H_c(t) with ϕ=ϕj\phi=\phi_j is used to realize the jjth π\pi pulse which has the propagator Uj=exp⁡[−iπ2(cos⁡ϕjσz+sin⁡ϕjσx)]U_j=\exp[-i\frac{\pi}{2}(\cos\phi_j\sigma_z+\sin\phi_j\sigma_x)]. After nn such π\pi pulses, we have

Uc=UnUn−1⋯U2U1.(18)U_c=U_nU_{n-1}\cdots U_2U_1. \tag*{(18)}

This gives the transformation after nn π\pi pulses, Uc†σzUc=σzcos⁡Θn+σxsin⁡ΘnU_c^\dagger\sigma_zU_c=\sigma_z\cos\Theta_n+\sigma_x\sin\Theta_n, where Θn=2∑j=1n(−1)j+1ϕj\Theta_n=2\sum_{j=1}^{n}(-1)^{j+1}\phi_j. Using the relation ϕj=ωscanTscan2j−12N\phi_j=\omega_{\mathrm{scan}}T_{\mathrm{scan}}\frac{2j-1}{2N}, we obtain Θn=(−1)n+1nNωscanTscan\Theta_n=(-1)^{n+1}\frac{n}{N}\omega_{\mathrm{scan}}T_{\mathrm{scan}}, where the factor (−1)n(-1)^n is caused by the π\pi phase shifts of the dynamic phases. Neglecting the fast oscillating term σxsin⁡Θn\sigma_x\sin\Theta_n in a way similar to the rotating wave approximation, we obtain after nn π\pi pulses Uc†σzUc≈σzcos⁡ϕnU_c^\dagger\sigma_zU_c\approx\sigma_z\cos\phi_n. Since ϕj=ϕ(Tj)=ωscanTj\phi_j=\phi(T_j)=\omega_{\mathrm{scan}}T_j, we have Uc†σzUc≈σzcos⁡ϕ(t)U_c^\dagger\sigma_zU_c\approx\sigma_z\cos\phi(t).

Sensing of single nuclear spins by GDy_y

Here we provide details of quantum sensing by geometric evolution for the case of sensing of single nuclear spins by a single NV center. The Hamiltonian of a NV electron spin and its surrounding nuclear spins reads

H=−∑jγjI⃗j⋅(Bzz^)+∣1⟩⟨1∣∑jA⃗j⋅I⃗j+Hc(t),(19)H=-\sum_j\gamma_j\vec{I}_j\cdot(B_z\hat{z})+|1\rangle\langle1|\sum_j\vec{A}_j\cdot\vec{I}_j+H_c(t), \tag*{(19)}

where I⃗j\vec{I}_j is the spin-12\frac{1}{2} operator for the jjth nuclear spin, γj\gamma_j is the gyromagnetic ratio, and A⃗j=(Ajx,Ajy,Ajz)\vec{A}_j=(A_j^x,A_j^y,A_j^z) describes the hyperfine interaction at the spin location. Bzz^B_z\hat{z} is a magnetic field with an amplitude BzB_z and a direction denoted by the unit vector z^\hat{z}. Hc(t)=Δ(t)σz2+Ω(t)σx2H_c(t)=\Delta(t)\frac{\sigma_z}{2}+\Omega(t)\frac{\sigma_x}{2} as in the main text for GDy_y. We has neglected the week nuclear-nuclear coupling in the Hamiltonian because it is much smaller than the electron-nuclear coupling.

The nuclear spins have the effective precession frequency ν⃗j=νjν^j=γjBzz^−12A⃗j\vec{\nu}_j=\nu_j\hat{\nu}_j=\gamma_jB_z\hat{z}-\frac{1}{2}\vec{A}_j, where ν^j\hat{\nu}_j is a unit vector. In the rotating frame with respect to −∑jν⃗j⋅I⃗j-\sum_j\vec{\nu}_j\cdot\vec{I}_j, the Hamiltonian becomes

H~(t)=Hint(t)+Hc(t),(20)\widetilde{H}(t)=H_{\mathrm{int}}(t)+H_c(t), \tag*{(20)}
Hint(t)=σz2B(t),(21)H_{\mathrm{int}}(t)=\frac{\sigma_z}{2}B(t), \tag*{(21)}

where

B(t)=∑j12ajx(Ij+e−iνjt+h.c.)+12ajzIjz,(22)B(t)=\sum_j\frac{1}{2}a_j^x\left(I_j^+e^{-i\nu_jt}+\mathrm{h.c.}\right)+\frac{1}{2}a_j^zI_j^z, \tag*{(22)}

ajz=A⃗j⋅ν^ja_j^z=\vec{A}_j\cdot\hat{\nu}_j, and ajx=∣A⃗j−ajzν^j∣a_j^x=\left|\vec{A}_j-a_j^z\hat{\nu}_j\right|. Defining z^j≡ν^j\hat{z}_j\equiv\hat{\nu}_j, x^j≡(A⃗j−ajzν^j)/ajx\hat{x}_j\equiv(\vec{A}_j-a_j^z\hat{\nu}_j)/a_j^x, and y^j≡z^j×x^j\hat{y}_j\equiv\hat{z}_j\times\hat{x}_j, we write Ijμ=I⃗j⋅μ^jI_j^\mu=\vec{I}_j\cdot\hat{\mu}_j with μ∈{x,y,z}\mu\in\{x,y,z\} and Ij±=Ijx±iIjyI_j^\pm=I_j^x\pm iI_j^y. In the rotating frame with respect to Hc(t)H_c(t), the Pauli operator σz\sigma_z becomes

Uc†σzUc≡∑α=x,y,zFα(t)σα,(23)U_c^\dagger\sigma_zU_c\equiv\sum_{\alpha=x,y,z}F_\alpha(t)\sigma_\alpha, \tag*{(23)}

where Fα(t)F_\alpha(t) are oscillating factors. See Figure S3 and Figure S1 for exemplified plots for GDy_y and XY8 sequences, respectively. For the case of GDy_y,

Uc†σzUc≈F(t)σz(24)U_c^\dagger\sigma_zU_c\approx F(t)\sigma_z \tag*{(24)}

in the relevant frequency range. With a sufficiently large NN,

F(t)=Fz(t)≈cos⁡(ωscant).(25)F(t)=F_z(t)\approx\cos(\omega_{\mathrm{scan}}t). \tag*{(25)}

Therefore, when ωscan=νn\omega_{\mathrm{scan}}=\nu_n for the nnth spin and assuming that ωscan≫ajx\omega_{\mathrm{scan}}\gg a_j^x and ajxa_j^x are much smaller than the differences between νn\nu_n and other νj\nu_j, we obtain

H~int≈12anxσzInx,(26)\widetilde{H}_{\mathrm{int}}\approx\frac{1}{2}a_n^x\sigma_zI_n^x, \tag*{(26)}

where the qubit sensor couples to only one nucler spin.

Heterodyne sensing by GDz_z

We consider a single-qubit sensor exposed to a high-frequency signal field B⃗(t)=(Bx,By,Bz)\vec{B}(t)=(B_x,B_y,B_z) and a dephasing noise δt\delta_t, with the Hamiltonian

H0=(ωq+δt+Bz)σz2+Bxσx2+Byσy2,(27)H_0=(\omega_q+\delta_t+B_z)\frac{\sigma_z}{2}+B_x\frac{\sigma_x}{2}+B_y\frac{\sigma_y}{2}, \tag*{(27)}
=(ωq+δt+Bz)σz2+(12B⊥σ++h.c.),(28)=(\omega_q+\delta_t+B_z)\frac{\sigma_z}{2}+\left(\frac{1}{2}B_\perp\sigma_++\mathrm{h.c.}\right), \tag*{(28)}

where B⊥≡Bx−iBy=∑jbjeiαjcos⁡(νjt+θj)B_\perp\equiv B_x-iB_y=\sum_jb_je^{i\alpha_j}\cos(\nu_jt+\theta_j) and σ+=12(σx+iσy)\sigma_+=\frac{1}{2}(\sigma_x+i\sigma_y). Adding a control Hctr=Ω(t)cos⁡(ωctrt+ϕ)σxH_{\mathrm{ctr}}=\Omega(t)\cos(\omega_{\mathrm{ctr}}t+\phi)\sigma_x with a detuning Δscan=ωctr−ωq\Delta_{\mathrm{scan}}=\omega_{\mathrm{ctr}}-\omega_q, the total Hamiltonian is H=H0+HctrH=H_0+H_{\mathrm{ctr}}. In the rotating frame with respect to 12ωctrσz\frac{1}{2}\omega_{\mathrm{ctr}}\sigma_z, the Hamiltonian becomes

H′=(δt+Bz−Δscan)σz2+[12B⊥(t)σ+eiωctrt+h.c.]+Ω(t)σϕ2,(29)H'=(\delta_t+B_z-\Delta_{\mathrm{scan}})\frac{\sigma_z}{2} +\left[\frac{1}{2}B_\perp(t)\sigma_+e^{i\omega_{\mathrm{ctr}}t}+\mathrm{h.c.}\right]+\Omega(t)\frac{\sigma_\phi}{2}, \tag*{(29)}

where

σϕ≡σxcos⁡ϕ+σycos⁡ϕ.(30)\sigma_{\phi} \equiv\sigma_x \cos\phi+ \sigma_y \cos\phi. \tag*{(30)}

The control part Hc=Ω(t)σϕ2H_c = \Omega(t)\frac{\sigma_{\phi}}{2} realizes a sequence of π\pi pulses for accelerated quantum adiabatic evolution along the geodesic and prolongs the coherence time of the qubit by suppressing the nosie in a way similar to DD [70]. On the other hand, BzB_z is a fast oscillating field at a frequency much larger than the Rabi frequency Ω(t)\Omega(t). Therefore, the effect of the term (δt+Bz−Δscan)σz2(\delta_t+B_z-\Delta_{\mathrm{scan}})\frac{\sigma_z}{2} can be neglected. Similar to Eq. (17) in § 1.5 B, in the rotating frame of Uc(t)=Te−i∫0tHc(t′)dt′U_c(t)=Te^{-i\int_0^t H_c(t')dt'}, Uc†σxUc≈cos⁡ϕσxU_c^\dagger\sigma_xU_c\approx\cos\phi\sigma_x and Uc†σyUc≈sin⁡ϕσxU_c^\dagger\sigma_yU_c\approx\sin\phi\sigma_x, which implies

Uc†σ+Uc≈12σxeiϕ≈12σxeiωscant.(31)U_c^\dagger\sigma_+U_c\approx\frac{1}{2}\sigma_xe^{i\phi}\approx\frac{1}{2}\sigma_xe^{i\omega_{\mathrm{scan}}t}. \tag*{(31)}

Therefore, in the rotating frame of Uc(t)U_c(t), we obtain the effective Hamiltonian

H~int(t)≈Uc†[12B⊥(t)σ+eiωctrt+h.c.]Uc≈14B⊥(t)σxei(ωctr+ωscan)t+h.c.,≈18∑jbjei(αj−θj)+i(ωctr+ωscan−νj)tσx+h.c.,=∑jbj4cos⁡[(ωctr+ωscan−νj)t+(αj−θj)]σx.(32)\begin{aligned} \widetilde{H}_{\mathrm{int}}(t)&\approx U_c^\dagger\left[\frac{1}{2}B_{\perp}(t)\sigma_+e^{i\omega_{\mathrm{ctr}}t}+\mathrm{h.c.}\right]U_c\\ &\approx\frac{1}{4}B_{\perp}(t)\sigma_xe^{i(\omega_{\mathrm{ctr}}+\omega_{\mathrm{scan}})t}+\mathrm{h.c.},\\ &\approx\frac{1}{8}\sum_j b_je^{i(\alpha_j-\theta_j)+i(\omega_{\mathrm{ctr}}+\omega_{\mathrm{scan}}-\nu_j)t}\sigma_x+\mathrm{h.c.},\\ &=\sum_j\frac{b_j}{4}\cos\left[(\omega_{\mathrm{ctr}}+\omega_{\mathrm{scan}}-\nu_j)t+(\alpha_j-\theta_j)\right]\sigma_x. \tag*{(32)} \end{aligned}

When the frequency of control matches a signal frequency νn\nu_n, i.e.,

ωctr+ωscan=νn,(33)\omega_{\mathrm{ctr}}+\omega_{\mathrm{scan}}=\nu_n, \tag*{(33)}

and the signal is weak ∣bj/(νj−νn)∣≪1\lvert b_j/(\nu_j-\nu_n)\rvert\ll1, the qubit sensor extracts the signal at this frequency with the Hamiltonian

H~int(t)≈12bncos⁡(αn−θn)σx2,(34)\widetilde{H}_{\mathrm{int}}(t)\approx\frac{1}{2}b_n\cos(\alpha_n-\theta_n)\frac{\sigma_x}{2}, \tag*{(34)}

by applying the rotating wave approximation.

For a state ρ\rho after the control of a duration TT, the population remaining at the initial state is P∣ψ0⟩⟨ψ0∣≡Tr⁡(ρ(T)∣ψ0⟩⟨ψ0∣)P_{\lvert\psi_0\rangle\langle\psi_0\rvert}\equiv\operatorname{Tr}(\rho(T)\lvert\psi_0\rangle\langle\psi_0\rvert). For the resonant case ωctr+ωscan=νn\omega_{\mathrm{ctr}}+\omega_{\mathrm{scan}}=\nu_n, P∣ψ0⟩⟨ψ0∣=cos⁡2[14bncos⁡(θn−αn)T]P_{\lvert\psi_0\rangle\langle\psi_0\rvert}=\cos^2\left[\frac{1}{4}b_n\cos(\theta_n-\alpha_n)T\right], while for the offresonant case, P∣ψ0⟩⟨ψ0∣≈1P_{\lvert\psi_0\rangle\langle\psi_0\rvert}\approx1.

Simulation details

AC field sensing

The AC field used for Figure 2 (a)-(Figure 2b) of the main text is B(t)=∑jbjcos⁡(νjt+θj)B(t)=\sum_j b_j\cos(\nu_jt+\theta_j), where {νj}=2π×{500,1500.05,2499.875} kHz\{\nu_j\}=2\pi\times\{500,1500.05,2499.875\}\ \mathrm{kHz}, {bj}=2π×{5,10,10} kHz\{b_j\}=2\pi\times\{5,10,10\}\ \mathrm{kHz}, and θj=0\theta_j=0.

Without control errors and dephasing noise, the Hamiltonian for the simulation reads H=B(t)σz2+Hc(t)H=B(t)\frac{\sigma_z}{2}+H_c(t). For the XY8 sequences, Hc(t)=12Ω(t)σμH_c(t)=\frac{1}{2}\Omega(t)\sigma_\mu when applying μ\mu (μ=x,y\mu=x,y) pulses (see Figure S1 for the pulse sequence). For GDy_y, Hc(t)=Δ(t)σz2+Ω(t)σx2H_c(t)=\Delta(t)\frac{\sigma_z}{2}+\Omega(t)\frac{\sigma_x}{2}. In the simulations of both protocols, we choose Ω(t)=Ω=2π×10 MHz\Omega(t)=\Omega=2\pi\times10\ \mathrm{MHz} for the Rabi frequency during the π\pi pulses when there is no control errors. In GDy_y, we use the detuning of the control Δ=Ωcot⁡ϕj\Delta=\Omega\cot\phi_j for the jjth pulse according to the ideal value of Ω\Omega. For the sensing using XY8 sequences, we use the resonant frequency 1/(2τ)1/(2\tau) at the 1st harmonic to detect the signal, and we repeat the XY8 unit 16 times implying the whole XY8 sequence has N′=128N'=128 π\pi pulses and a total duration Tseq=128τT_{\mathrm{seq}}=128\tau. To compare the results for the same sequence duration TseqT_{\mathrm{seq}}, we apply M=64M=64 cycles of the cyclic evolution when we use the GDy_y (with N=25N=25). In both protocols the qubit is initially prepared in the superposition state ∣+⟩=(∣0⟩+∣1⟩)/2\lvert+\rangle=(\lvert0\rangle+\lvert1\rangle)/\sqrt{2} and we measure the population remaining at the initial state ∣+⟩\lvert+\rangle as the signal.

For the results with dephasing noise and control errors, we add a noise term δB(t)\delta B(t) to B(t)B(t) and an amplitude fluctuation to the Rabi frequency Ω(t)\Omega(t) in both protocols. δB(t)\delta B(t) is modeled by a stochastic Ornstein-Uhlenbeck (OU) process [S5] with a correlation time τB=4 ms\tau_B=4\ \mathrm{ms} and a standard deviation δB=2π×0.1 MHz\delta_B=2\pi\times0.1\ \mathrm{MHz} [43], which gives a decoherence time T2∗≈2 μsT_2^* \approx2\ \mu\mathrm{s}, i.e., a time when the phase coherence decays to 1/e1/e of its original value in the absence of quantum control. In the modeling of the control fluctuation, we change the value of Ω(t)\Omega(t) to [1+η(t)](1+ϵ)Ω(t)[1+\eta(t)](1+\epsilon)\Omega(t), where the static bias ϵ=20%\epsilon=20\% and the time-dependent fluctuation η(t)\eta(t) is modeled by a stochastic OU process with a correlation time of 1 ms and a standard deviation 2.4×10−32.4 \times10^{-3} [S1]. The results are the average of 6500 Monte Carlo simulations. The initial state is ∣ψ0⟩=(∣0⟩+∣1⟩)/2|\psi_0\rangle=(|0\rangle+|1\rangle)/\sqrt{2}, and we measure the population remaining at the initial state as the signal.

Arbitrary high-resolution spectroscopy

Combining the sensing sequences with the method of synchronized readout [S3, S4] can improve the spectral resolution. In the simulations for Figure 2 (c) and (Figure 2d) of the main text, we use the same AC field [B(t)=∑jbjcos⁡(νjt+θj)[B(t)=\sum_j b_j\cos(\nu_j t+\theta_j) with {νj}=2π×{500,1500.05,2499.875} kHz\{\nu_j\}=2\pi\times\{500,1500.05,2499.875\}\ \mathrm{kHz}, {bj}=2π×{5,10,10} kHz\{b_j\}=2\pi\times\{5,10,10\}\ \mathrm{kHz}, θj=0]\theta_j=0] and sequence parameters (Ω=2π×10 MHz(\Omega=2\pi\times10\ \mathrm{MHz}, N′=128N'=128, M=64M=64, and N=25)N=25) as in Sec. II A.

We repeatedly apply the same sequence (each has the duration Tseq=128τT_{\mathrm{seq}}=128\tau) for 10410^4 times to obtain the synchronized readout signals [see the lower panel of Figure 1(a) in the main text]. At the end of each sequence we make a projection measurement and record the measured values. There is a delay of trt_r for readout and initialization between the end of a sequence and the starting time of its subsequent sequence. The readout process at the end of the nnth sequence yields at the moment tt a random photon number yn(t)=Pois⁡[C(1−rBn[pn])]y_n(t)=\operatorname{Pois}[C(1-rB_n[p_n])], where Pois denotes a Poisson process and BnB_n is a Bernoulli process that takes the value 1 with a probability of pnp_n and the value 0 with probability 1−pn1-p_n [S3]. We initialize the qubit sensor at the state ∣ψ0⟩=(∣0⟩+∣1⟩)/2|\psi_0\rangle=(|0\rangle+|1\rangle)/\sqrt{2} before each sequence. The probability pnp_n is the population signal of finding the qubit in the state ∣+y⟩=(∣0⟩+i∣1⟩)/2|+y\rangle=(|0\rangle+i|1\rangle)/\sqrt{2} at the end of the sequence. We choose r=0.35r=0.35, tr=2.32 μst_r=2.32\ \mu\mathrm{s}, and C≈0.105nC\approx0.105n with n=10n=10 according to the parameters in Ref. [S3]. A discrete Fourier transform y~\tilde{y} of the photon numbers {yn(t)}\{y_n(t)\} results in the power spectrum ∣y~∣/N|\tilde{y}|/\mathcal{N} shown in Figure 2 (c) and (Figure 2d). Here N\mathcal{N} is a fixed scaling factor for all the protocols.

Sensing of single nuclear spins

In Figure 2 (e) and (Figure 2f) of the main text for the sensing of a proton spin, we perform the simulations by using the Hamiltonian Eq. (19). The control Hc(t)=Δ(t)σz/2+Ω(t)σx/2H_c(t)=\Delta(t)\sigma_z/2+\Omega(t)\sigma_x/2 as in the main text for GDy_y. For the XY8 sequences, Hc(t)=12Ω(t)σμH_c(t)=\frac{1}{2}\Omega(t)\sigma_\mu (μ∈{x,y}\mu\in\{x,y\}). The parameters for the proton spin are γ1H/2π≈4.2576 kHz/G\gamma_{\mathrm{^{1}H}}/2\pi\approx4.2576\ \mathrm{kHz/G}, (AxH)2+(AyH)2/2π=0.5 kHz\sqrt{(A_x^{\mathrm{H}})^2+(A_y^{\mathrm{H}})^2}/2\pi=0.5\ \mathrm{kHz}, and AzH/2π=0.5 kHzA_z^{\mathrm{H}}/2\pi=0.5\ \mathrm{kHz}. We assume that one 13C{}^{13}\mathrm{C} nucleus (with a natural abundance 1.1% and γC≈1.0705 kHz/G\gamma_{\mathrm{C}}\approx1.0705\ \mathrm{kHz/G}) in the diamond is coupled to the NV with (AxC)2+(AyC)2/2π=100 kHz\sqrt{(A_x^{\mathrm{C}})^2+(A_y^{\mathrm{C}})^2}/2\pi=100\ \mathrm{kHz} and AzC/2π=28 kHzA_z^{\mathrm{C}}/2\pi=28\ \mathrm{kHz}. The magnetic field Bz=100 GB_z=100\ \mathrm{G}. The Rabi frequency for the NV is Ω=2π×10 MHz\Omega=2\pi\times10\ \mathrm{MHz}. The XY8 sequence is repeated 50 times for a total number of N′=400 πN'=400\ \pi pulses. We change B(t)→B(t)+ΔerrorB(t)\rightarrow B(t)+\Delta_{\mathrm{error}} to include detuning error. We set Δerror=2π×2 MHz\Delta_{\mathrm{error}}=2\pi\times2\ \mathrm{MHz} as it is the typical error due to unpolarized nitrogen spin of the NV center. An amplitude error ϵ\epsilon will change Ω(t)\Omega(t) to (1+ϵ)Ω(t)(1+\epsilon)\Omega(t), where we set ϵ=30%\epsilon=30\%. The initial state of the nuclear spins are assumed to be in a room-temperature thermal state. We measure the population of the NV qubit sensor remaining at its initial state ∣ψ0⟩=(∣0⟩+∣1⟩)/2|\psi_0\rangle=(|0\rangle+|1\rangle)/\sqrt{2} as the signal.

High-frequency AC field sensing

In Figure 3 of the main text, the signal field B⃗(t)=(Bx,By,Bz)\vec{B}(t)=(B_x,B_y,B_z) has the transverse component B⊥≡Bx−iBy=∑jbjeiαjcos⁡(νjt+θj)B_\perp\equiv B_x-iB_y=\sum_j b_j e^{i\alpha_j}\cos(\nu_j t+\theta_j). In the simulation, {(νj−ωq)/2π}={−84,−68,−56,−50,−42,−2,72,90} kHz\{(\nu_j-\omega_q)/2\pi\}=\{-84,-68,-56,-50,-42,-2,72,90\}\ \mathrm{kHz}, with the corresponding {bj/2π}={5.3,4.6,2.7,4.2,3.4,4.5,2.2,3.1} kHz\{b_j/2\pi\}=\{5.3,4.6,2.7,4.2,3.4,4.5,2.2,3.1\}\ \mathrm{kHz}. The angles αj\alpha_j and phases θj\theta_j are random real numbers. The control field has a Rabi frequency Ω=2π×50 MHz\Omega=2\pi\times50\ \mathrm{MHz}. For the results with dephasing noise δt\delta_t, in each run of the simulation we model δt\delta_t by a OU process with a correlation time τB=4 ms\tau_B=4\ \mathrm{ms} and a standard deviation δB=2π×0.1 MHz\delta_B=2\pi\times0.1\ \mathrm{MHz} for a decoherence time T2∗≈2 μsT_2^*\approx2\ \mu\mathrm{s}. The results are the average of 10410^4 Monte Carlo simulations.

For the heterodyne spectroscopy using CPMG sequences, the initial state is the superposition state ∣ψ0⟩=(∣0⟩+∣1⟩)/2|\psi_0\rangle=(|0\rangle+|1\rangle)/\sqrt{2} and the population remained at ∣ψ0⟩|\psi_0\rangle is the measured signal.

For the heterodyne spectroscopy using GDz_z, the initial state is ∣ψ0⟩=∣0⟩|\psi_0\rangle=|0\rangle. For the case of NV center, ∣0⟩|0\rangle is the state after optical illumination and we do not need to use the initial π/2\pi/2 pulse to prepare the qubit state. We measure the population remained at ∣ψ0⟩|\psi_0\rangle as the signal. There are 16 cycles of geodesic evolution where for each cycle there are N=10 πN=10\ \pi pulses.

Robustness and Combination with composite pulses

In Figure 4 of the main text we demonstrate the control robustness in the absence of signals. A Rabi frequency Ω=2π×50\Omega= 2\pi\times50 MHz is used for the simulations. For DD sequences (CPMG and XY8 sequences) we choose N′=40N' = 40 pulses for each sequence and the pulse interval τ\tau satisfies 1/(2τ)=11/(2\tau) = 1 MHz. Therefore each DD sequence has a time length Tseq=2τ×N′/2=40T_{\mathrm{seq}} = 2\tau\times N'/2 = 40 μ\mus. For the protocols of geodesic driving, we use ωscan/(2π)=1\omega_{\mathrm{scan}}/(2\pi) = 1 MHz for the same sensing frequency as DD, Δscan=0\Delta_{\mathrm{scan}} = 0, and the same Tseq=40T_{\mathrm{seq}} = 40 μ\mus. To investigate the robustness of the protocols, we change the Rabi frequency to (1+ϵ)Ω(1+\epsilon)\Omega by adding a relative fluctuation, and add a detuning error Δ\Delta (i.e., adding Δσz\Delta\sigma_z to the qubit Hamiltonian). We denote U(Δ/Ω,ϵ)U(\Delta/\Omega,\epsilon) the actual evolution realized by the sequences in the presence of control errors. In the absence of control errors, the ideal evolution after the sequence is an identity operator, Uideal=IU_{\mathrm{ideal}} = I. The control fidelity is calculated in the standard way as [S9]

fidelity=Tr⁡[U(ΔΩ,ϵ)Uideal†]/Tr⁡I.(35)\mathrm{fidelity} = \operatorname{Tr}\left[U\left(\frac{\Delta}{\Omega},\epsilon\right)U_{\mathrm{ideal}}^\dagger\right]/\operatorname{Tr}I. \tag*{(35)}

References

  1. [1]J. Preskill, Quantum Computing in the NISQ era and beyond, Quantum 2, 79 (2018).
  2. [2]E. Paladino, Y. M. Galperin, G. Falci, and B. L. Altshuler, 1/f noise: Implications for solid-state quantum information, Rev. Mod. Phys. 86, 361 (2014).DOI
  3. [3]C. L. Degen, F. Reinhard, and P. Cappellaro, Quantum sensing, Rev. Mod. Phys. 89, 035002 (2017).
  4. [4]J. F. Barry, J. M. Schloss, E. Bauch, M. J. Turner, C. A. Hart, L. M. Pham, and R. L. Walsworth, Sensitivity optimization for NV-diamond magnetometry, Rev. Mod. Phys. 92, 015004 (2020).DOI
  5. [5]L. Viola, E. Knill, and S. Lloyd, Dynamical decoupling of open quantum systems, Phys. Rev. Lett. 82, 2417 (1999).
  6. [6]W. Yang, Z.-Y. Wang, and R.-B. Liu, Preserving qubit coherence by dynamical decoupling, Front. Phys. 6, 2 (2011).
  7. [7]C. Munuera-Javaloy, R. Puebla, and J. Casanova, Dynamical decoupling methods in nanoscale NMR, Europhys. Lett. 134, 30001 (2021).
  8. [8]M. W. Doherty, N. B. Manson, P. Delaney, F. Jelezko, J. Wrachtrup, and L. C. Hollenberg, The Nitrogen-Vacancy Colour Centre in Diamond, Phys. Rep. 528, 1 (2013).
  9. [9]Y. Wu, F. Jelezko, M.B. Plenio, and T. Weil, Diamond Quantum Devices in Biology, Angewandte Chemie – International Edition 55, 6586 – 6598 (2016).
  10. [10]J. R. Weber, W. F. Koehl, J. B. Varley, A. Janotti, B. B. Buckley, C. G. Van de Walle, D. D. Awschalom, Quantum computing with defects. Proc. Natl. Acad. Sci. U.S.A. 107, 8513–8518 (2010).
  11. [11]S. Kolkowitz, Q. P. Unterreithmeier, S. D. Bennett, and M. D. Lukin, Sensing Distant Nuclear Spins with a Single Electron Spin, Phys. Rev. Lett. 109, 137601 (2012).
  12. [12]N. Zhao, J. Honert, B. Schmid, M. Klas, J. Isoya, M. Markham, D. Twitchen, F. Jelezko, R. B. Liu, H. Fedder, and J. Wrachtrup, Sensing single remote nuclear spins, Nat. Nanotechnol. 7, 657 (2012).DOI
  13. [13]T. Staudacher, F. Shi, S. Pezzagna, J. Meijer, J. Du, C. A. Meriles, F. Reinhard, and J. Wrachtrup, Nuclear Magnetic Resonance Spectroscopy on a <5-Nanometer3 Sample Volume, Science 339, 561 (2013).
  14. [14]C. Müller, X. Kong, J.-M. Cai, K. Melentijević, A. Stacey, M. Markham, D. Twitchen, J. Isoya, S. Pezzagna, J. Meijer, J. Du, M. B. Plenio, B. Naydenov, L. P. McGuinness, and F. Jelezko, Nuclear magnetic resonance spectroscopy with single spin sensitivity, Nat. Commun. 5, 4703 (2014).DOI
  15. [15]D. Rugar, H. J. Mamin, M. H. Sherwood, M. Kim, C. T. Rettner, K. Ohno, and D. D. Awschalom, Proton magnetic resonance imaging using a nitrogen-vacancy spin sensor, Nat. Nanotechnol. 10, 120 (2015).DOI
  16. [16](2015).DOI
  17. [17]D. R. Glenn, D. B. Bucher, J. Lee, M. D. Lukin, H. Park, R. L. Walsworth, High-resolution magnetic resonance spectroscopy using a solid-state spin sensor. Nature 555, 351–354 (2018).
  18. [18]E. Lang, J. Casanova, Z.-Y. Wang, M. B. Plenio, and T. S. Monteiro, Enhanced Resolution in Nanoscale NMR via Quantum Sensing with Pulses of Finite Duration, Phys. Rev. Applied 7, 054009 (2017).DOI
  19. [19]M. Pfender, P. Wang, H. Sumiya, S. Onoda, W. Yang, D. B. Rao Dasari, P. Neumann, X.-Y. Pan, J. Isoya, R.-B. Liu, and J. Wachtrup, High-resolution spectroscopy of single nuclear spins via sequential weak measurements, Nat. Commun. 10, 594 (2019).DOI
  20. [20]D. B. Bucher, D. P. L. Aude Craik, M. P. Backlund, M. J. Turner, O. Ben Dor, D. R. Glenn, R. L. Walsworth, Quantum diamond spectrometer for nanoscale NMR and ESR spectroscopy. Nat. Protoc. 14, 2707–2747 (2019).DOI
  21. [21]J. Casanova, E. Torrontegui, M. B. Plenio, J. J. García-Ripoll, and E. Solano, Modulated Continuous Wave Control for Energy-Efficient Electron-Nuclear Spin Coupling, Phys. Rev. Lett. 122, 010407 (2019).
  22. [22]N. Aharon, I. Schwartz, and A. Retzker, Quantum Control and Sensing of Nuclear Spins by Electron Spins under Power Limitations, Phys. Rev. Lett. 122, 120403 (2019).DOI
  23. [23]C. Munuera-Javaloy, Y. Ban, X. Chen, and J. Casanova, Robust Detection of High-Frequency Signals at the Nanoscale, Phys. Rev. Appl. 14, 054054 (2020).DOI
  24. [24]J. Meinel, V. Vorobyov, P. Wang, B. Yavkin, M. Pfender, H. Sumiya, S. Onoda, J. Isoya, R.-B. Liu, and J. Wachtrup, Quantum nonlinear spectroscopy of single nuclear spins, Nat. Commun. 13, 5318 (2022).
  25. [25]P. Wang, W. Yang, and R. Liu, Using Weak Measurements to Synthesize Projective Measurement of Nonconserved Observables of Weakly Coupled Nuclear Spins, Phys. Rev. Applied 19, 054037 (2023).DOI
  26. [26]C. Munuera-Javaloy, A. Tobalina, and J. Casanova, High-Resolution NMR Spectroscopy at Large Fields with Nitrogen Vacancy Centers, Phys. Rev. Lett. 130, 133603 (2023).
  27. [27]F. Shi, Q. Zhang, P. Wang, H. Sun, J. Wang, X. Rong, M. Chen, C. Ju, F. Reinhard, H. Chen, J. Wachtrup, J. Wang, and J. Du, Single-protein spin resonance spectroscopy under ambient conditions, Science 347, 1135 (2015).DOI
  28. [28]I. Lovchinsky, A. O. Sushkov, E. Urbach, N. P. de Leon, S. Choi, K. De Greve, R. Evans, R. Gertner, E. Bersin, C. Müller, L. McGuinness, F. Jelezko, R. L. Walsworth, H. Park, M. D. Lukin, Nuclear magnetic resonance detection and spectroscopy of single proteins using quantum logic. Science 351, 836 (2016).
  29. [29]C. Munuera-Javaloy, R. Puebla, B. D. Anjou, M. B. Plenio, and J. Casanova, Detection of molecular transitions with nitrogen-vacancy centers and electron-spin la- bels, npj Quantum Inf. 8, 140 (2022).DOI
  30. [30]N. Zhao, J.-L. Hu, S.-W. Ho, J. T. K. Wan, and R. B. Liu, Atomic-scale magnetometry of distant nuclear spin clusters via nitrogen-vacancy spin in diamond, Nat. Nanotechnol. 6, 242 (2011).DOI
  31. [31]F. Shi, X. Kong, P. Wang, F. Kong, N. Zhao, R. B. Liu, and J. Du, Sensing and Atomic-Scale Structure Analysis of Single Nuclear-Spin Clusters in Diamond, Nat. Phys. 10, 21 (2014).
  32. [32]J. E. Lang, R. B. Liu, and T. S. Monteiro, Dynamical Decoupling-Based Quantum Sensing: Floquet Spectroscopy, Phys. Rev. X 5, 041016 (2015).DOI
  33. [33]Z.-Y. Wang, J. F. Haase, J. Casanova, and M. B. Plenio, Positioning nuclear spins in interacting clusters for quantum technologies and bioimaging, Phys. Rev. B 93, 174104 (2016).DOI
  34. [34]K. Sasaki, K. M. Itoh, and E. Abe, Determination of the position of a single nuclear spin from free nuclear precessions detected by a solid-state quantum sensor, Phys. Rev. B 98, 121405(R) (2018).
  35. [35]J. Zopes, K. S. Cujia, K. Sasaki, J. M. Boss, K. M. Itoh, and C. L. Degen, Three-dimensional localization spectroscopy of individual nuclear spins with sub-Angstrom resolution, Nat. Commun. 9, 4678 (2018).DOI
  36. [36]J. Zopes, K. Herb, K. S. Cujia, and C. L. Degen, Three Dimensional Nuclear Spin Positioning Using Coherent Radio-Frequency Control, Phys. Rev. Lett. 121, 170801 (2018).DOI
  37. [37]M. H. Abobeih, J. Randall, C. E. Bradley, H. P. Bartling, M. A. Bakker, M. J. Degen, M. Markham, D. J. Twitchen, and T. H. Tamminiau, Atomic-scale imaging of a 27-nuclear-spin cluster using a quantum sensor, Nature 576, 411 (2019).
  38. [38]K. S. Cujia, K. Herb, J. Zopes, J. M. Abendroth, and C. L. Degen, Parallel detection and spatial mapping of large nuclear spin clusters, Nat. Commun. 13 1260 (2022).DOI
  39. [39]S. Schmitt, T. Gefen, F. M. Stürner, T. Unden, G. Wolff, C. Müller, J. Scheuer, B. Naydenov, M. Markham, S. Pezzagna, J. Meijer, I. Schwarz, M. Plenio, A. Retzker, L. P. McGuinness, and F. Jelezko, Submillihertz magnetic spectroscopy performed with a nanoscale quantum sensor, Science 356, 832 (2017).
  40. [40]J. M. Boss, K. S. Cujia, J. Zopes, and C. L. Degen, Quantum sensing with arbitrary frequency resolution, Science 356, 837 (2017).DOI
  41. [41]T. Joas, A. M. Waeber, G. Braunbeck, and F. Reinhard, Quantum sensing of weak radio-frequency signals by pulsed Mollow absorption spectroscopy, Nat. Commun. 8, 964 (2017).DOI
  42. [42]A. Stark, N. Aharon, T. Unden, D. Louzon, A. Huck, A. Retzker, U. L. Andersen, and F. Jelezko, Narrow-bandwidth sensing of high-frequency fields with continuous dynamical decoupling, Nat. Commun. 8, 1105 (2017).DOI
  43. [43]Y. Chu, P. Yang, M. Gong, M. Yu, B. Yu, M. B. Plenio, A. Retzker, and J. Cai, Precise Spectroscopy of High-Frequency Oscillating Fields with a Single-Qubit Sensor, Phys. Rev. Applied 15, 014031 (2021).DOI
  44. [44]J. Meinel, V. Vorobyov, B. Yavkin, D. Dasari, H. Sumiya, S. Onoda, J. Isoya, and J. Wrachtrup, Heterodyne sensing of microwaves with a quantum sensor, Nat. Commun. 12, 1 (2021).
  45. [45]G. Wang, Y.-X. Liu, J. M. Schloss, S. T. Alsid, D. A. Braje, and P. Cappellaro, Sensing of Arbitrary-Frequency Fields Using a Quantum Mixer, Phys. Rev. X 12, 021061 (2022).DOI
  46. [46]Z. Jiang, H. Cai, R. Cernansky, X. Liu, and W. Gao, Quantum sensing of radio-frequency signal with NV centers in SiC, Sci. Adv. 9, eadg2080 (2023).DOI
  47. [47]T. H. Taminiau, J. T. W. Wagenaar, T. van der Sar, F. Jelezko, V. V. Dobrovitski, and R. Hanson, Detection and Control of Individual Nuclear Spins using a Weakly Coupled Electron Spin, Phys. Rev. Lett. 109, 137602 (2012).
  48. [48]Z.-Y. Wang, J. Casanova, and M. B. Plenio, Delayed entanglement echo for individual control of a large number of nuclear spins, Nat. Commun. 8, 14660 (2017).DOI
  49. [49]J. F. Haase, Z.-Y. Wang, J. Casanova, and M. B. Plenio, Soft quantum control for highly selective interactions among joint quantum systems, Phys. Rev. Lett. 121, 050402 (2018).
  50. [50]M. A. Perlin, Z.-Y. Wang, J. Casanova, and M. B. Plenio, Noise-resilient architecture of a hybrid electron-nuclear quantum register in diamond, Quantum Sci. Technol. 4, 015007 (2019).DOI
  51. [51]S. S. Hegde, J. Zhang, and D. Suter, Efficient quantum states for individual nuclear spin qubits by indirect control, Phys. Rev. Lett. 124, 220501 (2020).DOI
  52. [52]H. P. Bartling, M. H. Abobeih, B. Pingault, M.J. Degen, S. J. H. Loenen, C. E. Bradley, J. Randall, M. Markham, D. J. Twitchen, and T. H. Taminiau, Entanglement of Spin-Pair Qubits with Intrinsic Dephasing Times Exceeding a Minute, Phys. Rev. X 12, 011048 (2022).DOI
  53. [53]J. Casanova, Z.-Y. Wang, and M. B. Plenio, Noise-Resilient Quantum Computing with a Nitrogen-Vacancy Center and Nuclear Spins, Phys. Rev. Lett. 117, 130502 (2016).
  54. [54]S. Pezzagna and J. Meijer, Quantum computer based on color centers in diamond, Appl. Phys. Rev. 8, 011308 (2021).
  55. [55]J. Cai, A. Retzker, F. Jelezko, and M. B. Plenio, A largescale quantum simulator on a diamond surface at room temperature, Nat. Phys. 9, 168 (2013).
  56. [56]L. Childress and R. Hanson, Diamond NV centers for quantum computing and quantum networks, MRS Bull. 38, 134 (2013).
  57. [57]N. Kalb, A. A. Reiserer, P. C. Humphreys, J. J. W. Bakermans, S. J. Kamerling, N. H. Nickerson, S. C. Benjamin, D. J. Twitchen, M. Markham, and R. Hanson, Entanglement Distillation between Solid-State Quantum Network Nodes, Science 356, 928 (2017).DOI
  58. [58]P. C. Humphreys, N. Kalb, J. P. J. Morits, R. N. Schouten, R. F. L. Vermeulen, D. J. Twitchen, M. Markham, and R. Hanson, Deterministic delivery of remote entanglement on a quantum network, Nature (London) 558, 268 (2018).
  59. [59]M. Pompili, S. L. N. Hermans, S. Baier, H. K. C. Beukers, P. C. Humphreys, R. N. Schouten, R. F. L. Vermeulen, M. J. T. Tiggelen, L. dos Santos Martins, B. Dirksе, S. Wehner, and R. Hanson, Realization of a Multinode Quantum Network of Remote Solid-State Qubits, Science 372, 259 (2021).DOI
  60. [60]H. Y. Carr and E. M. Purcell, Effects of diffusion on free precession in nuclear magnetic resonance experiments, Phys. Rev. 94, 630 (1954).
  61. [61]S. Meiboom and D. Gill, Modified spin-echo method for measuring nuclear relaxation times, Rev. Sci. Instrum. 29, 688 (1958).
  62. [62]T. Gullion, D. B. Baker, and M. S. Conradi, New, compensated Carr-Purcell sequences, J. Magn. Reson. 89, 479 (1990).
  63. [63]M. Loretz, J. M. Boss, T. Rosskopf, H. J. Mamin, D. Rugar, and C. L. Degen, Spurious Harmonic Response of Multipulse Quantum Sensing Sequences, Phys. Rev. X 5, 021009 (2015).DOI
  64. [64]V. M. Frey, S. Mavadia, L. M. Norris, W. de Ferranti, D. Lucarelli, L. Viola, and M. J. Biercuk, Application of optimal band-limited control protocols to quantum noise sensing, Nat. Commun. 8, 2189 (2017).
  65. [65]N. Zhao, J. Wrachtrup, and R.-B. Liu, Dynamical decoupling design for identifying weakly coupled nuclear spins in a bath, Phys. Rev. A 90, 032319 (2014).
  66. [66]J. Casanova, Z.-Y. Wang, J. F. Haase, and M. B. Plenio, Robust dynamical decoupling sequences for individual nuclear-spin addressing, Phys. Rev. A 92, 042304 (2015).
  67. [67]Z.-Y. Wang, J. E. Lang, S. Schmitt, J. Lang, J. Casanova, L. McGuinness, T. S. Monteiro, F. Jelezko, and M. B. Plenio, Randomization of Pulse Phases for Unambiguous and Robust Quantum Sensing, Phys. Rev. Lett. 122, 200403 (2019).
  68. [68]Z.-Y. Wang, J. Casanova, and M. B. Plenio, Enhancing the Robustness of Dynamical Decoupling Sequences with Correlated Random Phases, Symmetry 12, 730 (2020).
  69. [69]Z.-Y. Wang and M. B. Plenio, Necessary and sufficient condition for quantum adiabatic evolution by unitary control fields, Phys. Rev. A 93, 052107 (2016).
  70. [70]K. Xu, T. Xie, F. Shi, Z.-Y. Wang, X. Xu, P. Wang, Y. Wang, M. B. Plenio, and J. Du, Breaking the quantum adiabatic speed limit by jumping along geodesics, Sci. Adv. 5, eaax3800 (2019).
  71. [71]Y. Liu and Z.-Y. Wang, Shortcuts to adiabaticity with inherent robustness and without auxiliary control, arXiv:2211.02543.arxiv.org/abs/2211.02543
  72. [72]M. G. Gong, M. Yu, R. Betzholz, Y. Chu, P. Yang, Z. Wang, and J. Cai, Accelerated quantum control in a three-level system by jumping along the geodesics, Phys. Rev. A 107, L040602 (2023).
  73. [73]M. V. Berry, Transitionless quantum driving, J. Phys. A: Math. Theor. 42, 365303 (2009).
  74. [74]D. Géúry-Odelin, A. Ruschhaupt, A. Kiely, E. Torrentégui, S. Marínez-Garaot, and J. G. Muga, Shortcuts to adiabaticity: Concepts, methods, and applications, Rev. Mod. Phys. 91, 045001 (2019).
  75. [75]D. Chruscinski, A. Jamiolkowski, Geometric Phases in Classical and Quantum Mechanics (Birkhäuser, 2004).
  76. [76]See Supplemental Material at [URL] for additional information.
  77. [77]L. Cywiński, R. M. Lutchyn, C. P. Nave, and S. Das Sarma, How to enhance dephasing time in superconducting qubits, Phys. Rev. B 77, 174509 (2008).
  78. [78]C. A. Ryan, J. S. Hodges, and D. G. Cory, Robust Decoupling Techniques to Extend Quantum Coherence in Diamond, Phys. Rev. Lett. 105, 204002 (2010).
  79. [79]G. T. Genov, D. Schrafft, N. V. Vitanov, and T. Halfmann, Arbitrarily Accurate Pulse Sequences for Robust Dynamical Decoupling, Phys. Rev. Lett. 118, 133202 (2017).
  80. [80]G. A. Alvarez and D. Suter, Measuring the spectrum of colored noise by dynamical decoupling, Phys. Rev. Lett. 107, 230501 (2011).
  81. [81]R. Tycko and A. Pines, Iterative schemes for broad-band and narrow-band population inversion in NMR, Chem. Phys. Lett. 111, 462 (1984).

Paper details

vlnc:W_tw6ftccf
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

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

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents