PHYSICAL REVIEW A 103, 032616 (2021)

Implementing conventional and unconventional nonadiabatic geometric quantum

gates via SU(2) transformations
*
Jian-jian Cheng and Lin Zhang
School of Physics and Information Technology, Shaanxi Normal University, Xi’an 710119, People’s Republic of China

(Received 7 October 2020; accepted 16 March 2021; published 29 March 2021)

We propose a simple but versatile protocol to engineer time-dependent Hamiltonians inversely for geometric
quantum computation. By utilizing SU(2) transformation, a speedup goal on gate operation is achieved with
more freedom to design the control parameters. As an application, this protocol enables the conventional and
unconventional nonadiabatic geometric quantum gates with desired evolution paths by controlling the microwave
pulses in the diamond nitrogen-vacancy center system. We show that the inversely designed Hamiltonian
can fulfill the geometric gate with more economical evolution time and further reduces the influence of the
environment noise on gate fidelity.

DOI: 10.1103/PhysRevA.103.032616

I. INTRODUCTION they still take a longer time to realize geometric computation

beyond decoherence time and do not integrate well with the
Quantum computation exhibits more efficiency than the
experiments. Hence, how to optimize the evolutionary time
classical one in solving some problems, such as factoring
for realizing nonadiabatic geometric quantum computation
large integers, searching big databases, and finding optimal
becomes a topic with great interest.
solutions by quantum annealing [1–3]. However, it still faces
In this paper, we propose a reverse design scheme for GQC
enormous challenges in theory and applications mainly due
in two-level systems. By using a universal SU(2) transforma-
to the inevitable noises introduced by the control errors and
tion to design the evolutionary path of the system, a speedup
the system-environment couplings, which destroy the state
goal of the geometric gate can be achieved in principle with
coherence that plays a kernel role in the parallel computation.
more freedom to select the control parameters. Due to the flex-
To overcome this challenge, geometric quantum computa-
ibility of this approach, the conventional and unconventional
tion (GQC) has been proposed [4]. As the geometric phase
nonadiabatic geometric gates with desired evolution paths can
is solely related to the structure of an evolution path and
be freely designed under the conditions of cyclic evolution and
independent of the middle details, a quantum gate designed
parallel transportation. This approach is much more powerful
with geometric phase is immune to local disturbances during
to find better evolutionary paths and can be well integrated
the evolution [5–7]. However, the scheme of the geometric
with the experiments. As a demonstration, we adopt the dia-
gate based on adiabatic geometric phase [8,9] suffers slowly
mond nitrogen-vacancy (NV) center system as a platform to
cyclic evolutions. Although the adiabatic geometric gate is
illustrate this approach. The nonadiabatic geometric gate in
robust against control errors, the lengthy gate operation time is
this system can be realized by manipulating solid-state spins
still vulnerable to the environment-induced decoherence [10].
in the NV center by appropriately controlling the amplitude,
To relax the limit of evolution speed, nonadiabatic geometric
phase, and frequency of the pulsing fields. Our approach
quantum computation and nonadiabatic holonomic quantum
allows us to execute geometric gates with shorter operation
computation were proposed [11–23]. It has been proved that
times than those in the previous schemes, so the influence of
the implementation of high-speed gates for quantum compu-
environment noises on the quantum gates can be reduce fur-
tation is plausible [24].
ther. In particular, our inverse-engineering framework is more
As the traditional GQC should undergo cyclic evolutions
compatible and conducive to solving the parametric matching
and eliminate the dynamical phase to keep its gauge invari-
problem for different computing platforms.
ance, the evolution paths were mainly restricted to special
forms such as the former multiple loops and the newly orange-
slice-shaped loops [25–29]. The multiple-loop scheme adopts
II. GENERAL FRAMEWORK BY
several closed loops to cancel out the dynamical phase and
TRANSFORMATION METHOD
the orange-slice-shaped-loop takes the geodesic path on Bloch
sphere to eliminate the dynamical phase during the evolution. In this section, we will give a general idea of the trans-
Although the paths in the orange-slice-shaped-loop scheme formation method to realize feasible nonadiabatic geometric
are generally shorter than those in the multiple-loop scheme, quantum computation. Consider a two-level (one-qubit) quan-
tum system described by ( h̄ = 1)

*
zhanglincn@snnu.edu.cn Ĥ0 (t ) = hx (t )σ̂x + hy (t )σ̂y + hz (t )σ̂z , (1)

2469-9926/2021/103(3)/032616(7) 032616-1 ©2021 American Physical Society

JIAN-JIAN CHENG AND LIN ZHANG
where hk (t )(k = x, y, z) are arbitrary real functions of a time
to be designed and σ̂x , σ̂y , σ̂z are Pauli operators. The state
evolution of the system is given by
|(t ) = Û0 (t )|(0),
where the initial state |(0) is reset by initialization and the
evolution operator Û0 (t ) will be properly designed. To engi-
neer feasible Hamiltonians H0 (t ) that give desired dynamics
Û0 (t ), we adopt a unitary transformation of R(t ) parameter-
ized by [30]
 
cos θ 2(t ) −e−iϕ(t ) sin θ 2(t )
R(t ) ≡ R(θ , ϕ) = iϕ(t )
e sin θ 2(t ) cos θ 2(t )
which enables a representation transformation of |ψ (t ) =
R† |(t ). The corresponding Schördinger equation is [31]
∂
i|ψ (t ) = ĤR (t )|ψ (t ),
∂t
and the transformed Hamiltonian is
ĤR (t ) = R† Ĥ0 (t )R + i∂t R† R,
where R† Ĥ0 (t )R is often called the dynamical part, which
is related to dynamical phase and i∂t R† R is the non-Abelian
part, which often brings geometric phase [32]. Then the time-
evolution operator in R representation reads
  t  Although the nonadiabatic geometric gates are robust
ÛR (t ) = T̂ exp −i ĤR (t  )dt  . (5)
0 by the evolution path, an unavoidable challenge remains
To remove the time-ordering operator T̂ to calculate ÛR (t ),
we can design a diagonal form of
ĤR (t ) = F (t )σ̂z
by opportunely choosing the transformation parameters θ (t )
and ϕ(t ) in R(t ). In this case, the time evolution operator
becomes
t  
ÛR (t ) = e−i 0 ĤR (t )dt = e−iγ (t )σ̂z , (7)
t
where γ (t ) = 0 F (t  )dt  . Therefore, the time-evolution oper-
ator in the previous representation can be obtained by
Û0 (t ) = R(t )ÛR (t )R† (0).
To design ĤR (t ) to retain only the diagonal part, a good
choice is to make the nondiagonal parts of R† Ĥ0 (t )R and
i∂t R† R cancel out at any time. Then the diagonal matri-
ces K (t ) ≡ dig[R† H0 (t )R] and A(t ) ≡ dig[i∂t R† R] will safely
lead to Eq. (6). This diagonalization procedure is equivalent
to set a parallel transportation condition for the quantum
computation, which means no transitions admit between two
evolution states in R representation. To confine the control
freedom for a reliable design, we consider a special scheme
that K (t ) is proportional to A(t ), i.e., K (t ) = ηA(t ) [33,34],
which gives
 t  t
γ (t )σ̂z = F (t  )σ̂z dt  = (1 + η)A(t  )dt  , (9)
0 0
where η is a newly introduced constant parameter and η = −1
to avoid a trivial case. The physical meaning of η can be
seen if we set η = 0, the rotation phase, γ (t )σ̂z ≡ γg (t )σ̂z =
t  
0 A(t )dt , reduces to a pure geometric phase without any
PHYSICAL REVIEW A 103, 032616 (2021)
dynamical component. To discriminate them, we denote the
total phase γ (t ) as γ (t ) = (1 + η)γg (t ).
Another condition that needs to be satisfied in quantum
geometric gate is the cyclic evolution, which leads to recy-
(2)
cling transformation R(τ ) = R(0), i.e., R(θτ , ϕτ ) = R(θ0 , ϕ0 ),
where the parameters are labeled by θ (t ) ≡ θt , ϕ(t ) ≡ ϕt for
convenience, then the previous evolution operator will be
Û0 (τ ) = R(0)ÛR (τ )R† (0) = e−iγ n0 ·σ̂ (10)
where n0 ≡ (sin θ0 cos ϕ0 , sin θ0 sin ϕ0 , cos θ0 ) is a unit vector
and Û0 (τ ) conducts a rotation around n0 by an angle 2γ (τ ) =
2(1 + η)γg (τ ). Therefore, an arbitrary geometric gate for a
,
single-qubit can be designed for the original system Ĥ0 (t ).
In the spherical parametric space of (1, θ , ϕ), the transition
curve traces a closed path C of a cyclic evolution during t = 0
to τ and γg (τ ) represents a half of the solid angle enclosed
(3) by path C. This implies that the geometric phase γg (τ ) is
only determined by the evolution path of the parameters θ (t )
and ϕ(t ), and independent of the evolutionary details such as
changing rates of the parameters. This demonstrates that the
(4)
gate is robust against the control errors and depends only on
the topological aspects of the evolution path. As a matter of
fact, γg (τ ) is invariant as long as the area enclosed by the path
does not change, and the rotation axis depends only on the
initial parameters θ0 and ϕ0 .
against the control errors that do not change the area enclosed
that the gates are still vulnerable to environment-induced
decoherence and sequentially impede the experimental imple-
mentation. Fortunately, a nonadiabatic geometric gate with a
(6) determined area has many different evolutionary paths which
can conditionally explore. As some paths have shorter evolu-
tion times than others, the influences of environmental noises
on the quantum gates can reduce.
Based on the above strategy, we propose a scheme of
nonadiabatic GQC with an unconventional geometric phase
that the dynamical phase does not need to be avoided. Al-
though the total phase γ (t ) accumulated in the designed gate
operation contains the dynamical component, it still relies on
(8) global geometric features and the corresponding gate is also
a kind of geometric one in a general sense. The total phase
is proportional to the geometric phase and the correspond-
ing proportional coefficient is constant, independent of (or at
least some) parameters of the qubit system [33]. Under the
circumstances, the total phase possesses the same geometric
robustness as the geometric phase and can be directly used
to realize geometric gate [35]. When η = 0 the dynamical
phase is removed, our scheme will recover the conventional
nonadiabatic schemes.
III. DESIGNED HAMILTONIAN
The main features of our scheme can be summarized as
follows. (1) According to Eq. (8), the evolution operator
Û0 (t ) of system Ĥ0 (t ) with feasible hx,y,z (t ) is designed by
properly choosing R(t ) and ÛR (t ). (2) By the diagonalization
procedure, the target Û0 (t ) satisfies the parallel transport con-
dition. (3) By choosing the closed path in parametric space,
R(t ) meets the cyclic evolution condition. (4) The control
032616-2
IMPLEMENTING CONVENTIONAL AND UNCONVENTIONAL … PHYSICAL REVIEW A 103, 032616 (2021)

functions hx,y,z (t ) are then determined and the inversely de-

signed Hamiltonian H0 (t ) is achieved. In the following, we
elucidate the details.
We adopt a unitary transformation R(t ) on the Hamiltonian
Eq. (1) and obtain explicit forms of R† H0 (t )R and i∂t R† R by
R† Ĥ0 (t )R = fx (t )σ̂x + fy (t )σ̂y + fz (t )σ̂z ,
where
fx (θ , ϕ) = (cos θ cos2 ϕ + sin2 ϕ)hx (t )
θ
− sin2 sin 2ϕhy (t ) FIG. 1. Two Zeeman levels |ms = 0 and |ms = −1 of the NV
2
− sin θ cos ϕhz (t ), spin-triplet ground state are encoded as the qubit states |0 and |1.

hx (t ), hy (t ), and hz (t ) as follows:
θ
fy (θ , ϕ) = − sin 2
sin 2ϕhx (t ) ϕ̇ θ̇
2 hx (t ) = [η − (1 + η) cos θ ] sin θ cos ϕ − sin ϕ,
2 2
+ (cos θ sin2 ϕ + cos2 ϕ)hy (t )
ϕ̇ θ̇
− sin θ sin ϕhz (t ), hy (t ) = [η − (1 + η) cos θ ] sin θ sin ϕ + cos ϕ,
2 2
 
ϕ̇ θ
hz (t ) = sin2 θ + 2η cos θ sin2 ,
fz (θ , ϕ) = sin θ cos ϕhx (t ) 2 2
+ sin θ sin ϕhy (t ) + cos θ hz (t ), and this inversely designed Hamiltonian Ĥ0 (t ) is equivalent to
a general two-level model of
and  −iφ(t )

1 (t ) R (t )e
i∂t R† R = gx (t )σ̂x + gy (t )σ̂y + gz (t )σ̂z , Ĥ0 (t ) = , (13)
2 R (t )eiφ(t ) − (t )
where where
θ̇ ϕ̇ 
gx (θ , ϕ) = sin ϕ + sin θ cos ϕ, R (t ) =θ̇ 2 + ϕ̇ 2 sin2 θ [η − (1 + η) cos θ ]2 ,
2 2

η − 1+η sin 2θ ϕ̇ sin ϕ + θ̇ cos ϕ
θ̇ ϕ̇ φ(t ) = arctan 2
,
gy (θ , ϕ) = − cos ϕ + sin θ sin ϕ, η − 2 sin 2θ ϕ̇ cos ϕ − θ̇ sin ϕ
1+η

2 2  
θ
(t ) = ϕ̇ sin2 θ + 2η cos θ sin2 .
ϕ̇ 2
gz (θ , ϕ) = (1 − cos θ ).
2 It is well known that Eq. (13) can fulfill on many physical
platforms such as the laser-driven atomic system, the quantum
To diagonalize ĤR (t ) according to Eq. (4), the nondiagonal dot spin, or Josephson junction system [37–41], as well as the
components of R† Ĥ0 (t )R and i∂t R† R should cancel out leading diamond NV center controlled by tailoring the parameters of
to fx (t ) = −gx (t ) and fy (t ) = −gy (t ). The diagonal compo- microwave or laser fields [42–46]. This method is generally
nents should be proportional to each other: fz (t ) = ηgz (t ). beyond the adiabatic dynamics without any confinement of
Then we can naturally arrive at the familiar geometric phase slowly varying parameters.
[36]

1 τ IV. GATE IMPLEMENTATION
γg (τ ) = [1 − cos θ (t )]ϕ̇(t )dt. (11)
2 0
To demonstrate our approach, we design geometric rotation
By putting the above integral on a closed path in the para- gates in the diamond NV center system by controlling the
metric space of (1, θ , ϕ), it will be light pulses as an example [47]. A given geometric rotation
 can realize by tailoring (t ) and R (t ) of the microwave
1
γg = (1 − cos θ )dϕ (12) pulses along the desired evolution path in the parametric space
2 C
of (1, θ , ϕ). In the following, we take ÛZ (τ ) = e−iπ σ̂z /2 as a
with a cyclic condition along the closed path, and the total sample target gate.
rotation phase in Eq. (10) reads γ = (1 + η)γg, which indeed As shown in Fig. 1, the NV center has a spin-triplet ground
possesses a global geometric feature beyond the changing state and the nearby nuclear spins (15 N and 13 C) are polarized
rates of the control parameters. by a magnetic field of about 500G along the NV axis. We use
Based on the above conditions, the general geometric gates two lower Zeeman levels |ms = 0 ≡ |0 and |ms = −1 ≡
governed by Eq. (1) can be realized by opportunely choosing |1 of NV center to encode the qubit and the nuclear spins

032616-3
JIAN-JIAN CHENG AND LIN ZHANG PHYSICAL REVIEW A 103, 032616 (2021)

FIG. 2. (Left) The orange-slice-shaped-loop path for the realiza- FIG. 3. (Left) The larger triangular path for the realization of
tion of UZ (τ ) gate. (Right) The evolutions of state populations (|1: UZ (τ ) gate. (Right) The corresponding evolutions of state popula-
red dashed line; |0: blue dot-dashed line) and the fidelities (black tions (|1: red dashed line; |0: blue dot-dashed line) and the fidelities
solid line) along two separate paths. (black solid line) along three separate paths.

of 13C atom for further controls. The qubit is manipulated For this path, the Rabi frequencies of the laser pulses read
by a microwave pulse whose spectrum, intensity, and phase
⎧ −i(ϕ0 − π2 )
can be adjusted by a hybrid waveform generator. The pulse
parameters used here are 0 = 20 MHz (the maximal Rabi ⎨θ̇√(t )e
⎪ , 0  t  τ1 ,
−iφ(t )
frequency), 0 = 20 MHz (the maximal detuning), and the R (t )e = 4 ϕ̇(t )e
3 −iϕ(t )
, τ1 < t  τ2 , (16)
⎪
⎩ −i(ϕ0 + π6 )
π -pulse control time τ is in units of τ0 = π / 0 [48]. θ̇ (t )e , τ2 < t  τ.

Here, their respective pulse areas and detunings are

A. Conventional geometric gate with “orange-slice” path
−i π2 
To realize gate ÛZ (τ ) = e σ̂z , a usual evolution path is se- τ1
2π
lected in parametric space as shown in Fig. 2. The parameters R (t )dt = , (t ) = 0,
0 3
[θ (t ), ϕ(t )] start from the north pole (0, ϕ0 ) to the south pole  √
τ2
(π , ϕ0 ) along the great circle ϕ(t ) = ϕ0 , then return back to 3π 3
R (t )dt = , (t ) = ϕ̇(t ),
the north pole from the south pole along another great circle τ1 6 4
ϕ(t ) = ϕ0 + π2 . This path is the so-called resonant ( = 0)  τ
2π
orange-slice-shaped loop widely used in the control schemes R (t )dt =− , (t ) = 0. (17)
of nonadiabatic geometric quantum computation [49]. To this τ2 3
end, the nondiagonal terms in Eq. (13) are
The total evolution time of this triangular path by the square-
π
−iφ(t ) θ̇ (t )e−i(ϕ0 − 2 ) , 0  t  τ2 , wave pulse is 23 τ0 + 21 τ0 + 23 τ0 ≈ 1.833τ0 , which is shorter
R (t )e = (14)
θ̇ (t )e−iϕ0 , τ2 < t  τ. than that of the orange-slice-shaped loop. Further, a shorter
evolution path to realize this gate can also be designed by
Here, the pulse areas of the Rabi frequencies at their respec- this method if the Rabi frequency (t ) and detuning (t ) can
tive time intervals satisfy reach the experimental maximum at the same time, and more
 τ  τ optimal evolution time will be about 1.792τ0 .
2
R (t )dt = π , R (t )dt = −π . (15)
τ
0 2
C. Unconventional geometric gate
If a square-wave pulse is used to do the calculation, the op- Although the above triangular path evolves faster than
eration time is τ = 2τ0 = 2π / 0 . The geometric phase can the orange-slice one in the conventional geometric gate, the
be calculated from Eq. (12), γg = π /2, which is obtained by requirement of zero dynamical phases imposes stringent con-
the saltation of φ(t ) at the moment of t = τ2 at the south pole. straints on the driving Hamiltonian. However, our method
That is how the conventional nonadiabatic geometric quantum relaxes the experimental conditions and combines geometric
gate is realized via the “orange slice” as shown above. phase control with the nonadiabatic dynamics to validate the
dynamical phase (η = 0) in the gate design. Therefore, we can
B. Conventional geometric gate beyond “orange-slice” path provide better evolution paths with more relaxed experimental
conditions to design (t ) and (t ).
We can choose an alternative evolution path to realize UZ
For example, we can choose the path in this way (see
gate to avoid the singular point at the pole without saltation of
Fig. 4): the parameters [θ (t ), ϕ(t )] start from the north pole
φ. As shown in Fig. 3, the parameters [θ (t ), ϕ(t )] start from
(0, ϕ0 ) to the point ( π2 , ϕ0 ) along the great circle ϕ(t ) = ϕ0 ,
the north pole (0, ϕ0 ) to the point ( 2π , ϕ0 ) along the great
3 then evolve along the equator to ( π2 , ϕ0 + π2 ), and finally
circle ϕ(t ) = ϕ0 . Then the parameters evolve from ( 2π 3
, ϕ0 ) return back to the north pole along the great circle ϕ0 + π2 .
to ( 2π
3
, ϕ0 + 2π
3
) along the arc θ (t ) = 2π
3
, and finally return Along this path, we set η = 1 to make Rabi frequency (t )
back to the north pole along the great circle ϕ(t ) = ϕ0 + 2π 3
. and detuning (t ) both reaching maximum at the same time.

032616-4
IMPLEMENTING CONVENTIONAL AND UNCONVENTIONAL … PHYSICAL REVIEW A 103, 032616 (2021)

that is
⎡ ⎤
↑ (t ) (t )e−iφ(t ) 0 0
1⎢
⎢ (t )e
iφ(t )
− ↑ (t ) 0 0 ⎥
⎥
H2 =
2⎣ 0 0 ↓ (t ) (t )e−iφ(t ) ⎦
0 0 (t )eiφ(t ) − ↓ (t )

in the two-qubit basis. In their respective subspaces of

{|1, ↑, |0, ↑} and {|1, ↑, |0, ↓}, H↑ and H↓ can selectively
satisfy Eq. (8) if the pulse frequency is on resonance with
one computational state |1, ↑ and |0, ↑, and far detuned
from another computational state |1, ↓ and |0, ↓ with
FIG. 4. (Left) The unconventional triangular path for the real- a detuning δω = ↑ − ↓ . The unwanted mixing effect
ization of UZ (τ ) gate. (Right) The corresponding evolutions of state caused by the coupling with the subspace of nuclear spin
populations (|1: red dashed line; |0: blue dot-dashed line) and the pointing downward can neglect when ↓ < δω is satisfied for
fidelities (black solid line) along three separate paths. geometric gate [48,51,52]. With the same routine design as
that of a single-qubit gate Ûsq , we can achieve the nontrivial
geometric two-qubit gate as
Therefore, the Rabi frequencies of the control pulses are
⎧ π Ûtq = |↑ ↑| ⊗ Ûsq + |↓ ↓| ⊗ I. (21)
⎨θ̇ (t )e−i(ϕ0 − 2 ) , 0  t  τ1 ,
−iφ(t ) −iϕ(t )
R (t )e = ϕ̇(t )e , τ1 < t  τ2 , (18)
⎩
θ̇ (t )e−iϕ0 , τ2 < t  τ, V. GATE PERFORMANCE
and their respective pulse areas and detunings satisfy Now we check the reliability of the quantum geometric
 τ1 gates designed by our method in an open system. The per-
π
R (t )dt = , (t ) = 0, formance of a UZ gate, in this case, can be simulated by using
0 2 the Lindblad master equation as
 τ2
π
R (t )dt = , (t ) = ϕ̇(t ), ρ̇(t ) = i[ρ(t ), Ĥ (t )] + 21 [γ1 L(σ̂ + ) + γ2 L(σ̂z )], (22)
τ1 2
 τ
π where ρ(t ) is the density matrix of the designed system,
R (t )dt = − , (t ) = 0. (19) and L(Â) = 2Âρ † − † Âρ − ρ †  is the Lindbladian of op-
τ2 2
erator  (σ̂ + ≡ |1 0|, σ̂z ≡ |1 1| − |0 0|). The decoherent
The total operation time along this path by the square-wave effects of the environment are considered by the damping rates
pulses is 21 τ0 + 21 τ0 + 21 τ0 = 1.5τ0 , which clearly demon- γ1 and γ2 , respectively. In our simulations, the decay and de-
strates that the unconventional GQC owns the shortest gate phasing rates are set γ1 = γ2 = 4 × 104 Hz [48]. Suppose that
time comparing with the conventional ones. In addition, al- the qubit is initially prepared in the state |(0) = √12 (|0 +
though the conventional geometric gate driven by Eq. (17) |1) and the target state is |Z , the time dependence of the
accumulates only pure geometric phase and the unconven- state populations and the state fidelities F (τ ) = | Z |(τ )|2
tional geometric gate driven by Eq. (19) involves nonzero of the UZ gates in different designed paths are shown in
dynamical phase, both target gates acquire the same total Figs. 2, 3, and 4, respectively. We can obtain a final fideli-
phase that is only dependent on the global geometric features. ties of 99.35% for the conventional geometric gate with an
Accordingly, the Hamiltonians for convectional and uncon- “orange-slice” path, 99.57% for the conventional geometric
ventional gates share similarities in the designed forms. More gate in the triangular path, and 99.67% for the unconventional
importantly, the freely adjusting parameter η in our theory can geometric gate.
propose more optimal strategies to do GQC under relaxing
experimental conditions.
VI. CONCLUSION
D. Two-qubit gate In conclusion, we proposed a design method to implement
Moreover, a conditional two-qubit gate will be realized if conventional and unconventional nonadiabatic GQC under
we use two different pairs of orthogonal cyclic states of the the framework of SU(2) transformation, which can realize
target qubit, conditioned on the state of another control qubit. the nonadiabatic geometric gate with a desired evolutionary
The target qubit is exploited by the electron spin of NV center path. Our scheme can minimize the operation time needed for
and one nearby 13 C nuclear spin as the control qubit. In this high-fidelity geometric gates and is better combined with ex-
case, a product one-qubit basis {|0, |1} ⊗ {|↓, |↑} serves as perimental techniques on a general computational platform. In
a two-qubit computational basis, which are coupled by differ- other words, our schemes have more selectivity than the pre-
ent state-selective pulses and radio-frequency fields [50,51]. vious ones towards possible designs of optimal protocols. To
Under the parametric controls of the pulses, the effective show its potential applications, we simulate the performances
Hamiltonian of this two-qubit system has an extensible form of the geometric gates along different parametric paths in the
NV center system, and shorter evolution times and higher gate
Ĥ2 = Ĥ↑ ⊕ Ĥ↓ , (20) fidelities than the previous schemes are explicitly confirmed.

032616-5
JIAN-JIAN CHENG AND LIN ZHANG
ACKNOWLEDGMENTS
We thank Bao-Jie Liu for fruitful discussions. This work
was supported by the National Natural Science Foundation
[1] L. Vandersypen, M. Steffen, G. Breyta, C. Yannoni, M.
Sherwood, and I. Chuang, Nature (London) 414, 883 (2001).
[2] L. K. Grover, Phys. Rev. Lett. 79, 4709 (1997).
[3] M. W. Johnson, M. H. S. Amin, S. Gildert, T. Lanting, F.
Hamze, N. Dickson, R. Harris, A. J. Berkley, J. Johansson, P.
Bunyk, E. M. Chapple, C. Enderud, J. P. Hilton, K. Karimi,
E. Ladizinsky, N. Ladizinsky, T. Oh, I. Perminov, C. Rich,
M. C. Thom, E. Tolkacheva, C. J. S. Truncik, S. Uchaikin,
J. Wang, B. Wilson, and G. Rose, Nature (London) 473, 194
(2015).
[4] P. Zanardi and M. Rasetti, Phys. Lett. A 264, 94 (1999).
[5] S. L. Zhu and P. Zanardi, Phys. Rev. A 72, 020301(R) (2005).
[6] P. J. Leek, J. M. 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).
[7] S. Berger, M. Pechal, A. A. Abdumalikov, Jr., C. Eichler, L.
Steffen, A. Fedorov, A. Wallraff, and S. Filipp, Phys. Rev. A
87, 060303(R) (2013).
[8] M. V. Berry, Proc. R. Soc. Lond. A 392, 45 (1984).
[9] F. Wilczek and A. Zee, Phys. Rev. Lett. 52, 2111 (1984).
[10] Y. X. Du, Z. T. Liang, H. Yan, and S. L. Zhu, Adv. Quantum
Technol. 2, 1900013 (2019).
[11] W. Xiang-Bin and M. Keiji, Phys. Rev. Lett. 87, 097901 (2001).
[12] S. L. Zhu and Z. D. Wang, Phys. Rev. Lett. 89, 097902 (2002).
[13] G. F. Xu, J. Zhang, D. M. Tong, E. Sjöqvist, and L. C. Kwek,
Phys. Rev. Lett. 109, 170501 (2012).
[14] B. J. Liu, X. K. Song, Z. Y. Xue, X. Wang, and M. H. Yung,
Phys. Rev. Lett. 123, 100501 (2019).
[15] E. Sjöqvist, D. M. Tong, L. M. Andersson, B. Hessmo, M.
Johansson, and K. Singh, New J. Phys. 14, 103035 (2012).
[16] Z.-Y. Xue, J. Zhou, and Z. D. Wang, Phys. Rev. A 92, 022320
(2015).
[17] Z. Y. Xue, F. L. Gu, Z. P. Hong, Z. H. Yang, D. W. Zhang, Y.
Hu, and J. Q. You, Phys. Rev. Appl. 7, 054022 (2017).
[18] Z. P. Hong, B. J. Liu, J. Q. Cai, X. D. Zhang, Y. Hu, Z. D. Wang,
and Z. Y. Xue, Phys. Rev. A 97, 022332 (2018).
[19] P. Z. Zhao, K. Z. Li, G. F. Xu, and D. M. Tong, Phys. Rev. A
101, 062306 (2020).
[20] M. Johansson, E. Sjöqvist, L. M. Andersson, M. Ericsson, B.
Hessmo, K. Singh, and D. M. Tong, Phys. Rev. A 86, 062322
(2012).
[21] H. Wu, E. M. Gauger, R. E. George, M. Möttönen, H. Riemann,
N. V. Abrosimov, P. Becker, H.-J. Pohl, K. M. Itoh, M. L. W.
Thewalt, and J. J. L. Morton, Phys. Rev. A 87, 032326 (2013).
[22] N. Ramberg and E. Sjöqvist, Phys. Rev. Lett. 122, 140501
(2019).
[23] J. Jing, C. H. Lam, and L. A. Wu, Phys. Rev. A 95, 012334
(2017).
032616-6
PHYSICAL REVIEW A 103, 032616 (2021)
of China for emergency management project (Grants No.
11447025 and No. 11847308).
[24] G. R. Feng, G. F. Xu, and G. L. Long, Phys. Rev. Lett. 110,
190501 (2013).
[25] S. L. Zhu and Z. D. Wang, Phys. Rev. A 67, 022319
(2003).
[26] Y. Ota and Y. Kondo, Phys. Rev. A 80, 024302 (2009).
[27] J. T. Thomas, M. Lababidi, and M. Tian, Phys. Rev. A 84,
042335 (2011).
[28] G. F. Xu and G. L. Long, Phys. Rev. A 90, 022323 (2014).
[29] T. Chen and Z. Y. Xue, Phys. Rev. Appl. 10, 054051 (2018).
[30] N. V. Vitanov, Phys. Rev. A 85, 032331 (2012).
[31] L. Zhang and W. P. Zhang, Ann. Phys. (NY) 373, 424 (2016).
[32] J. Anandan, Phys. Lett. A 133, 171 (1988).
[33] S. L. Zhu and Z. D. Wang, Phys. Rev. Lett. 91, 187902
(2003).
[34] J. F. Du, P. Zou, and Z. D. Wang, Phys. Rev. A 74, 020302(R)
(2006).
[35] P. Z. Zhao, G. F. Xu, and D. M. Tong, Phys. Rev. A 94, 062327
(2016).
[36] J. Samuel and R. Bhandari, Phys. Rev. Lett. 60, 2339 (1988).
[37] A. A. Abdumalikov, J. M. Fink, K. Juliusson, M. Pechal, S.
Berger, A. Wallraff, and S. Filipp, Nature (London) 496, 482
(2013).
[38] 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).
[39] 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).
[40] 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).
[41] Y. Xu, Z. Hua, T. Chen, X. Pan, X. Li, J. Han, W. Cai, Y. Ma, H.
Wang, Y. P. Song, Z. Y. Xue, and L. Sun, Phys. Rev. Lett. 124,
230503 (2020).
[42] C. Zu, W. B. Wang, L. He, W. G. Zhang, C. Y. Dai, F. Wang,
and L. M. Duan, Nature (London) 514, 72 (2014).
[43] S. Arroyo-Camejo, A. Lazariev, S. W. Hell, and G.
Balasubramanian, Nat. Commun. 5, 4870 (2014).
[44] Y. Sekiguchi, N. Niikura, R. Kuroiwa, H. Kano, and H. Kosaka,
Nat. Photonics. 11, 309 (2017).
[45] B. B. Zhou, P. C. Jerger, V. O. Shkolnikov, F. J. Heremans, G.
Burkard, and D. D. Awschalom, Phys. Rev. Lett. 119, 140503
(2017).
[46] K. Nagata, K. Kuramitani, Y. Sekiguchi, and H. Kosaka, Nat.
Commun. 9, 3227 (2018).
[47] M. W. Doherty, N. B. Manson, P. Delaney, F. Jelezko, J.
Wrachtrupe, and L. C. L. Hollenberg, Phys. Rep. 528, 1
(2013).
IMPLEMENTING CONVENTIONAL AND UNCONVENTIONAL … PHYSICAL REVIEW A 103, 032616 (2021)

[48] Y. Y. Huang, Y. K. Wu, F. Wang, P. Y. Hou, W. B. Wang, W. G.
Zhang, W. Q. Lian, Y. Q. Liu, H. Y. Wang, H. Y. Zhang, L.
He, X. Y. Chang, Y. Xu, and L. M. Duan, Phys. Rev. Lett. 122,
010503 (2019).
[49] P. Z. Zhao, X. D. Cui, G. F. Xu, E. Sjöqvist, and D. M. Tong,
Phys. Rev. A 96, 052316 (2017).
[50] Z. T. Liang, X. Yue, Q. Lv, Y. X. Du, W. Huang, H. Yan, and
S. L. Zhu, Phys. Rev. A 93, 040305(R) (2016).
[51] F. Jelezko, T. Gaebel, I. Popa, M. Domhan, A. Gruber, and J.
Wrachtrup, Phys. Rev. Lett. 93, 130501 (2004).
[52] X. Rong, J. Geng, F. Shi, Y. Liu, K. Xu, W. Ma, F. Kong, Z.
Jiang, Y. Wu, and J. Du, Nat. Commun. 6, 8748 (2015).

032616-7