This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

thanks: These authors contributed equally to this work.thanks: These authors contributed equally to this work.thanks: These authors contributed equally to this work.

Experimental Realization of Universal Time-optimal non-Abelian Geometric Gates

Zhikun Han National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Yuqian Dong National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Baojie Liu Institute for Quantum Science and Engineering, and Department of Physics, Southern University of Science and Technology, Shenzhen 518055, China    Xiaopei Yang National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Shuqing Song National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Luqing Qiu National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Danyu Li National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Ji Chu National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Wen Zheng National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Jianwen Xu National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Tianqi Huang National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Zhimin Wang National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Xiangmin Yu National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Xinsheng Tan [email protected] National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Dong Lan National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China    Man-Hong Yung [email protected] Institute for Quantum Science and Engineering, and Department of Physics, Southern University of Science and Technology, Shenzhen 518055, China    Yang Yu National Laboratory of Solid State Microstructures, School of Physics, Nanjing University, Nanjing 210093, China
Abstract

Based on the geometrical nature of quantum phases, non-adiabatic holonomic quantum control (NHQC) has become a standard technique for enhancing robustness in constructing quantum gates. However, the conventional approach of NHQC is sensitive to control instability, as it requires the driving pulses to cover a fixed pulse area. Furthermore, even for small-angle rotations, all operations need to be completed with the same duration of time. Here we experimentally demonstrate a time-optimal and unconventional approach of NHQC (called TOUNHQC), which can optimize the operation time of any holonomic gate. Compared with the conventional approach, TOUNHQC provides an extra layer of robustness to decoherence and control errors. The experiment involves a scalable architecture of superconducting circuit, where we achieved a fidelity of 99.51%99.51\% for a single qubit gate using interleaved randomized benchmarking. Moreover, a two-qubit holonomic control-phase gate has been implemented where the gate error can be reduced by as much as 18%18\% compared with NHQC.

Introduction. Quantum computation Nielsen and Chuang (2002); Arute et al. (2019), which can provide an unprecedented computational power over classical computation, relies heavily on high fidelity quantum manipulations. However, there are two main obstacles to achieve high fidelity quantum gates. One is the decoherence caused by inevitable interaction between environment and quantum systems, and the other is the imperfection of control pulses caused by the crosstalk between control lines and signal deformation. To cope with these problems, different approaches have been developed, including quantum error correction (QEC) Gottesman ; Knill (2005) and decoherence-free systems (DFS) Duan and Guo (1997); Zanardi and Rasetti (1997); Lidar et al. (1998). Recently, it has been proposed to utilize geometric phases Wilczek and Zee (1984); Berry (1984); Aharonov and Anandan (1987) in quantum system to construct gate operation, for constructing geometric gates due to their inherent robustness under fluctuations of the orbital trajectory in control space De Chiara and Palma (2003); Zhu and Zanardi (2005); Filipp et al. (2009); Berger et al. (2013).

Geometric phase, as the key element in this protocol, has been studied theoretically and experimentally for decades Falci et al. (2000); Whitney et al. (2005); Leek et al. (2007); Möttönen et al. (2008): geometric phase can be either Abelian (Berry phase) or non-Abelian. The non-abelian geometric phase, due to its noncommutativity, can naturally lead to universal quantum computation (so-called Holonomic quantum computation) Zanardi and Rasetti (1999); Pachos et al. (1999); Duan et al. (2001); Albert et al. (2016). Originally, Holonomic quantum computation (HQC) was constructed via adiabatic evolutions, which lead to lengthy gate time and hence induce higher decoherence error. Subsequently, Nonadiabatic Holonomic quantum computation (NHQC)Sjöqvist et al. (2012); Xu et al. (2015); Zhu et al. (2019); Herterich and Sjöqvist (2016), in which the adiabatic constrain does not exist hence shorten the gate time, was proposed and several experimental demonstrations of NHQC have been realized on NMR Feng et al. (2013); Li et al. (2017), NV center in diamond Arroyo-Camejo et al. (2014); Zu et al. (2014); Sekiguchi et al. (2017); Zhou et al. (2017); Nagata et al. (2018); Ishida et al. (2018) and superconducting circuits Abdumalikov et al. (2013); Xu et al. (2018); Egger et al. (2019). However, this NHQC condition imposes stringent requirements on the driving Hamiltonian; the systematic errors would introduce additional fluctuating phase shifts, smearing the geometric phases Zheng et al. (2016); Yan et al. (2019); Liu et al. (2019). Furthermore, even for small-angle rotations, all operations need to be completed with the same duration of time.

Here, we experimentally demonstrate a new scheme of holonomic quantum gates, where a unconventional nonadiabatic and non-Abelian geometric phase is employed to construct quantum gates. The scheme is called time-optimal unconventional non-adiabatic holonomic quantum computation (TOUNHQC) Liu et al. , which can construct arbitrary nonadiabatic and non-Abelian geometric gates with the minimal time. Thus, TOUNHQC offers the unique stability of robustness to decoherence errors and control parameter variations.

For an experimental demonstration, we experimentally realize a universal set of single-qubit and two-qubit unconventional holonomic quantum gates in an Xmon-type of superconducting circuit. Compared with the conventional NHQC implementations, this scheme is more robust against control error by removing the fixed pulse area limitation. Moreover, by combining with time-optimal technology, we further minimize the pulse time for both single and two qubit gate, hence achieving less decoherence error than conventional NHQC. We use gate in single qubit and C-phase gate in two qubits as examples to demonstrate the advantage of our routine.

Setting the stage. Similar to the traditional holonomic gate, an Λ\Lambda shape Abdumalikov et al. (2013); Gu et al. (2017) Hamiltonian is constructed in our quantum system by applying driving microwave, which can be written as

H1(t)=Ω0e(t)2eiϕ0|0e|+Ω1e(t)2eiϕ1|1e|+H.c.,H_{1}(t)=\frac{\Omega_{0e}(t)}{2}e^{i\phi_{0}}|0\rangle\langle{e}|+\frac{\Omega_{1e}(t)}{2}e^{i\phi_{1}}|1\rangle\langle{e}|+H.c., (1)

where energy levels |0|0\rangle and |1|1\rangle are computational basis which are both coupled to an auxiliary level |e|e\rangle. By setting Ω0e(t)=Ω(t)sin(θ/2)\Omega_{0e}(t)=\Omega(t)\sin(\theta/2) and Ω1e(t)=Ω(t)cos(θ/2)\Omega_{1e}(t)=\Omega(t)\cos(\theta/2), we can obtain bright state |b=sin(θ/2)eiϕ|0+cos(θ/2)|1|b\rangle=\sin(\theta/2)e^{i\phi}|0\rangle+\cos(\theta/2)|1\rangle and dark state |d=cos(θ/2)eiϕ|0sin(θ/2)|1|d\rangle=\cos(\theta/2)e^{i\phi}|0\rangle-\sin(\theta/2)|1\rangle, where ϕϕ0ϕ1\phi\equiv\phi_{0}-\phi_{1}. We shall keep θ\theta time-independent, the above Hamiltonian Eq. (1) can then be rewritten as:

H1(t)=Ω(t)2eiϕ1(t)|be|+H.c.H_{1}(t)=\frac{\Omega(t)}{2}e^{i\phi_{1}(t)}|b\rangle\langle{e}|+H.c. (2)

Time-optimal unconventional holonomic gates. To realize holonomic gates, we need to choose a set of auxiliary states {|μk(t)}\{|\mu_{k}(t)\rangle\}, which satisfy the following conditions of (i) the cyclic evolution Π1(0)=Π1(τ)=|dd|\Pi_{1}(0)=\Pi_{1}(\tau)=|d\rangle\langle d|, Π2(0)=Π2(τ)=|bb|\Pi_{2}(0)=\Pi_{2}(\tau)=|b\rangle\langle b|, and (ii) the von Neumann equation ddtΠk(t)=i[H1(t),Πk(t)]\frac{d}{dt}\Pi_{k}(t)=-i\left[H_{1}(t),\Pi_{k}(t)\right], where Πk(t)|μk(t)μk(t)|\Pi_{k}(t)\equiv\left|\mu_{k}(t)\right\rangle\left\langle\mu_{k}(t)\right| denotes the projector of the auxiliary basis. Here, we choose a decoupled dark state |μ1=|d|\mu_{1}\rangle=|d\rangle and an orthogonal state |μ2(t)|\mu_{2}(t)\rangle, which can be parameterized with two angles χ\chi, α(t)\alpha(t) and η(t)\eta(t) as |μ2(t)=(cosη(t)isinη(t)cosχ)|bisinη(t)sinχeiα(t)|e\left|\mu_{2}(t)\right\rangle=(\cos\eta(t)-i\sin\eta(t)\cos\chi)|b\rangle-i\sin\eta(t)\sin\chi e^{i\alpha(t)}|e\rangle. Using the von Neumann equation, we obtain the following coupled differential equations,

Ω(t)=α˙(t)tanχ,ϕ1˙(t)=α˙(t)=2η˙(t)cosχ\Omega(t)=-\dot{\alpha}(t)\tan\chi,\quad\dot{\phi_{1}}(t)=\dot{\alpha}(t)=-2\dot{\eta}(t)\cos\chi (3)

In fact, we have many choices to pick the variables η(t)\eta(t), α(t)\alpha(t) as longle as they meet the coupled equation and the cyclic evolution condition η(0)=0\eta(0)=0 and η(τ)=π\eta(\tau)=\pi.

Refer to caption
Figure 1: (a) A geometric picture of the single-qubit Holonomic gates in Bloch sphere representation. The orange line is the evolution trajectory of TOUNHQC protocol while the blue one illustrates conventional NHQC protocol. γg\gamma_{g} and γ\gamma are geometric phases acquired from these two gate operations respectively. (b) Optical image of the six qubits sample (three not shown) used in our experiments.

Now, we demonstrate how to build up universal arbitrary holonomic single-qubit gates. Let us start with the an arbitrary initial state as |ψ(0)=a1|μ1+a2|μ2(0)|\psi(0)\rangle=a_{1}|\mu_{1}\rangle+a_{2}|\mu_{2}(0)\rangle with a1,2=μ1,2(0)|ψ(0)a_{1,2}=\left\langle\mu_{1,2}(0)|\psi(0)\right\rangle. Under a cyclic evolution, the state |μ2|\mu_{2}\rangle gains a unconventional geometric phase Liu et al. ; Zhu and Wang (2003); Du et al. (2006) that is |μ2(τ)=ei(γg+γd)|μ2(0)=eiϕg|μ2(0)|\mu_{2}(\tau)\rangle=e^{-i(\gamma_{g}+\gamma_{d})}|\mu_{2}(0)\rangle=e^{-i\phi_{g}}|\mu_{2}(0)\rangle corresponding to the geometric phase γg=α(0)α(τ)sin2η(t)sin2χdα\gamma_{g}=\int^{\alpha(\tau)}_{\alpha(0)}\sin^{2}\eta(t)\sin^{2}\chi d\alpha and the dynamical phase γd=γg+ϕg\gamma_{d}=-\gamma_{g}+\phi_{g}. Note that the unconventional geometric phase γγg=α(τ)α(0)\gamma\equiv\gamma_{g}=\alpha(\tau)-\alpha(0) is just the half of the azimuthal angle difference between initial and final points, which is also only determined by the geometric feature of the evolution path. As a result, the final time evolution operator on the subspace {|b,|d}\{|b\rangle,|d\rangle\} is given by U(τ)=eiγ|bb|+|dd|U(\tau)=e^{i\gamma}|b\rangle\langle b|+|d\rangle\langle d|.

Consequently, the unconventional non-Abelian geometric gate can be spanned by the logical {|0,|1}\{|0\rangle,|1\rangle\} basis as,

U(θ,ϕ,γ)=(cγ2isγ2cθisγ2sθeiϕ1isγ2sθeiϕcγ2+isγ2cθ)=exp(iγ2𝐧σ),\begin{aligned} U(\theta,\phi,\gamma)&=\left(\begin{array}[]{cc}c_{\frac{\gamma}{2}}-is_{\frac{\gamma}{2}}c_{\theta}&-is_{\frac{\gamma}{2}}s_{\theta}e^{i\phi_{1}}\\ -is_{\frac{\gamma}{2}}s_{\theta}e^{-i\phi}&c_{\frac{\gamma}{2}}+is_{\frac{\gamma}{2}}c_{\theta}\end{array}\right)\\ &=\exp\left(-i\frac{\gamma}{2}\mathbf{n}\cdot\sigma\right)\end{aligned}\ , (4)

where cxcosx{c_{x}}\equiv\cos x and sxsinx{s_{x}}\equiv\sin x. Note this operation corresponds to an arbitrary rotation around the axis n=(sinθcosϕ,sinθsinϕ,cosθ)\textbf{n}=(\sin{\theta}\cos{\phi},\sin{\theta}\sin{\phi},\cos{\theta}) by an angle of γ\gamma, which picks up a global phase γ/2\gamma/2. Therefore, it is feasible to implement the unconventional holonomic gate by designing particular ϕ\phi and γ\gamma .

To further speed up the nonadiabatic holonomic gates, we combined our unconventional scheme with time-optimal control technology Carlini and Koike (2012); Wang et al. (2015) to shorten the loop path and minimum the gate time against the influence of the environmental induced decoherence effect. In the time-optimal scheme, the phase ϕ1(t)\phi_{1}(t) in the Eq. 2 equals to 2(πγ)t/τ2(\pi-\gamma)t/\tau, with operation time τ=2π2(πγ)2/Ω0\tau=2\sqrt{\pi^{2}-(\pi-\gamma)^{2}}/\Omega_{0}. Comparing with the conventional NHQC, the TOUNHQC scheme generally has a shorter gate time Liu et al. .

Experimental results of time-optimal holonomic single-qubit gate These experiments are demonstrated on a chain of superconducting Xmon-shaped qubits Clarke and Wilhelm (2008); Devoret and Schoelkopf (2013); Barends et al. (2013), as shown in Fig. 1(b). There are six qubits on the chip (three of them not shown) and each qubit is individually coupled to a λ/4\lambda/4 resonator, which is coupled to a common feed line for readoutaSu . These qubits are coupled to each other with a static capacitive coupling strength of g/2π25g/2\pi\approx 25 MHz. Among them, two qubits highlighted in red (denoted as Q1Q_{1} and QaQ_{a}) are tunable transmons while the one in blue (denoted as Q2Q_{2}) is a fixed-frequency transmon.

We first demonstrate single-qubit time-optimal holonomic gates by following the single-loop protocol in the reference. The experiment is performed on the three lowest energy levels |0,|e,|1|0\rangle,|e\rangle,|1\rangle of a superconducting qubit (Q2Q_{2}), as shown in Fig. 2(a). Among them, the ground state |0|0\rangle and the second excited state |1|1\rangle are used to construct the qubit computational basis, while the first excited state |e|e\rangle is regarded as an auxiliary energy level.

Refer to caption
Figure 2: (a) The energy level configuration of a qutrit. Two microwave are driven resonantly with |0|e|0\rangle\leftrightarrow|e\rangle and |e|1|e\rangle\leftrightarrow|1\rangle transitions. (b) States transfer during the X\sqrt{X} gate. The experimental data (blue circles, red diamonds and green triangles) shows a good agreement with the simulation results. (c) State evolution of X\sqrt{X} gate of TO-UNHQC demonstrated in the |0|1|0\rangle-|1\rangle subspace, where the pentagram and triangle represent the beginning and end of the evolution procedure respectively. Solid dots and lines represent experimental data and theoretical simulation respectively. Theoretical simulation of the gate based on NHQC protocol is also shown for comparison. (d) To evaluate the performance of our routine, we choose Clifford-based randomized benchmarking (RB) Chow et al. (2009); Knill et al. (2008); Magesan et al. (2011) to extract the fidelity.

Here we take a X\sqrt{X} gate as an example, which is performed on the logical qubit initialized in state |0|0\rangle. The rotation operator in Eq. 4 is set as U(π/2,0,π/2)U(\pi/2,0,\pi/2). In our approach, we fix Ω0/2π=8.660\Omega_{0}/2\pi=8.660 MHz, making the driving amplitudes Ω0e/2π=Ωe1/2π=6.124\Omega_{0e}/2\pi=\Omega_{e1}/2\pi=6.124 MHz, with the evolution time τ=100\tau=100 ns. To characterize state evolution in the gate operation, we execute state tomography during entire procedure. The populations of |0|0\rangle, |1|1\rangle and |e|e\rangle are illustrated in Fig. 2(b), while state trajectory in Bloch sphere spanned by computational basis is shown in Fig. 2(c). There is a good agreement between the experimental data and the theoretical result. Furthermore, we show the direct comparison of TOUNHQC (red) and NHQC (blue) in terms of the X\sqrt{X} gate. Notice that in our approach, the evolution trajectory in the Bloch sphere is shorter due to time-optimal control. The total procedure times for the protocols are τTOUNHQC=100\tau_{TOUNHQC}=100 ns and τNHQC=115.47\tau_{NHQC}=115.47 ns respectively, displaying advantages of our protocol in operation time.

Refer to caption
Figure 3: (a)Q1Q_{1} (control qubit) and Q2Q_{2} (target qubit), form the Hilbert space spanned by {|0C0T,|0C1T,|1C0T,|1C1T}\{|\rm 0_{C}0_{T}\rangle,|0_{C}1_{T}\rangle,|1_{C}0_{T}\rangle,|1_{C}1_{T}\rangle\}. The two-qubit gates are realized by interaction between the |01|01\rangle and an ancillary state |a|a\rangle of QaQ_{a}. The interaction is activated by modulating the frequency of QaQ_{a} with an ac flux through the flux bias line (FBL). (b) The pulse sequence to perform Ramsey interferometer. (c) The results of Ramsey interferometer with control-T gate on (red circle) or off (blue triangle). The dash lines are used as guidelines. (d) States transfer during the control-T gate. Experimental results fit well with numerical simulations (dash lines). The error bars indicate standard deviation of the data.

Experimental results of time-optimal holonomic two-qubit entangled gate. A universal set of gates requires a kind of two qubit operation. We can demonstrate nontrivial two-qubit C-phase gates of TOUNHQC using the three qubits sample. In practice, we choose control-RkR_{k} gate as an example,

Rk=(100ei2π2k),R_{k}=\begin{pmatrix}1&0\\ 0&e^{i\frac{2\pi}{2^{k}}}\end{pmatrix}, (5)

which is important for realizing a fast quantum Fourier transform Nielsen and Chuang (2002). Here we utilize the lowest two levels of Q1Q_{1} and Q2Q_{2} to form the logical computation space spanned by {|00,|01,|10,|11}\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\} while use the first excited state of QaQ_{a} (denoted as |a|a\rangle) as an ancilla state, as seen in the left panel of Fig. 3a. In the C-phase routine, |b|b\rangle and |e|e\rangle in Eq. 2 are regarded as |01|01\rangle and |a|a\rangle respectively. To implement C-phase gate, we use parametric modulation to activate the interaction between |01|01\rangle and |a|a\rangle Caldwell et al. (2018); Didier et al. (2018); Reagor et al. (2018).

Without loss of generality, we demonstrate the Control-T in practice, by setting value of nn in Eq. 5 as 3. The parameters θ,ϕ,γ\theta,\phi,\gamma in Eq.4 are set as {0,0,π/4}\{0,0,\pi/4\} respectivelySup . Consequently, we set ξ=π/4\xi=\pi/4 in the experiment, then the unitary operator in the subspace {|00,|01,|10,|11}\{|00\rangle,|01\rangle,|10\rangle,|11\rangle\} is written as

U=(10000eiπ40000100001)U=\begin{pmatrix}1&0&0&0\\ 0&e^{i\frac{\pi}{4}}&0&0\\ 0&0&1&0\\ 0&0&0&1\end{pmatrix} (6)
Refer to caption
Figure 4: (a) Superiority of the TOUNHQC scheme in terms of gate operation time. The Fidelity of the gate based on TOUNHQC (b) and NHQC (d) scheme varied as a function of both pulse amplitude error δA/A\delta A/A and the pulse detuning error δυ/υ\delta\upsilon/\upsilon. Both the amplitude and detuning error are range from -0.05 to 0.05. (c) and (e) are the corresponding simulation results.

To verify that Q2Q_{2} indeed picks up ξ=π/4\xi=\pi/4 after the gate, we execute the Ramsey interferometer, whose pulse sequence is shown in Fig. 3(b). We initially prepare Q2Q_{2} in (|0i|1)/2(|0\rangle-i|1\rangle)/\sqrt{2} while Q1Q_{1} and QaQ_{a} remain in |0|0\rangle. Then a π/2\pi/2 pulse around axis (cosθ,sinθ,0\rm cos\rm\theta,\rm sin\rm\theta,0) is applied right after the two-qubits Holonomic gate. As seen in Fig. 3(c), the final population with and without the Holonomic gate are plotted as the function of θ\theta respectively, demonstrating the qubit indeed acquires a geometric phase about π/4\pi/4. We also measured the populations of each states in the entire procedure besides the finite rising and falling edge, as shown in Fig. 3(d). It is noticed that our data is in good agreement with the simulation result Johansson et al. (2012).

The superiority of TOUNHQC. Next, we demonstrate the superiority of the TOUNHQC protocol in terms of gate operation time. Here we use the interleaved RB to calibrate the performance of time-optimal and conventional methods affected by decoherence. The pulse sequence of RB approach is set as: a randomly selected Clifford gate and a subsequent unit gate with the specific Holonomic T gate length. To compare these two approaches, we set the same amplitude of driving microwave Ω\Omega, resulting the gate times τTOUNHQC=7π/2Ω\tau_{TOUNHQC}=\sqrt{7}\pi/2\Omega and τNHQC=2π/Ω\tau_{NHQC}=2\pi/\Omega. The shorter operation time means less decoherence effect. Therefore, RB with our time-optimal protocol has higher fidelity. As shown in Fig. 4(a), by fitting the curve of experimental data with function F=Apm+BF=Ap^{m}+B, we obtain the experimental results for PTOUNHQCP_{TOUNHQC} and PNHQCP_{NHQC} as 0.9230 and 0.9051 respectively.

To further demonstrate that our scheme is more robust than the conventional NHQC, we measure gate fidelities under the imperfections of control parameters for both our unconventional Holonomic gate and the conventional Holonomic gate. The unattenuated fidelities here are defined by Funatt=Tr(ρthρout)/Tr(ρthρth)Tr(ρoutρout)F_{unatt}=\rm{Tr}(\rho_{th}\rho_{out})/\sqrt{\rm{Tr}(\rho_{th}\rho_{th})\rm{Tr}(\rho_{out}\rho_{out})}, where the theoretic output state is (|0+eiπ/4|1)/2(|0\rangle+e^{-i\pi/4}|1\rangle)/\sqrt{2} with initial state in (|0i|1)/2(|0\rangle-i|1\rangle)/\sqrt{2}. The maximal experimental unattenuated fidelity is found approach 0.996 for our scheme while the simulate one can reach 0.999 in absence of decoherence Sup . As shown in Fig. 4(b) and (d), our gate is less sensitive to the detuning error than the conventional Holonomic gate. This can be attribute to our scheme remove the fixed pulse area limitation, thus further improve the robustness against control noises Zhu and Wang (2003); Zheng (2004).

Summary. We have experimentally demonstrated the single- and two-qubit gates for the TOUNHQC scheme. We characterize that our routine has a high gate fidelity by interleaved randomized benchmark. Moreover, we prove our two-qubit gate indeed more robust against certain control noises than the conventional NHQC gate by measuring the fidelity under these control imperfections. Therefore our scheme is a promising candidate to realize quantum state transfer and quantum gates in future.

This work was supported by the Key-Area Research and Development Program of GuangDong Province (Grant No. 2018B030326001), the NKRDP of China (Grant No. 2016YFA0301802), and the NSFC (Grants No. 11604103, No. 11474153, No. 91636218, No. 11890704, No. 61521001, No. 11875160 and No. U1801661), the Natural Science Foundation of Guangdong Province (Grant No.2017B030308003), the Guangdong Innovative and Entrepreneurial Research Team Program(Grant No.2016ZT06D348), the Science, Technology and Innovation Commission of Shenzhen Municipality (Grants No. JCYJ20170412152620376, No. JCYJ20170817105046702, and No. KYTDPT20181011104202253), the Economy, Trade and Information Commission of Shenzhen Municipality (Grant No.201901161512).

References

  • Nielsen and Chuang (2002) M. A. Nielsen and I. Chuang, Quantum computation and quantum information (AAPT, 2002).
  • Arute et al. (2019) F. Arute, K. Arya, R. Babbush, D. Bacon, J. C. Bardin, R. Barends, R. Biswas, S. Boixo, F. Brandao, D. A. Buell, B. Burkett, Y. Chen, Z. Chen, B. Chiaro, R. Collins, et al., Nature (London) 574, 505 (2019).
  • (3) D. Gottesman, arXiv:quant-ph/9705052 .
  • Knill (2005) E. Knill, Nature (London) 434, 39 (2005).
  • Duan and Guo (1997) L. M. Duan and G. C. Guo, Phys. Rev. Lett. 79, 1953 (1997).
  • Zanardi and Rasetti (1997) P. Zanardi and M. Rasetti, Phys. Rev. Lett. 79, 3306 (1997).
  • Lidar et al. (1998) D. A. Lidar, I. L. Chuang,  and K. B. Whaley, Phys. Rev. Lett. 81, 2594 (1998).
  • Wilczek and Zee (1984) F. Wilczek and A. Zee, Phys. Rev. Lett. 52, 2111 (1984).
  • Berry (1984) M. V. Berry, Proc. R. Soc. London Ser. A 392, 45 (1984).
  • Aharonov and Anandan (1987) Y. Aharonov and J. Anandan, Phys. Rev. Lett. 58, 1593 (1987).
  • De Chiara and Palma (2003) G. De Chiara and G. M. Palma, Phys. Rev. Lett. 91, 090404 (2003).
  • Zhu and Zanardi (2005) S. L. Zhu and P. Zanardi, Phys. Rev. A 72, 020301(R) (2005).
  • Filipp et al. (2009) S. Filipp, J. Klepp, Y. Hasegawa, C. Plonka-Spehr, U. Schmidt, P. Geltenbort,  and H. Rauch, Phys. Rev. Lett. 102, 030404 (2009).
  • Berger et al. (2013) S. Berger, M. Pechal, A. A. Abdumalikov, C. Eichler, L. Steffen, A. Fedorov, A. Wallraff,  and S. Filipp, Phys. Rev. A 87, 060303(R) (2013).
  • Falci et al. (2000) G. Falci, R. Fazio, G. M. Palma, J. Siewert,  and V. Vedral, Nature (London) 407, 355 (2000).
  • Whitney et al. (2005) R. S. Whitney, Y. Makhlin, A. Shnirman,  and Y. Gefen, Phys. Rev. Lett. 94, 070407 (2005).
  • Leek et al. (2007) P. J. Leek, J. Fink, A. Blais, R. Bianchetti, M. Göppl, J. M. Gambetta, D. I. Schuster, L. Frunzio, R. J. Schoelkopf,  and A. Wallraff, Science 318, 1889 (2007).
  • Möttönen et al. (2008) M. Möttönen, J. J. Vartiainen,  and J. P. Pekola, Phys. Rev. Lett. 100, 177201 (2008).
  • Zanardi and Rasetti (1999) P. Zanardi and M. Rasetti, Phys. Lett. A 264, 94 (1999).
  • Pachos et al. (1999) J. Pachos, P. Zanardi,  and M. Rasetti, Phys. Rev. A 61, 010305(R) (1999).
  • Duan et al. (2001) L.-M. Duan, J. I. Cirac,  and P. Zoller, Science 292, 1695 (2001).
  • Albert et al. (2016) V. V. Albert, C. Shu, S. Krastanov, C. Shen, R.-B. Liu, Z.-B. Yang, R. J. Schoelkopf, M. Mirrahimi, M. H. Devoret,  and L. Jiang, Phys. Rev. Lett. 116, 140502 (2016).
  • Sjöqvist et al. (2012) E. Sjöqvist, D.-M. Tong, L. M. Andersson, B. Hessmo, M. Johansson,  and K. Singh, New J. Phys. 14, 103035 (2012).
  • Xu et al. (2015) G. F. Xu, C. L. Liu, P. Z. Zhao,  and D. M. Tong, Phys. Rev. A 92, 052302 (2015).
  • Zhu et al. (2019) Z. Zhu, T. Chen, X. Yang, J. Bian, Z.-Y. Xue,  and X. Peng, Phys. Rev. Appl. 12, 024024 (2019).
  • Herterich and Sjöqvist (2016) E. Herterich and E. Sjöqvist, Phys. Rev. A 94, 052310 (2016).
  • Feng et al. (2013) G. Feng, G. Xu,  and G. Long, Phys. Rev. Lett. 110, 190501 (2013).
  • Li et al. (2017) H. Li, Y. Liu,  and G. Long, Sci. China Phys. Mech. Astron. 60, 080311 (2017).
  • Arroyo-Camejo et al. (2014) S. Arroyo-Camejo, A. Lazariev, S. W. Hell,  and G. Balasubramanian, Nat. commun. 5, 4870 (2014).
  • Zu et al. (2014) C. Zu, W. B. Wang, L. He, W. G. Zhang, C. Y. Dai, F. Wang,  and L. M. Duan, Nature (London) 514, 72 (2014).
  • Sekiguchi et al. (2017) Y. Sekiguchi, N. Niikura, R. Kuroiwa, H. Kano,  and H. Kosaka, Nat. Photon. 11, 309 (2017).
  • Zhou et al. (2017) B. B. Zhou, P. C. Jerger, V. O. Shkolnikov, F. J. Heremans, G. Burkard,  and D. D. Awschalom, Phys. Rev. Lett. 119, 140503 (2017).
  • Nagata et al. (2018) K. Nagata, K. Kuramitani, Y. Sekiguchi,  and H. Kosaka, Nat. commun. 9, 3227 (2018).
  • Ishida et al. (2018) N. Ishida, T. Nakamura, T. Tanaka, S. Mishima, H. Kano, R. Kuroiwa, Y. Sekiguchi,  and H. Kosaka, Opt. Lett. 43, 2380 (2018).
  • Abdumalikov et al. (2013) J. Abdumalikov, A. A., J. M. Fink, K. Juliusson, M. Pechal, S. Berger, A. Wallraff,  and S. Filipp, Nature (London) 496, 482 (2013).
  • Xu et al. (2018) Y. Xu, W. Cai, Y. Ma, X. Mu, L. Hu, T. Chen, H. Wang, Y. P. Song, Z.-Y. Xue, Z.-Q. Yin,  and L. Sun, Phys. Rev. Lett. 121, 110501 (2018).
  • Egger et al. (2019) D. J. Egger, M. Ganzhorn, G. Salis, A. Fuhrer, P. Muller, P. K. Barkoutsos, N. Moll, I. Tavernelli,  and S. Filipp, Phys. Rev. Appl. 11, 014017 (2019).
  • Zheng et al. (2016) S.-B. Zheng, C.-P. Yang,  and F. Nori, Phys. Rev. A 93, 032313 (2016).
  • Yan et al. (2019) T. Yan, B.-J. Liu, K. Xu, C. Song, S. Liu, Z. Zhang, H. Deng, Z. Yan, H. Rong, K. Huang, M.-H. Yung, Y. Chen,  and D. Yu, Phys. Rev. Lett. 122, 080501 (2019).
  • Liu et al. (2019) B.-J. Liu, X.-K. Song, Z.-Y. Xue, X. Wang,  and M.-H. Yung, Phys. Rev. Lett. 123, 100501 (2019).
  • (41) B.-J. Liu, Z.-Y. Xue,  and M.-H. Yung, arXiv:2001.05182 .
  • Gu et al. (2017) X. Gu, A. F. Kockum, A. Miranowicz, Y.-x. Liu,  and F. Nori, Phys. Rep. 718, 1 (2017).
  • Zhu and Wang (2003) S.-L. Zhu and Z. D. Wang, Phys. Rev. Lett. 91, 187902 (2003).
  • Du et al. (2006) J. Du, P. Zou,  and Z. D. Wang, Phys. Rev. A 74, 020302(R) (2006).
  • Carlini and Koike (2012) A. Carlini and T. Koike, Phys. Rev. A 86, 054302 (2012).
  • Wang et al. (2015) X. Wang, M. Allegra, K. Jacobs, S. Lloyd, C. Lupo,  and M. Mohseni, Phys. Rev. Lett. 114, 170501 (2015).
  • Clarke and Wilhelm (2008) J. Clarke and F. K. Wilhelm, Nature (London) 453, 1031 (2008).
  • Devoret and Schoelkopf (2013) M. H. Devoret and R. J. Schoelkopf, Science 339, 1169 (2013).
  • Barends et al. (2013) R. Barends, J. Kelly, A. Megrant, D. Sank, E. Jeffrey, Y. Chen, Y. Yin, B. Chiaro, J. Mutus, C. Neill, P. O’Malley, P. Roushan, J. Wenner, T. C. White, A. N. Cleland,  and J. M. Martinis, Phys. Rev. Lett. 111, 080502 (2013).
  • (50) Please see the Supplemental Material for the parameters of transmon.
  • Chow et al. (2009) J. M. Chow, J. M. Gambetta, L. Tornberg, J. Koch, L. S. Bishop, A. A. Houck, B. R. Johnson, L. Frunzio, S. M. Girvin,  and R. J. Schoelkopf, Phys. Rev. Lett. 102, 090502 (2009).
  • Knill et al. (2008) E. Knill, D. Leibfried, R. Reichle, J. Britton, R. B. Blakestad, J. D. Jost, C. Langer, R. Ozeri, S. Seidelin,  and D. J. Wineland, Phys. Rev. A 77, 012307 (2008).
  • Magesan et al. (2011) E. Magesan, J. M. Gambetta,  and J. Emerson, Phys. Rev. Lett. 106, 180504 (2011).
  • Caldwell et al. (2018) S. A. Caldwell, N. Didier, C. A. Ryan, E. A. Sete, A. Hudson, P. Karalekas, R. Manenti, M. P. da Silva, R. Sinclair, E. Acala, N. Alidoust, J. Angeles, A. Bestwick, M. Block, B. Bloom,  and others., Phys. Rev. Appl. 10, 034050 (2018).
  • Didier et al. (2018) N. Didier, E. A. Sete, M. P. da Silva,  and C. Rigetti, Phys. Rev. A 97, 022330 (2018).
  • Reagor et al. (2018) M. Reagor, C. B. Osborn, N. Tezak, A. Staley, G. Prawiroatmodjo, M. Scheer, N. Alidoust, E. A. Sete, N. Didier, M. P. da Silva, E. Acala, J. Angeles, A. Bestwick, M. Block, B. Bloom,  and others., Sci. Adv. 4, eaao3603 (2018).
  • (57) Please see the Supplemental Material for two-qubits gate.
  • Johansson et al. (2012) J. R. Johansson, P. D. Nation,  and F. Nori, Comput. Phys. Commun. 183, 1760 (2012).
  • Zheng (2004) S.-B. Zheng, Phys. Rev. A 70, 052320 (2004).