TYPE Original Research

PUBLISHED 25 October 2022

DOI 10.3389/frqst.2022.1026025

Experimental validation of the

OPEN ACCESS Kibble-Zurek mechanism on a
EDITED BY
Erik Torrontegui,
Universidad Carlos III de Madrid, Spain
digital quantum computer
REVIEWED BY
Francesco Plastina, Santiago Higuera-Quintero 1, Ferney J. Rodríguez 1,
University of Calabria, Italy
Ricardo Puebla,
Luis Quiroga 1 and Fernando J. Gómez-Ruiz 2*
Universidad Carlos III de Madrid, Spain 1
Departamento de Física, Universidad de los Andes, Bogotá, Colombia, 2Instituto de Física
*CORRESPONDENCE Fundamental IFF-CSIC, Madrid, Spain
Fernando J. Gómez-Ruiz,
[email protected]

SPECIALTY SECTION
This article was submitted to Quantum The Kibble-Zurek mechanism (KZM) captures the essential physics of
Engineering,
a section of the journal nonequilibrium quantum phase transitions with symmetry breaking. KZM
Frontiers in Quantum Science and predicts a universal scaling power law for the defect density which is fully
Technology
determined by the system’s critical exponents at equilibrium and the quenching
RECEIVED 23 August 2022 rate. We experimentally tested the KZM for the simplest quantum case, a single
ACCEPTED 07 October 2022
PUBLISHED 25 October 2022
qubit under the Landau-Zener evolution, on an open access IBM quantum
computer (IBM-Q). We find that for this simple one-qubit model, experimental
CITATION
Higuera-Quintero S, Rodríguez FJ, data validates the central KZM assumption of the adiabatic-impulse
Quiroga L and Gómez-Ruiz FJ (2022), approximation for a well isolated qubit. Furthermore, we report on extensive
Experimental validation of the Kibble-
Zurek mechanism on a digital
IBM-Q experiments on individual qubits embedded in different circuit
quantum computer. environments and topologies, separately elucidating the role of crosstalk
Front. Quantum. Sci. Technol. between qubits and the increasing decoherence effects associated with the
1:1026025.
doi: 10.3389/frqst.2022.1026025 quantum circuit depth on the KZM predictions. Our results strongly suggest that
COPYRIGHT
increasing circuit depth acts as a decoherence source, producing a rapid
© 2022 Higuera-Quintero, Rodríguez, deviation of experimental data from theoretical unitary predictions.
Quiroga and Gómez-Ruiz. This is an
open-access article distributed under
the terms of the Creative Commons KEYWORDS

Attribution License (CC BY). The use, IBM quantum computing, Kibble-Zurek mechanism, Landau-Zener model, adiabatic-
distribution or reproduction in other
impulse approximation, quantum technologies
forums is permitted, provided the
original author(s) and the copyright
owner(s) are credited and that the
original publication in this journal is Characterizing the non-equilibrium dynamics in noisy intermediate-scale quantum
cited, in accordance with accepted
academic practice. No use, distribution (NISQ) devices plays an important role in developing both hardware and architecture
or reproduction is permitted which does designs in the search for scalable quantum computers. NISQ devices have recently
not comply with these terms.
attracted tremendous interest, resulting in rapid progress in fundamental studies of novel
hardware and architecture together with promising potential for quantum computing
(Preskill, 2018; Bharti et al., 2022). For example, advancements in NISQ devices
demonstrate a “quantum advantage” in solving sampling problems (Arute et al., 2019;
Zhong et al., 2020; Mooney et al., 2021). To further improve quantum advantage, it is
desirable that devices show important features such as high-fidelity gates, qubits with long
coherence times, control of state preparation and measurement. (Flammia and Liu, 2011;
da Silva et al., 2011; Proctor et al., 2019). Open-access/online NISQ devices have recently
become readily available, such as those provided publicly by the IBM Quantum
Experience platform (IBM-Corporation, 2022), showing a significant improvement in
the last few years. Despite suffering from noise and scalability limitations, this platform

Frontiers in Quantum Science and Technology 01 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

offers a unique possibility to experiment with actual few qubit range of parameters. In this work, we test the KZM adiabatic-
quantum devices in order to carry out a rigorous study of impulse assumption on the simplest, but important case of a
dynamical quantum properties in different settings along the single qubit (LZ model), through experiments on the Qiskit
real time-dynamics of quantum hardware. A key feature of merit (Andersson et al., 2020) simulator and real quantum
in the current NISQ regime is the ability to simulate non- hardware, establishing the limits required to obtain accurate
equilibrium quantum dynamics. The Kibble-Zurek mechanism results in each case. We successfully reproduced the LZ
(KZM) (Kibble, 1976; Kibble, 1980; Zurek, 1985; Zurek, 1993) is a dynamics under a discrete time evolution in current IBM
prominent paradigm to unravel signatures of universal dynamics quantum devices which can provide information about
in the scenario of a finite-rate spontaneous symmetry breaking. dynamics state evolution given that error mitigation
The KZM predicts the production of topological defects (kinks, procedures were implemented. Additionally, noticeable effects
vortices, strings) or in general, non-equilibrium excitations (in of decoherence are observed and explained by a simple
both short- and large-ranged interacting systems) in the course of phenomenological model of relaxation and dephasing for
either quantum (Dziarmaga, 2005; Zurek et al., 2005; Acevedo open quantum systems. Furthermore, analysis and estimation
et al., 2014) or classical (Kibble, 1980; Zurek, 1985) phase of the experimental asymptotic probability allows us to verify the
transitions. The key result of KZM is concerned with the fact universal KZM in a timescale appropriate for an almost closed
that the mean value of density of topological defects scales as a system under an adiabatic quench regime. In summary, the key
power law of the quench rate. Furthermore, new evidence of achievement of this work has been the validation of a central
scaling in the high-order cumulants has also been recently shown premise of KZM through a protocol to characterize and obtain an
(Del Campo, 2018; Gómez-Ruiz et al., 2020). These theoretical effective time-dependent dynamics on IBM realistic quantum
predictions have been observed in various experimental computers. For reaching such goal we performed LZ evolution
platforms such as Bose Gas (Goo et al., 2021), trapped ions under different annealing times, maintaining a fixed number of
(Cui et al., 2020), quantum annealer (Bando et al., 2020; King total gates, a basic benchmark procedure on quantum critical
et al., 2022), Bose-Einstein Condensate (Damski and Zurek, phenomena in near term quantum computers.
2007; Anquez et al., 2016) and Rydberg atoms (Keesling et al., This paper is organized as follows. A brief review on KZM,
2019). the LZ model and its close connection with KZM are presented in
Damski et al. (Damski, 2005; Damski and Zurek, 2006; Sect. 1. In Sect. 2 we present the experimental platform. The
Cucchietti et al., 2007) established a close relationship contrast between theoretical predictions and experimental results
between second order quantum phase transitions and avoided is collected in Sect. 3. Finally, we summarize the main
level crossing evolutions, thus establishing the Landau-Zener conclusions in Sect. 4.
(LZ) model itself as the simplest paradigmatic scenario for
probing KZM (Landau, 1932a; Landau, 1932b; Majorana,
1932; Stückelberg, 1932; Zener and Fowler, 1932). The density 1 Theoretical background
of topological defects can be expressed as a transition probability
for a two-level system. Therefore, this relationship can be tested 1.1 Brief review of the Kibble-Zurek
in generic single qubit platforms. This relationship has been Mechanism
probed by using optical interferometry (Xu et al., 2014),
superconducting qubits (Wang et al., 2014; Gong et al., 2016) The KZM describes the dynamics of a system across a
and trapped ion systems (Cui et al., 2016). continuous symmetry breaking second-order phase transition
IBM-Q currently grants access up to 5-qubit quantum induced by the change of a control parameter λ. When the system
machines based on superconducting transmon qubits which is driven through the critical point λc, both the correlation length
are controllable using Qiskit, an open-source software ξ and reaction time τ diverge as
development kit (Aleksandrowicz et al., 2019; Andersson
ξ
ξ 0 |ϵ|−] , τ
τ 0 |ϵ|−z] . (1)
et al., 2020). These machines have been successfully utilized in
simulating spin models (Cervera-Lierta, 2018; Rodriguez-Vega where, ϵ
(λ − λc )/λ marks the separation from the critical point.
et al., 2022), topological fermionic models (Koh et al., 2022), The spatial and dynamic equilibrium critical exponents are given by ]
quantum entanglement (Choo et al., 2018; Wang et al., 2018; and z, respectively, while the mesoscopic behavior of the system is
Cruz et al., 2019; Mooney et al., 2019; Pozzobom and Maziero, contained in the dimensional constants ξ0 and τ0. If the quench varies
2019), far-from-equilibrium dynamics (Zhukov et al., 2018), linearly in time, ϵ(t)
t/ta , where ta denotes a quench or annealing
non-equilibrium quantum thermodynamics (Gherardini et al., time scale, the system reaches the critical point at t = 0. Therefore, the
2021; Solfanelli et al., 2021), open-quantum systems (García- equilibrium effective reaction time diverges as Eq. 1. This
Pérez et al., 2020), among others. One of the future advantages of phenomenon is known as critical slowing down and can be used
IBM-Q is the possibility to do simulation of quantum systems to describe the time evolution across a phase transition as a sequence
beyond the maximum limits of classical computer over a wide of three stages. Initially, the system is prepared in the high symmetry

Frontiers in Quantum Science and Technology 02 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

FIGURE 1
Connection between KZM and avoided level crossing in a LZ transition. (A) In a continuous second order phase transition, the reaction time
diverges near the critical point. The KZM approximation takes into account the total dynamics divided in three stages (adiabatic, impulse and
adiabatic) represented by the graduated yellow-dark red-yellow colors and separated by the freeze out-time ^tKZM . (B) The inverse of the energy gap
in LZ exhibits a similar behavior of the reaction time. However, it is not divergent at the crossing point. Similarly, we divided the LZ dynamics in
the same three KZM regimes and separated by the Landau-Zener jump time ^tLZ . Inset: Avoided level crossing LZ.

phase from which it evolves within an adiabatic evolution stage. With σ^n the Pauli matrix along the n ∈ x, y, z direction. We define
Secondly, the evolution enters an impulse stage in the neighborhood the diabatic states as the Hamiltonian eigenvectors when Δ = 0 and
of the phase transition where the system is effectively frozen. Finally, consequently eigenvectors for the Pauli operator σ^z : σ^z |0〉
+1|0〉
when the system is far away from the critical point, the dynamics are and σ^z |1〉
−1|1〉. The respective (diabatic) energy levels are
adiabatic again. These three regimes are schematically represented in E0,1
∓ ε(t)/2. Now, the adiabatic instantaneous eigenvalues
Figure 1A. The three regions are separated by two points marked as E± (t) and eigenstates |E± (t)〉 are solutions of
−^tKZM and ^tKZM , in such a way that the freeze-out occurs at the ^
H(t)|E ± (t)〉
E± (t)|E± (t)〉. The instantaneous gap energy is

instant ^tKZM ~ (τ 0 tz]
a )
1/1+z]
. The main point of the KZM argument given by ΔE
E+ − E−
ε2 (t) + Δ2 (for more details see Ref.
is that the size average or correlation length, ξ,^ of domains in the (Ivakhnenko et al., 2022)). In the main panel of Figure 1B, we
broken symmetry phase is set by the equilibrium correlation length depicted the inverse of the energy gap as a function of time while the
evaluated at the freeze-out time. Therefore, the density of excitations instantaneous adiabatic eigenvalues are shown as an inset in the
−D
resulting from quench evolution scales as ρ ~ ξ^ and goes as Figure 1B. The eigenstates are written as a linear combination of the
D] diabatic states as |ψ(t)〉
α(t)|0〉 + β(t)|1〉. By solving the
1 τ 0 1+z]
ρKZM ~
 , (2) corresponding eigenequation in terms of parabolic cylinder
√
ξ 0 ta functions Dp (z), and using the substitution z
t exp[iπ/4]/ ta ,
where D is the dimensionality of the system. This result was initially we obtain the transition amplitudes

e−i 4
3π
derived in the classical domain (Kibble, 1980; Zurek, 1985) and
subsequently extended to quantum systems (Dziarmaga, 2005; Zurek α(z)
√ δχ 1 D−1−iδ (z) + χ 2 Diδ (iz) , (4)
δ
et al., 2005). Additionally, the KZM has also been extended to novel β(z)
χ 1 D−iδ (z) + χ 2 D−1+iδ (iz).
scenarios including long-range interactions (Acevedo et al., 2014;
where δ = Δ2ta/4 is the adiabaticity parameter. Moreover, χ1 and χ2
Puebla et al., 2019; Puebla et al., 2020), inhomogeneous systems
are found from the initial condition at z = zi (see the section:
(Collura and Karevski, 2010; Dziarmaga and Rams, 2010; Gómez-
Supplementary Data for details of the calculations and derivations):
Ruiz and del Campo, 2019) and nonlinear quenches (Barankov and
3π √
Polkovnikov, 2008; Sen et al., 2008).
ei 4 δ D−1+iδ (izi )α(zi ) − Diδ (izi )β(zi )
χ1
,
δD−1−iδ√(zi )D−1+iδ (izi ) − D−iδ (zi )Diδ (izi ) (5)
i3π
−e 4 δ D−iδ (zi )α(zi ) + δD−1−iδ (zi )β(zi )
1.2 Landau-Zener model χ2
.
δD−1−iδ (zi )D−1+iδ (izi ) − D−iδ (zi )Diδ (izi )
Consider a two-level system, with gap Δ, described by the Notice that, Eqs 4, 5 are valid for any arbitrary initial condition
time-dependent Hamiltonian (Z
1) and final time t. For the experimental implementation discussed
^ (t)
− ε(t)σ^z − Δσ^x .
below, we are interested in studying the system’s evolution from
H (3)
2 2 an initial state starting in the anticrossing point at t = 0. In the

Frontiers in Quantum Science and Technology 03 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

section: Supplementary Data, the formal solutions for this ψ (t)〉 ≈ eiΦ2 E− −^tLZ 〉.
particular initial condition are summarized.
• Adiabatic dynamics: from t
^tLZ to tf = ∞
2
1.3 Connection between the KZM and LZ 〈ψ (t)|E− (t)〉 ≈ A.
evolution
Where Φ1, Φ2 are global phases, and A is a constant.
Here we demonstrate how we can implement a controllable Following the AI, Damski in Refs (Damski, 2005; Damski and
evolution using an IBM-Q quantum simulation, in close analogy to Zurek, 2006). reported the probability of finding the LZ system in
the topological defect formation in KZM. Following the seminal the excited state at tf ≫ tLZ, a calculation we briefly summarize for
arguments exposed in Ref. (Damski, 2005; Damski and Zurek, 2006), the sake of completeness in view of our main experimental in
topological defects can be built into the LZ model by being associated terest.
to the diabatic states. Consider one of the states, such as |0〉, to be a From now on, we focus on the LZ dynamics for the evolution
topologically defected phase and |1〉 a defect-free phase. For example, starting in the ground state at the anticrossing point. The initial
√
in the case of vortices, state |0〉 may be an eigenstate of the angular state at t = 0 is then expressed as |E− (0)〉
(|1〉 − |0〉)/ 2, and
momentum operator L^z |0〉
n|0〉, while L^z |1〉
0. In this consequently the transition probability PAI

scenario, Damski introduces the normalized density of topological |〈E+ (^tLZ )|E− (0)〉|2 is given by (Damski, 2005; Damski and
defects as the average angular momentum Zurek, 2006)

1 1 1
ρKZM
〈ψ L^z ψ 〉
〈ψ|0〉 .
2
(6) PAI

1 − 2 
n 2 1+^ε


Then, a system evolving in time under the LZ model can be used 1 1 2

− 1−  . (9)
to study transitions between the phases through the probabilities 2 2 2 2
ηta + ηta ηta + 4 + 2
of the diabatic states. The similarity between the reaction time of
a second order phase transition and the inverse of energy gap in Where we have fixed the two-level system gap to Δ = 1.
the LZ Hamiltonian is shown in Figure 1B. In analogy with the ε
ε(^tLZ ) is the linear bias at time t
^tLZ .
Additionally, ^
KZM, this suggests that the adiabatic-impulse-adiabatic √
Expanding Eq. 9 into aseries of ta , we obtained (Damski,
approximation (AI) may be used to estimate the asymptotic 2005; Damski and Zurek, 2006)
probability when the system traverses the avoided level crossing,
√ √
thus elucidating the link between the KZM and LZ evolution. 1 η 1/2 η η 3/2
PAI
− ta + t + O t5/2
a . (10)
We divided the dynamics through the anti-crossing into 2 2 8 a
three stages like the AI scenario for KZM. Without loss of
which will be relevant for testing the predictions of the universal
generality, we assume that the system starts at ti → −∞ from
AI for KZM below.
the ground state |E− 〉, and then it evolves to tf → ∞. We define a
natural time scale given by the inverse of the energy gap
1

η^tLZ , (7) 2 Experimental IBM-Q platform
E+ LZ − E− ^tLZ
^
t
We implemented our experimental studies in two topologies or
where E± (t) are the adiabatic energy eigenvalues at time t
^tLZ
processors types. Figure 2A shows the device layout for the IBMQ 5-
and η is a constant. Using Eq. 3, we obtain
 qubit ibmq_bogota (Falcon r5.11L topology QC1 ) and ibmq_lima
  (Falcon r4T topology QC2 ). The topology of the device determines
^tLZ Δ 4

√ 1+ 2 2 − 1. (8) the possible placement of two-qubit gates. The qubits are furthermore
ta 2 Δ ηta
prone to decoherence, thereby requiring several runs of the
experiment to make up for statistical errors. We measure the LZ,
The AI assumes that the evolution wave function |ψ(t)〉 of and concomitant KZM relation, for each one of the IBM-Q
the system satisfies: transmons in QC1 and QC2 . Each transmon plays the role of a
qubit, evolving with its own dynamics, experimentally showing the
• Adiabatic dynamics: from ti = −∞ to t
−^tLZ effects of decoherence on the hardware. Generally, the physical
transmon type qubits of the same machine offer a variety of
ψ (t)〉 ≈ eiΦ1 |E− (t)〉. properties that describe the quality of the qubit, such as thermal
relaxation time (T1), dephasing time (T2), anharmonicity, and error
• Impulse dynamics: from t
−^tLZ to t
^tLZ properties detailed in the section: Supplementary Data, allowing us to

Frontiers in Quantum Science and Technology 04 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

FIGURE 2
Decoherence times in different IBM-Q and IBM-Q circuit simulation of the Landau-Zener process. (A) In the pie-like chart, we contrast the
thermal relaxation time (T1) and dephasing time (T2), in (μs), for two different topology circuits, simply called QC1 and QC2 (see text for details). Due to
in situ IBM machine calibration routines the times T1 and T2 may change. Every pie-like chart is divided into two sectors by a dashed line, where the
upper and lower sectors corresponding to decoherence times at two different dates. (B) Quantum circuit for the LZ simulation starting at the
^ n (θ)|0〉, where U
state |ψ(ti )〉
U ^ n (θ) is a unitary rotation along the axis n. (C) Schematic representation of the LZ transition probability: the solid line
corresponds to the exact result given by Eq. 13, with Δ = 1, ta = 2, ti = 0 and tf = 10, while the symbols illustrate expected results for a grid of points with
separation dt = tf − ti/Nt, being Nt the total circuit depth. The filled dots correspond to: the shortest circuit with depth 1 (gray dot) and an intermediate
circuit depth N (black dot). The inset shows the discrete approximation of the time-dependent component ε(t) of the LZ Hamiltonian.

compare the simulation’s performance with different physical N−1

^ (t, ti ) ≈  e−iH^ k dt ,
U (11)
parameters. In Figure 2A, the times T1 and T2 are depicted for
k
0
each considered circuit topology at two different dates, illustrating in
a graphical way how these times change every time that IBM where H ^ k
H(t
^ i + kdt).
performed a calibration of every device. Since we are interested in the evolution from an initial condition
where the LZ system is prepared in an equal weight superposition at
the anticrossing point, we start by applying the unitary rotation
3 Results U^ y (−π/2). The approximate time evolution operator is constructed
with general unitary gates. A general unitary single qubit gate
3.1 Simulation of the Landau-Zener describes rotations on the Bloch sphere and is defined by three
evolution on IBM-Q Euler angles

Unitary dynamics.– We are interested in the experimental θ θ

⎜ cos
 −eiλ sin
 ⎞
⎛
⎜
⎜
⎜ 2 2 ⎟ ⎟
⎟
⎟
determination, and respective simulation, on a digital open-access ^ θ, ϕ, λ
⎜
U ⎜
⎜
⎜
⎜
⎟
⎟
⎟
⎟
⎟ . (12)
IBM-Q of a single qubit evolution under a linearly time-dependent ⎜
⎜
⎝ iϕ θ θ ⎟
⎟
⎠
i(ϕ+λ)
e sin
 e cos

Hamiltonian (LZ problem). At time ti, a qubit in the processor is 2 2
initialized in the state |ψ(ti )〉
U ^ n^(θ)|0〉, where U ^ n (θ)

cos[θ/2]I^ − i^
σ n sin[θ/2] is a unitary rotation along the axis IBM-Q devices are equipped with the finite and complete set
√
n ∈ x, y, z with σ^n the usual Pauli matrix along the n-direction. {CX, I, Uz , X , X} of basis gates on which any quantum circuit
The whole evolution from ti to tf is performed by sampling the must be decomposed into. The general unitary gate can then be
Hamiltonian at regular intervals dt = (tf − ti)/Nt where Nt denotes the expressed using the previous set as
√ √
number of time steps or the total circuit depth (see blue region in ^ z (ϕ + π) XU
U(θ, ϕ, λ)
eiγ U ^ z (θ + π) XU ^ z (λ), where γ

Figure 2B). The equivalent circuit for the experimental IBM-Q (λ + ϕ + π)/2 is a global phase factor. Using this decomposition,
realization, and its simulation, is shown in Figure 2B. Assuming small time progressions as defined in Eq. 11 are simulated and finally
an evolution governed by a time-independent Hamiltonian and for the state |ψ(t)〉 is measured.
small enough intervals of duration dt, the time evolution operator at As already stated, the Landau-Zener dynamics can be exactly
time t = N dt, with 1 ≤ N ≤ Nt, can be approximated by solved (see Supplementary Data), thus allowing a direct benchmark

Frontiers in Quantum Science and Technology 05 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

test of the experimental results on a realistic quantum device against of gates due to the increase in computing times. We model every
exact results. For a LZ evolution starting at the anticrossing ground qubit on IBM-Q as a two-level system coupled to a Markovian
state, we obtain the LZ transition probability PLZ(t) at time t given as bath. The system evolution is described by a continuous map
ρt
etL ρt0 , t ≥ 0 generated by the Lindbladian L[•]
−i[H, ^ •] +
PLZ (t)
|χ 1 D−iδ (z) + χ 2 D−1+iδ (iz)|2 , (13) † 1 ^† ^
^ ^
n (Ln •Ln − 2 Ln Ln , • ) (Breuer and Petruccione, 2007), where,
with the amplitudes χ1 and χ2, see Eq. 5, given by: H^ is the Hamiltonian and !L^n " are Lindblad operators that
√ describe the system-bath interactions. Dissipative processes in a
2k exp[iπk] ⎡⎣ 2ik Γ(k) + (1 + i)Γ12 + k⎤⎦
χ1
− √ , (14) superconducting qubit such as relaxation, i.e., transitions from
4 ik Γ(2k) the higher energy level |1〉 to ground state |0〉, can be described
√ √
exp[iπk] ⎡⎣2ikΓ12 − k + (1 − i) 2ik Γ(1 − k)⎤⎦ phenomenologically by the operator L^1
Γ|0〉〈1| and
χ2
√
2k+1 Γ(1 − 2k)
. (15) dephasing by rotations around the z axis L^2
γ σ^z .
Additional transitions such as thermal excitations from the
Where z and δ are given in Section 1.2. ground state |0〉 to |1〉 may also be considered (Marquardt
Our first aim is to benchmark our LZ experimental results with and Püttmann, 2008), although for a superconducting
the above exact theoretical prediction. This is schematically illustrated transmon qubit this process is negligible. The rates Γ = 1/T1
in Figure 2C where we display the exact result, see Eq. 13, and a and γ = 1/T2 − 1/2T1 are related to the characteristic times of each
hypothetical grid of points representing expected target data with a physical qubit.
separation dt = (tf − ti)/Nt, being Nt the total circuit depth. For every In Figure 4, we establish a contrast between the unitary exact
experimental data, 5,000 shots have been realized on each quantum dynamics, numerical Lindblad dynamics (QuTip) and the
circuit, QC1 and QC2 .The unit of energy is set by choosing Δ = 1 in experimental results obtained for qubit 4, the noisiest qubit in
the LZ Hamiltonian (see Eq. 3). Therefore, in the following, we both quantum machines. QuTiP is an open-source framework for
express energy parameters and time as dimensionless quantities (Z = Python that allows for numerical simulations of quantum dynamics
1). Using the quantum circuits QC1 and QC2 , we implemented the of open systems under different solvers (Johansson et al., 2012;
corresponding gates in all qubits available on parallel and we did a Johansson et al., 2013). Specifically, we depicted the Landau-Zener
sweep of parameters in annealing time ta from 0.05 to 2.0. probability as a function of the number of layers in the circuit, N, for
Additionally, for both theoretical and experimental results, the two specific annealing times ta = 1 (colors green/purple) and ta = 0.1
final evolution time tf was chosen according with: tf = 4 for (colors blue/orange). Additionally, we show as an inset the ratio
annealing times in the interval 0.05 ≤ ta ≤ 0.17 and tf = 10 for between T2/T1, the bar scale shows the value of this proportion from
0.17 < ta ≤ 2. These particular choices have been supported by the fact 0 to 2. Although, ibmq_lima quantum computer has the ratio T2/T1
that as we are mainly interested in the asymptotic LZ probability almost constant, qubit 4 is the most prone to decoherence.
transition, a good asymptotic collapse is reached for these parameter
regimes. We also represent the experimental results PLZ (N) as a
function of the number of layers in the circuit instead of time. We 3.2 Simulation of the Kibble-Zurek
emphasize that an N-deep circuit corresponds to a physical qubit mechanism on IBM-Q
interaction time tInt = 2tSXN, where tSX is the gate length property for
√
X and it is fixed by IBM-Q as tSX = 35.555 ns. In Figure 3, we The main purpose of this work is to validate the adiabatic-
present a contrast of the LZ transition probability for both the impulse approximation of the Kibble-Zurek mechanism through the
theoretical and experimental results. In the panel Figure 3A, we nonequilibrium dynamics of the Landau-Zener model on IBM-Q.
choose the most robust qubit that better reproduced the theoretical Using Eq. 13 with Δ = 1, the asymptotic probability can be exactly
PLZ. Specifically, we found that the qubit 3 and 2 for ibmq_bogota calculated as
and ibmq_lima, respectively, have the best performance. In order to
1 3πδ
better appreciate the experimental agreement and differences for PLZ (t → ∞)
1 − exp#− $|χ 2 |2 . (16)
δ 2
every single-qubit over QC1 and QC2 , we show the LZ transition
probability as a function of the number of applied gates in Figure 3B. Expanding the asymptotic probability into series, we obtain
In the next subsection, we address the influence of the number of (Damski, 2005; Damski and Zurek, 2006)
layers in the LZ simulation circuit and the role of decoherence. √ √
1 π 1/2 π
Open system dynamics.- The performance of the hardware PLZ (t → ∞)
− ta + (π − ln(4))ta3/2 + O ta5/2 .
2 4 32
worsens with an increasing depth of the circuit. The assumption
(17)
of a closed quantum system rapidly breaks down for qubits with
short relaxation (T1) and dephasing (T2) timescales, thus We find the value of η by directly comparing the adiabatic-impulse
requiring for a theoretical analysis that resorts to a quantum approximation given by Eq. 10 and the expansion of the LZ
open system approach. The effects of quantum decoherence are asymptotic probability at first-order (η = π/4). However, non-
noticeable in the measured probability when scaling the number trivial corrections for high-order terms appear. In both main

Frontiers in Quantum Science and Technology 06 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

FIGURE 3
Measurement of LZ probabilities on IBM-Q. In panel (A), we establish a contrast between the exact and experimental results for the LZ transition
probability as a function the number of layers or circuit depth N, and the annealing time ta. In this panel, all figures share the same color vertical scale.
The initial condition, PLZ (t
0)
0.5, is represented by a red dashed line. Note that in some region of parameters a probability larger than 0.5 for the
experimental results is obtained. In panel (B), the behavior of the LZ probability for every qubit available in each processor is shown, identifying in
this way the most isolated (larger decoherence time) qubit in each case. We fixed the maximum number of layers in the circuit as Nt = 50.

FIGURE 4
Contrast between close and open quantum dynamics for LZ on IBM-Q. The Landau-Zener transition probability is shown as a function of the
number of layers N in the circuit implemention for qubit 4, the noisiest qubit for each QC1 and QC2 quantum circuit. We contrast the theoretical
prediction for a close system (unitary dynamics) given by Eq. 13 (dashed line), the decoherent dynamics given by the numerical solution of the
Lindblad equation (solid line) and experimental results (symbols). The experimental results clearly depart from the unitary evolution prediction
as the number of layers Nt increases in the circuit (see Figure 2B blue region). Additionally, in every panel, we present as inset the ratio between the
dephasing time (T2) and the thermal relaxation time (T1).

Frontiers in Quantum Science and Technology 07 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

FIGURE 5
Simulation of the Kibble-Zurek mechanism on IBM-Q. In both upper and lower panels, we contrast the adiabatic-impulse approximation (Eq. 9),
asymptotic Landau-Zener probability (Eq. 16), and the experimental data. In panel (A), we show experimental data retrieved from ibmq_bogota. In
panel (B), we present the experimental results from ibmq_lima. In the inset, we present the protocol to calculate the asymptotic experimental
Landau-Zener probability. The error bars with length 2ζEB, calculated from the finite-time effect, are also shown. Solid symbols are consistent
with the best qubit behavior as depicted in Figure 3.

panels of Figures 5A,B, we depict the agreement of the theoretical

prediction for the adiabatic-impulse approximation (Eq. 9) and
asymptotic Landau-Zener probability (Eq. 16). We note the role
of the corrections for large quench times. For finite-time LZ
simulations, estimating the asymptotic transition probability
becomes challenging and similar to experimental data. To this
end, we introduced the Landau-Zener jump-time t+ as the fist
zero in the second derivative of the Landau-Zener probability, thus:

d2 PLZ (t)

0. (18)
dt2 t
t+

In this way, we propose that the estimated finite-time FIGURE 6

asymptotic Landau-Zener probability can be approximated by KZM adiabatic-impulse approximation fitting parameters.
From the best qubit experimental data (solid symbols in Figure 5),
the average of all values of PLZ (t) with t ≥ t+. In the inset of the fitting to the KZM adiabatic-impulse approximation PAI (ta)
Figure 5B, we display the protocol implemented to calculate the given by Eq. 19 is probed (the dashed gray lines correspond to
the theoretical predictions). The experimental data at
finite-time asymptotic Landau-Zener probability. Therefore, we ibmq_bogota and ibmq_lima are depicted in blue and red,
establish a finite-time error regime depicted in the main panel of respectively.
Figures 5A,B as error bars using the experimental values of the
annealing time. The estimation of the Landau-Zener jump-time
t+ has been implemented uniquely from the theoretical
prediction, assuming it will be the same for the experimental
In order to further testing the KZM adiabatic-impulse
data. Note that the adiabatic-impulse approximation and the
approximation, from our experimental data, we rewrite the
asymptotic Landau-Zener probability are equivalent in the
Eq. 9 in terms of 3 fitting parameters, as
regime of our experimental data giving confidence in our

validation of the KZM on the IBM-Q platform. 
2
For the qubit with the largest decoherence T1 and T2 times (the PAI (ta ) ≃ x1 − x2 1 −  . (19)
best qubit from now on), the experimental data show an excellent (x3 ta ) + x3 ta (x3 ta )2 + 4 + 2
2

agreement with the theoretical predictions for the impulse-adiabatic

approximation. For large annealing time ta, the experimental data has In Figure 6, we depict the comparison of the fitting parameters x1,
a significant deviation for some qubits in the ibmq_bogota quantum x2 and x3 for the best qubit at ibmq_bogota and ibmq_lima. The
computer. Indeed, the adiabatic-impulse approximation relationship structure of the fitting expression allows us a direct comparison
with the Landau-Zener problem assumes a close system’s quantum with the theoretical predictions (xT1 , xT2 , xT3 ). The first fitting
dynamics. However, since IBM-Q is benchmarked as an open- parameter x1 provides information about how robust the
quantum system, deviations are to be expected. qubit is to decoherence for fast LZ driving. Note that the

Frontiers in Quantum Science and Technology 08 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

theoretical prediction is xT1
1/2 as it is fixed by the initial considered in each device and low annealing times. We find
condition at the anticrossing initial point. Moreover, it fixes the that larger annealing times demand a greater time resolution in
value of the impulse-adiabatic approximation for small annealing the evolution operator discretization, putting practical limits on
times, PAI (ta → 0)
1/2. It is evident from Figure 5 that some the performance achieved, as it becomes limited by the conflict
qubits deviate from this ideal value in this regime, confirming between computing depth and decoherence times. However, the
that these qubits are already highly sensible to decoherence. rapid rate of quantum hardware advances may soon change this.
Nonetheless, for these results, we used the smallest number of Furthermore, an interesting follow-up research direction would
layers considered. The second fitting parameter x2 gives consist in focusing on richer open quantum platforms, where
information about the higher annealing time regime, with KZM has been poorly explored. Thus, using real quantum
theoretical value xT2
1/2. The asymptotic value of the hardware to test quantum universal dynamical behaviors, in
adiabatic-impulse approximation is zero for large annealing both closed and open systems, represent an interesting
times. However, large annealing times imply that the LZ extension of the results presented in this work.
transition probability has several oscillations as a function of
time. Consequently, it is necessary to manage large simulation
times to obtain the asymptotic LZ probability. It is to be expected
Data availability statement
that, our results show deviations due to finite simulation time
The datasets presented in this study can be found in online
effects. Finally, the third parameter x3 validates the Kibble-Zurek
repositories. The names of the repository/repositories and
scaling in the adiabatic-impulse approximation (xT3
π/4). We
accession number(s) can be found below: https://github.com/
found an excellent agreement with the theoretical predictions for
sanhq17/Testing_KZM_IBMQ.
these QC1 and QC2 robust qubits. Thus, by using the close
relationship between the KZM and the LZ transition probability,
we validated and tested the KZM on IBM-Q. These results can be Author contributions
part of a sequence of major steps to fully understand the strength
and limitations of time-dependent quantum simulations. It may FG-R and LQ initiated and guided the project. SH-Q took the
provide insights for designing top efficient quantum simulation experimental measurements. FG-R developed numerical
protocols for more involved out-of-equilibrium and interacting simulations and prepared the figures. All authors contributed
systems. to the analysis of the results and the writing of the manuscript.

4 Conclusion Funding

In this work we explored the dynamics of a two level system SH-Q, FR, and LQ are thankful for the financial support from
under the time-dependent Landau-Zener Hamiltonian on digital Facultad de Ciencias-UniAndes projects: INV-2021-128-2292, and
IBM Quantum computers. Time evolution was simulated by INV-2019-84-1841. FG-R acknowledges financial support from
discretization of the time dependent Hamiltonian and application European Commission FET-Open project AVaQus GA 899561.
of subsequent single-qubit unitary gates representing finite time
progressions. We studied the Landau Zener transition probability as
a function of time by running parallel quantum circuits on 5-qubit Acknowledgments
machines ibmq_lima and ibmq_bogota with different topologies.
We find a strong agreement with the theoretical solution of the LZ The authors thank to Bogdan Damski for useful comments
problem for robust qubits from both machines. We also considered and suggestions.
the effect of decoherence on an open LZ system, modeling the
dissipation using collapse operators for relaxation and dephasing.
For greater trotterizations of the time evolution operator, increasing Conflict of interest
computing time cause noticeable deviations from the theoretical LZ
solution. The numerical solution of the Lindblad master equation The authors declare that the research was conducted in the
accurately depicts the open system’s relaxation towards the ground absence of any commercial or financial relationships that could
state, supported by the measured probabilities. be construed as a potential conflict of interest.
The above positive LZ results allowed us to demonstrate the
first simulation on a realistic quantum computer of the universal Publisher’s note
Kibble-Zurek mechanism by estimating the asymptotic
transition probability obtained from LZ experimental data. All claims expressed in this article are solely those of the
Results show excellent agreement for the best qubits authors and do not necessarily represent those of their

Frontiers in Quantum Science and Technology 09 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

affiliated organizations, or those of the publisher, the Supplementary material

editors and the reviewers. Any product that may be
evaluated in this article, or claim that may be made by The Supplementary Material for this article can be found
its manufacturer, is not guaranteed or endorsed by the online at: https://www.frontiersin.org/articles/10.3389/frqst.2022.
publisher. 1026025/full#supplementary-material
.

References
Abbas, A., Andersson, S., Asfaw, A., Corcoles, A., Bello, L., Ben-Haim, Y., et al. evolutions and back again. Phys. Rev. A 73, 063405. doi:10.1103/PhysRevA.73.
(2020). Learn quantum computation using Qiskit. 063405
Acevedo, O. L., Quiroga, L., Rodríguez, F. J., and Johnson, N. F. (2014). New Damski, B., and Zurek, W. H. (2007). Dynamics of a quantum phase transition in
dynamical scaling universality for quantum networks across adiabatic quantum a ferromagnetic bose-einstein condensate. Phys. Rev. Lett. 99, 130402. doi:10.1103/
phase transitions. Phys. Rev. Lett. 112, 030403. doi:10.1103/PhysRevLett.112.030403 PhysRevLett.99.130402
Aleksandrowicz, G., Alexander, T., Barkoutsos, P., Bello, L., Ben-Haim, Y., Del Campo, A. (2018). Universal statistics of topological defects formed in a
Bucher, D., et al. (2019). Qiskit: An open-source framework for quantum quantum phase transition. Phys. Rev. Lett. 121, 200601. doi:10.1103/PhysRevLett.
computing. doi:10.5281/zenodo.2562111 121.200601
Anquez, M., Robbins, B. A., Bharath, H. M., Boguslawski, M., Hoang, T. M., and Dziarmaga, J. (2005). Dynamics of a quantum phase transition: Exact solution of the
Chapman, M. S. (2016). Quantum Kibble-Zurek mechanism in a spin-1 Bose- quantum Ising model. Phys. Rev. Lett. 95, 245701. doi:10.1103/PhysRevLett.95.245701
Einstein condensate. Phys. Rev. Lett. 116, 155301. doi:10.1103/PhysRevLett.116.
Dziarmaga, J., and Rams, M. M. (2010). Dynamics of an inhomogeneous quantum
155301
phase transition. New J. Phys. 12, 055007. doi:10.1088/1367-2630/12/5/055007
Arute, F., Arya, K., Babbush, R., Bacon, D., Bardin, J. C., Barends, R., et al. (2019).
Flammia, S. T., and Liu, Y. K. (2011). Direct fidelity estimation from few pauli
Quantum supremacy using a programmable superconducting processor. Nature
measurements. Phys. Rev. Lett. 106, 230501. doi:10.1103/PhysRevLett.106.230501
574, 505–510. doi:10.1038/s41586-019-1666-5
García-Pérez, G., Rossi, M. A. C., and Maniscalco, S. (2020). IBM Q experience as
Bando, Y., Susa, Y., Oshiyama, H., Shibata, N., Ohzeki, M., Gómez-Ruiz, F. J.,
a versatile experimental testbed for simulating open quantum systems.
et al. (2020). Probing the universality of topological defect formation in a quantum
npj Quantum Inf. 6, 1. doi:10.1038/s41534-019-0235-y
annealer: Kibble-Zurek mechanism and beyond. Phys. Rev. Res. 2, 033369. doi:10.
1103/PhysRevResearch.2.033369 Gherardini, S., Belenchia, A., Paternostro, M., and Trombettoni, A. (2021). End-
point measurement approach to assess quantum coherence in energy fluctuations.
Barankov, R., and Polkovnikov, A. (2008). Optimal nonlinear passage through a
Phys. Rev. A 104, L050203. doi:10.1103/PhysRevA.104.L050203
quantum critical point. Phys. Rev. Lett. 101, 076801. doi:10.1103/PhysRevLett.101.
076801 Gómez-Ruiz, F. J., and del Campo, A. (2019). Universal dynamics of
inhomogeneous quantum phase transitions: Suppressing defect formation. Phys.
Bharti, K., Cervera-Lierta, A., Kyaw, T. H., Haug, T., Alperin-Lea, S., Anand, A.,
Rev. Lett. 122, 080604. doi:10.1103/PhysRevLett.122.080604
et al. (2022). Noisy intermediate-scale quantum algorithms. Rev. Mod. Phys. 94,
015004. doi:10.1103/RevModPhys.94.015004 Gómez-Ruiz, F. J., Mayo, J. J., and del Campo, A. (2020). Full counting statistics of
topological defects after crossing a phase transition. Phys. Rev. Lett. 124, 240602.
Breuer, H. P., and Petruccione, F. (2007). The theory of open quantum systems.
doi:10.1103/PhysRevLett.124.240602
Oxford, United Kingdom: Oxford University Press. doi:10.1093/acprof:oso/
9780199213900.001.0001 Gong, M., Wen, X., Sun, G., Zhang, D. W., Lan, D., Zhou, Y., et al. (2016).
Simulating the Kibble-Zurek mechanism of the ising model with a superconducting
Cervera-Lierta, A. (2018). Exact Ising model simulation on a quantum computer.
qubit system. Sci. Rep. 6, 22667. doi:10.1038/srep22667
Quantum 2, 114. doi:10.22331/q-2018-12-21-114
Goo, J., Lim, Y., and Shin, Y. (2021). Defect saturation in a rapidly quenched Bose
Choo, K., von Keyserlingk, C. W., Regnault, N., and Neupert, T. (2018).
gas. Phys. Rev. Lett. 127, 115701. doi:10.1103/PhysRevLett.127.115701
Measurement of the entanglement spectrum of a symmetry-protected
topological state using the IBM quantum computer. Phys. Rev. Lett. 121, Higuera-Quintero, S., Gómez-Ruiz, F. J., Rodríguez, F., and Quiroga, L. (2022).
086808. doi:10.1103/physrevlett.121.086808 Repository for “experimental validation of the kibble-zurek mechanism on a digital
quantum computer”. This repository will be Available at: https://arxiv.org/abs/
Collura, M., and Karevski, D. (2010). Critical quench dynamics in confined
2208.01050 (Accessed Aug 23, 2022).
systems. Phys. Rev. Lett. 104, 200601. doi:10.1103/PhysRevLett.104.200601
IBM-Corporation (2022). Quantum computing IBM. Available at: https://
Cruz, D., Fournier, R., Gremion, F., Jeannerot, A., Komagata, K., Tosic, T., et al.
quantum-computing.ibm.com (Accessed 07 27, 2022).
(2019). Efficient quantum algorithms for GHZ and W states, and implementation
on the IBM quantum computer. Adv. Quantum Technol. 2, 1900015. doi:10.1002/ Ivakhnenko, O. V., Shevchenko, S. N., and Nori, F. (2022). Quantum control via
qute.201900015 landau-zener-stückelberg-majorana transitions. Available at: https://arxiv.org/abs/
2203.16348 (Accessed Oct 3, 2022). doi:10.48550/ARXIV.2203.16348
Cucchietti, F. M., Damski, B., Dziarmaga, J., and Zurek, W. H. (2007). Dynamics
of the bose-hubbard model: Transition from a mott insulator to a superfluid. Phys. Johansson, J., Nation, P., and Nori, F. (2013). QuTiP 2: A python framework for
Rev. A 75, 023603. doi:10.1103/PhysRevA.75.023603 the dynamics of open quantum systems. Comput. Phys. Commun. 184, 1234–1240.
doi:10.1016/j.cpc.2012.11.019
Cui, J. M., Gómez-Ruiz, F. J., Huang, Y. F., Li, C. F., Guo, G. C., and del Campo, A.
(2020). Experimentally testing quantum critical dynamics beyond the Kibble–Zurek Johansson, J., Nation, P., and Nori, F. (2012). QuTiP: An open-source python
mechanism. Commun. Phys. 3, 44. doi:10.1038/s42005-020-0306-6 framework for the dynamics of open quantum systems. Comput. Phys. Commun.
183, 1760–1772. doi:10.1016/j.cpc.2012.02.021
Cui, J. M., Huang, Y. F., Wang, Z., Cao, D. Y., Wang, J., Lv, W. M., et al. (2016).
Experimental trapped-ion quantum simulation of the Kibble-Zurek dynamics in Keesling, A., Omran, A., Levine, H., Bernien, H., Pichler, H., Choi, S., et al. (2019).
momentum space. Sci. Rep. 6, 33381. doi:10.1038/srep33381 Quantum Kibble-Zurek mechanism and critical dynamics on a programmable
Rydberg simulator. Nature 568, 207–211. doi:10.1038/s41586-019-1070-1
da Silva, M. P., Landon-Cardinal, O., and Poulin, D. (2011). Practical
characterization of quantum devices without tomography. Phys. Rev. Lett. 107, Kibble, T. W. B. (1980). Some implications of a cosmological phase transition.
210404. doi:10.1103/PhysRevLett.107.210404 Phys. Rep. 67, 183–199. doi:10.1016/0370-1573(80)90091-5
Damski, B. (2005). The simplest quantum model supporting the Kibble-Zurek Kibble, T. W. B. (1976). Topology of cosmic domains and strings. J. Phys. A Math.
mechanism of topological defect production: Landau-Zener transitions from a new Gen. 9, 1387–1398. doi:10.1088/0305-4470/9/8/029
perspective. Phys. Rev. Lett. 95, 035701. doi:10.1103/PhysRevLett.95.035701
King, A. D., Suzuki, S., Raymond, J., Zucca, A., Lanting, T., Altomare, F., et al.
Damski, B., and Zurek, W. H. (2006). Adiabatic-impulse approximation for (2022). Coherent quantum annealing in a programmable 2000-qubit ising chain.
avoided level crossings: From phase-transition dynamics to Landau-Zener Nat. Phys.. doi:10.1038/s41567-022-01741-6

Frontiers in Quantum Science and Technology 10 frontiersin.org

Higuera-Quintero et al. 10.3389/frqst.2022.1026025

Koh, J. M., Tai, T., Phee, Y. H., Ng, W. E., and Lee, C. H. (2022). Stabilizing quantum computers. Phys. Rev. Res. 4, 013196. doi:10.1103/
multiple topological fermions on a quantum computer. npj Quantum Inf. 8, 16. physrevresearch.4.013196
doi:10.1038/s41534-022-00527-1
Sen, D., Sengupta, K., and Mondal, S. (2008). Defect production in nonlinear
Landau, L. (1932a). Zur theorie der energieubertragung. Phyz. Z. Sowjetunion quench across a quantum critical point. Phys. Rev. Lett. 101, 016806. doi:10.1103/
1, 88. PhysRevLett.101.016806
Landau, L. (1932b). Zur theorie der energieubertragung II. Phyz. Z. Sowjetunion Solfanelli, A., Santini, A., and Campisi, M. (2021). Experimental verification of
2, 46. fluctuation relations with a quantum computer. PRX Quantum 2, 030353. doi:10.
1103/PRXQuantum.2.030353
Majorana, E. (1932). Atomi orientati in campo magnetico variabile. Nuovo Cim.
9, 43–50. 1924-1942. doi:10.1007/BF02960953 Stückelberg, E. C. G. (1932). Theory of inelastic collisions between atoms. Helv.
Phys. Acta 5, 369.
Marquardt, F., and Püttmann, A. (2008). Introduction to dissipation and
decoherence in quantum systems. Available at: https://arxiv.org/abs/0809.4403. Wang, L., Zhou, C., Tu, T., Jiang, H. W., Guo, G. P., and Guo, G. C. (2014).
(Accessed 25 Sep 2008). Quantum simulation of the Kibble-Zurek mechanism using a semiconductor
electron charge qubit. Phys. Rev. A 89, 022337. doi:10.1103/PhysRevA.89.022337
Mooney, G. J., Hill, C. D., and Hollenberg, L. C. L. (2019). Entanglement in a 20-
qubit superconducting quantum computer. Sci. Rep. 9, 13465. doi:10.1038/s41598- Wang, Y., Li, Y., qi Yin, Z., and Zeng, B. (2018). 16-qubit IBM universal quantum
019-49805-7 computer can be fully entangled. npj Quantum Inf. 4, 46. doi:10.1038/s41534-018-
0095-x
Mooney, G. J., White, G. A. L., Hill, C. D., and Hollenberg, L. C. L. (2021). Whole-
device entanglement in a 65-qubit superconducting quantum computer. Adv. Xu, X. Y., Han, Y. J., Sun, K., Xu, J. S., Tang, J. S., Li, C. F., et al. (2014). Quantum
Quantum Technol. 4, 2100061. doi:10.1002/qute.202100061 simulation of Landau-Zener model dynamics supporting the Kibble-Zurek
mechanism. Phys. Rev. Lett. 112, 035701. doi:10.1103/PhysRevLett.112.035701
Pozzobom, M. B., and Maziero, J. (2019). Preparing tunable Bell-diagonal states
on a quantum computer. Quantum Inf. process. 18, 142. doi:10.1007/s11128-019- Zener, C., and Fowler, R. H. (1932). Non-adiabatic crossing of energy levels. Proc.
2264-z R. Soc. Lond. Ser. A, Contain. Pap. a Math. Phys. Character 137, 696–702. doi:10.
1098/rspa.1932.0165
Preskill, J. (2018). Quantum computing in the NISQ era and beyond. Quantum 2,
79. doi:10.22331/q-2018-08-06-79 Zhong, H. S., Wang, H., Deng, Y. H., Chen, M. C., Peng, L. C., Luo, Y. H., et al.
(2020). Quantum computational advantage using photons. Science 370, 1460–1463.
Proctor, T. J., Carignan-Dugas, A., Rudinger, K., Nielsen, E., Blume-Kohout, R.,
doi:10.1126/science.abe8770
and Young, K. (2019). Direct randomized benchmarking for multiqubit devices.
Phys. Rev. Lett. 123, 030503. doi:10.1103/PhysRevLett.123.030503 Zhukov, A. A., Remizov, S. V., Pogosov, W. V., and Lozovik, Y. E. (2018).
Algorithmic simulation of far-from-equilibrium dynamics using quantum
Puebla, R., Marty, O., and Plenio, M. B. (2019). Quantum Kibble-Zurek physics in
computer. Quantum Inf. process. 17, 223. doi:10.1007/s11128-018-2002-y
long-range transverse-field Ising models. Phys. Rev. A 100, 032115. doi:10.1103/
PhysRevA.100.032115 Zurek, W. H. (1993). Cosmological experiments in condensed matter systems.
Phys. Rep. 276, 177–221. doi:10.1016/s0370-1573(96)00009-9
Puebla, R., Smirne, A., Huelga, S. F., and Plenio, M. B. (2020). Universal anti-
Kibble-Zurek scaling in fully connected systems. Phys. Rev. Lett. 124, 230602. Zurek, W. H. (1985). Cosmological experiments in superfluid helium? Nature
doi:10.1103/PhysRevLett.124.230602 317, 505–508. doi:10.1038/317505a0
Rodriguez-Vega, M., Carlander, E., Bahri, A., Lin, Z. X., Sinitsyn, N. A., and Zurek, W. H., Dorner, U., and Zoller, P. (2005). Dynamics of a quantum phase
Fiete, G. A. (2022). Real-time simulation of light-driven spin chains on transition. Phys. Rev. Lett. 95, 105701. doi:10.1103/PhysRevLett.95.105701

Frontiers in Quantum Science and Technology 11 frontiersin.org