Quick charging of a quantum battery with superposed trajecotries

Po-Rong Lai, Jhen-Dong Lin, Yi-Te Huang, and Yueh-Nan Chen∗

Department of Physics and Center for Quantum Frontiers of Research &
Technology (QFort), National Cheng Kung University, Tainan 701, Taiwan

We propose novel charging protocols for quantum batteries based on quantum superpositions of
trajectories. Specifically, we consider that a qubit (the battery) interacts with multiple cavities or a
single cavity at various positions, where the cavities act as chargers. Further, we introduce a quan-
tum control prepared in a quantum superposition state, allowing the battery to be simultaneously
charged by multiple cavities or a single cavity with different entry positions. To assess the battery’s
performance, we evaluate the maximum extractable work, referred to as ergotropy. Our main result
is that the proposed protocols can utilize quantum interference effects to speed up the charging
process. For the protocol involving multiple cavities, we observe a substantial increase in ergotropy
as the number of superposed trajectories increases. In the case of the single-cavity protocol, we
arXiv:2307.09010v1 [quant-ph] 18 Jul 2023

show that two superposed trajectories (entry positions) are sufficient to achieve the upper limit of
the ergotropy throughout the entire charging process. Furthermore, we propose circuit models for
these charging protocols and conduct proof-of-principle demonstrations on IBMQ and IonQ quan-
tum processors. The results validate our theoretical predictions, demonstrating a clear enhancement
in ergotropy.

I. INTRODUCTION We consider a qubit acting as the quantum battery,

gaining energy through interactions with cavities func-
Quantum batteries (QBs) have emerged as a popular tioning as chargers. To assess the QB’s performance, we
research topic, providing valuable insights into how ther- focus on the maximum extractable work, known as er-
modynamics functions at the quantum scale [1, 2]. Re- gotropy [21]. The ergotropy is bounded by the stored
cent development has demonstrated that various quan- energy, i.e., the change in the qubit’s internal energy,
tum resources such as entanglement [3, 4] or coher- which can be regarded as a consequence of energy con-
ence [5, 6] can enhance the performance of quantum servation. To utilize superposed trajectories, we propose
bat- teries in terms of charging [7–14], storage [15–20] two charging protocols using an interferometric setup
and work extraction [21–29], etc. One of the intriguing akin to scenarios in Ref. [45]. The first one is called
phe- nomena used to achieve these enhancements is the multiple-charger protocol, which consists of multi-
the col- lective effects triggered by a group of QBs [8, 9, ple identical chargers (cavities), and the QB can interact
16, 30– 33]. In his seminal paper [34], Dicke with these chargers in a manner of quantum superposi-
characterized one of the collective effects, tion via a multi-port beam splitter. In principle, there
superradiance, by the quantum interference of are multiple output beams of QB when it exits the in-
emissions from an ensemble of atoms. Re- cent terferometer. This enable us to adjust the work extrac-
investigations have also demonstrated the utility of the tion strategy for each output and obtain an average er-
time-reversed phenomenon, known as superabsorp- gotropy, also known as the daemonic ergotropy [22]. The
tion [31, 32, 35, 36], on enhancing the capabilities of primary result of this protocol is an “activation” of the
QBs. In this work, our focus lies on an interferometric ergotropy. Specifically, we demonstrate that when the
ap- proach known as “superpositions of trajectories” battery is charged by a single cavity (without utilizing
[37–48]. This approach treats an atom’s space-time superposed trajectories), the ergotropy remains zero for
trajectories as a quantum system, enabling the a certain period, despite storing energy immediately
exploration of quantum interference of these after interaction with the cavity. Thus, there exists a
trajectories. A notable outcome of this approach is the finite de- lay before the battery can store “useful
effective noise mitigation in various quantum energy”, i.e., ex- tractable work. According to the
information tasks [41, 42, 44, 47, 48]. In our recent definition of ergotropy, population inversion, i.e., the
work [45], we have further advanced the under- excited state population of the qubit being larger than its
standing by interpreting this noise mitigation as a Zeno- ground state population, is required to obtain a
like state freezing phenomenon [49] within the nontrivial ergotropy. Therefore, one must wait until the
framework of open quantum systems [50]. battery reaches the inversion point in order to obtain
Additionally, we have demonstrated that this approach extractable work.
can also manifest Dicke- like collective effects even
Remarkably, we demonstrate that by considering this
when only one single atom is involved. Building upon
multiple-charger protocol, non-zero ergotropy can be ob-
these insights, the present work aims to delve into the
tained right after the charging process begins. This im-
potential of leveraging superposi- tions of trajectories to
plies that this protocol enables the achievement of
enhance the performance of QBs.
“quick charging” for the QB, where the ergotropy can be
acti- vated before reaching the inversion point.
Furthermore, we observe that the ergotropy increases as
∗ the number of
yuehnan@mail.ncku.edu.tw
 2

superposed trajectories grows. In the limit of an infinite ⇢1

number of superposed trajectories and considering the
rotating-wave approximation, the ergotropy even satu-
rates to its upper bound (i.e., the stored energy)
through- out the charging process. This saturation
indicates a complete conversion of stored energy into
extractable work. In addition, we reveal that the
enhancement of the average ergotropy stems from the
increased average pu- rity of the QB’s output states,
which can be regarded as a manifestation of the Zeno-
like phenomenon described in Ref. [45].
The second protocol is coined the single-charger pro-
tocol with only one cavity (charger) involved. In this
protocol, the QB can enter from different positions into
the cavity, experiencing different coupling strengths with
the cavity. By using the superposed trajectories, the QB
can enter these various positions simultaneously. As in-
dicated in Ref. [45], this particular setup can induce the
collective interference effect. It is important to note that
both the magnitude and relative phases of the coupling FIG. 1. The Quantum battery Q is first sent into a multi-
port beam splitter (MPBS1), which allows the quantum bat-
strengths between the QB and the charger can influence
tery to travel along N different trajectories (denoted by |j⟩D ,
the collective interference effect and, consequently, the
where j = 1 · · · N in the following) in a manner of quantum
ergotropy. We show that two superposed trajectories superposition. We consider two charging processes: (a) the
(positions) are sufficient to reach the upper limit of the trajectories each lead to a charger (cavities) {Cj }, causing
ergotropy, i.e., the stored energy, throughout the entire the quantum battery to interact with all chargers simultane-
charging process. This is achievable if the two coupling ously, (b) the trajectories lead to a single charger C but at
strengths possess the same magnitude and are completely different positions {rj }, causing the quantum battery to in-
out of phase. Our analytical analysis demonstrates that teract with the charger with various coupling strengths.
this phenomenon originates from the collective construc- Once the charging process is completed, a second multi-port
tive and destructive quantum interferences, which beam splitter (MPBS2) is used to perform measurement on
ensure that the output states remain pure throughout the tra- jectories degree of freedom D. This measurement
the charg- ing process. captures the quantum interference effect between different
Additionally, we present quantum circuits designed for trajectories and results in N possible reduced states ρj . We
the aforementioned charging protocols, requiring fewer then extract work from each ρj , where the maximum amount of
than twenty two-qubit gates. We implement and extractable work is called the ergotropy. The work extraction
execute these circuits on both IonQ quantum processors operations are described by unitary operators Uj , which
(based on trapped ions) and IBMQ quantum processors transforms each of the batteries to a passive state φj .
(based on superconducting circuits). The experimental
results obtained from these implementations further
validate the increase in ergotropy, which is consistent
with our theo- retical predictions.
The rest of the paper is organized as follows. In and gets charged when it passes through one of the
Sec. II, we characterize the multiple-charger protocol. In charg- ers. Suppose that the cavity length is l. Then, the
Sec. III, we further investigate the single-charger proto- inter- action time reads τ = l/v. To simplify our
col. In Sec. IV, we consider the circuit implementations discussions, we assume that the cavity is homogeneous
and present the experimental results of the devices from such that the interaction strength between the QB and
the charger re- mains constant during the charging
IBMQ and IonQ . Finally, we draw our conclusions in
process [51]. (iii) We characterize the trajectory degrees
Sec. V.
of freedom as an N dimensional qudit D, wherein we
associate N different trajectories inside the
interferometer with N basis states
II. MULTIPLE-CHARGER PROTOCOL
{|j⟩D}j=1···N . When the QB takes the path labeled by
j, it interacts with the charger Cj. In other words, D
We now formulate the multiple-charger scenario, which acts as a quantum control that determines which charger
can be described by an interferometric setup as shown in the QB interacts with. The total Hamiltonian involving
Fig. 1(a). The charging protocol consists of three differ- these three components can then be written as
ent components: (i) a qubit Q, which acts as the
quantum battery. (ii) N identical single-mode cavities
{Cj}j=1···N , which act as the chargers. The QB moves
N
at a speed v Σ
Htot = |j⟩ ⟨j|D ⊗ HQCj . (1)
j=1
 3

The Hamiltonian HQCj of the quantum battery Q and

the charger Cj can be expressed as Therefore, the (unnormalized) reduced state for the sys-
tem Q with the output k reads
h i
′
HQCj
= HQ + HCj + QCj , σk (τ) = TrCD Pk |ψ (τ)⟩ ⟨ψ (τ)|DQC Pk . (8)
¯h H
¯h
HQ = ω a σˆ z = Note that the probability of obtaining the outcome k
ωc σˆz ,2 2 (2) is pk(τ) = Tr [σk(τ )]. Thus, the normalized state con-
HCj = h¯ωc aˆ† aˆj , ditioned on the outcome k can be written as ρk(τ) =
′
HQCj σk(τ)/pk.
j Throughout this work, we choose
= h¯ωc λσˆx aˆ j +
N

aˆ † .
Here, aˆ j aˆ † annihilates (creates) a photon in Cj with
j 1 Σ
frequency ωc, ¯h ω a represents the energy splitting between |m⟩ ⟨n|D . (9)
the ground state |g⟩ and the excited state |e⟩ of Q. The |ξk=1⟩ ⟨ξk=1|D ≡
N
m ,
n=1
| ⟩e⟨ |−|
Pauli operators are therefore given by σˆ z = e g⟩ ⟨ g
| According to the assumption that all chargers are iden-
and σˆ x = |e⟩ ⟨g |+ g| ⟩e⟨ . Moreover, the dimensionless
tical as well as the orthonormality of the projectors, in
constant λ |represents the coupling strength between Q
and all the chargers. Throughout this work, we focus on Appendix A, we show that the explicit form of the rest
the resonant regime ωa = ωc. of the projectors is irrelevant, enabling us to further sim-
We first send the battery Q into the multiport beam plify the analysis.
splitter (MPBS1 in Fig. 1). In general, the beam splitter Here, we evaluate the performance of a QB by con-
can prepare the trajectories in a quantum superposition sidering the ergotropy, which quantifies the maximum
state so that Q can be charged by these N chargers extractable work. Given a charged state of the QB ρ(τ),
simul- taneously. For simplicity, we assume the the ergotropy is defined as
superposition state of the trajectories is W (ρ(τ)) ≡ Tr (ρ(τ)HQ) − min Tr Uρ(τ)U † H Q
N U (10)
Σ = Tr (ρ(τ)HQ) − Tr (φ(τ)HQ) ,
|ψ⟩D = √1 |j⟩D . (3)
N i=1 where U represents the unitary operation for work ex-
We consider that the battery and chargers are initialized traction. Also, φ is known as the passive state [21] (as-
in the ground state and the single-photon Fock state, sociated with ρ), which cannot provide useful work for
respectively. Therefore, the total initial state all possible work extraction operations U. According to
O reads Ref. [21], the passive state of the battery can be written
1 Σ as
N N
|ψ (0)⟩DQC = √ |j⟩D ⊗ |g⟩Q |1⟩Cj . (4)
N j=1 φ = s0 |e⟩ ⟨e| + s1 |g⟩ ⟨g| . (11)

j=1

After Q interacts with the chargers, according to Eq. (1),

the total states becomes Here, s0 and s1 denote the eigenvalues of ρ with s0 < s1.
Note that the upper limit of the ergotropy is set by the
1 Σ
N
stored energy quantified by the difference in the internal
|ψ (τ)⟩DQC = √ |j⟩D ⊗ |ϕj (τ)⟩QC , (5) energy of the QB before and after charging, namely
N j=1
N

where |ϕj (τ)⟩QC is defined as E (τ) ≡ Σ pk (τ) Tr [HQρk (τ)] − Tr (HQ |g⟩
⟨g|) (12)
k=1

, N  = Tr [HQρ(τ)] − Tr (HQ |g⟩ ⟨g|) ,

| Σ
|ϕj (τ)⟩QC = exp −i τ O |1⟩ .
C where ρ(τ) =
HQCj ¯h g⟩Q j=1 (6)
j
k

Finally, we make these trajectories interfere with pro- jectors {Pk}k acting on D, namely
one another by using another beam splitter (MPBS2
in Fig. 1). In principle, MPBS2 has N different outputs, Pk = |ξk⟩ ⟨ξk|D ,
which can be described using a set of orthonormal
Σ
 4
σk(τ). We can obtain W ≤ E, be-
cause Tr(HQ g g ) Tr(HQφ) in general. In addition, | ⟩ ⟨outputs
are N different | ≤ (labeled as k ). In principle,
the inequality saturates if and only if φ = g g , which | ⟩ ⟨ for each
one can find the optimal work extraction strategies
implies that ρ is a pure state. output and obtain the average ergotropy, i.e.,|
As aforementioned, in our charging protocol, there Σ
N W= pkW (ρk(τ)) { }
Pk = 11, (7) Σk
(13)
k=1
′ = pk [Tr (ρk(τ)HQ) − Tr (φk(τ)HQ)] ,
⟨ξk|ξk′ ⟩ = δk,k′ ∀ k, k . k
 5

where φk denotes the passive state associated with ρk. Here, for the case k = 1, the post-measurement state
Following similar reasoning as mentioned earlier, the av-
is passive during the time√period τ ∈ [0, TN ] with the inver-
erage ergotropy is also upper bonded by the stored en-
ergy, and the optimal extractable work can be obtained, sion time TN = tan−1( N)T/2π. Remarkably, for the
i.e., W (τ) = E(τ), if and only if pk, ρk forms a pure cases of k ̸ = 1, the post-measurement states are
{ } exactly the excited state, implying that the maximal
state decomposition of ρ.
extractable work h¯ ω c can be obtained. Therefore,
To gain some analytical insight, we now consider the the average er- gotropy can be expressed as
rotating-wave approximation, which is usually valid for
coupling strength λ ≤0.1 [52], so that the interaction W (τ)
Hamiltonian of the QB and the chargers in Eq. (2) can = W (σk=1(τ)) + (N − 1)W (σk̸=1(τ))

T
¯hω 2 sin2 (ωcλτ) − 1 if TN ≤ τ ≤ .
c
 ( 6
be reduced to the Jaynes-Cummings model, namely ¯h ω N−1 2
sin (ω λτ) if 0 ≤ τ ≤ T (20)
′ †
c N c N
H˜ =
j
σˆ + aˆ j + , (14) =
h¯ωcλ
σˆ− aˆj 4
QC
where σˆ + =| e⟩ ⟨g|and σˆ − =| g⟩ ⟨e represent the creation In Fig. 2, we present the time-dependent stored energy
|
and annihilation operators of Q, respectively. We can and the average ergotropy for different values of N. In
then evaluate Eq. (6) in this case: contrast to the case of N = 1, we observe non-zero av-
O erage ergotropy for the entire interval of interest
N
|ϕj (τ)⟩QC = − i sin (ωcλτ) |e⟩Q ⊗ aˆ j |1⟩C ′ because the states with k ̸= 1 are non-passive right after
the QB-
j′=1
j
chargers interaction is turned on (τ > 0). Therefore, the
(15) protocol can be used for “quick charging”, enabling im-
N
O mediate storage of useful work after the charging
+ cos (ωcλτ) |g⟩Q |1⟩Cj . process begins. Furthermore, the result indicates that
j=1
increasing N delays the inversion time TN and
enhances the av-
Let us start from the simplest case with only one charger erage ergotropy before TN . According to Eq. (20), this
(i.e., N = 1), where the reduced state of Q is expressed enhancement originates from the increase of the purity
as for the output k = 1 as its excited state population de-
ρ(τ) = sin2 (ωcλτ) |e⟩ ⟨e| + cos2 (ωcλτ) |g⟩ ⟨g| . (16) creases with N. This result aligns with the Zeno-like
state freezing effect described in Ref. [45]. In the asymp-
In this case, the stored energy is totic limit (N → ), we can further obtain a pure state
E(τ) = h¯ ω c sin2 (ωcλτ) , (17) decomposition,∞i.e., σk=1(τ) ∝ g | ⟩ ⟨g | σk̸=1(τ) ∝ e| ⟩ e⟨ ,
implying that the stored energy| can be fully converted
which oscillates with a period T = 2π/(ωcλ). We now into extractable work for the whole time interval, i.e,
focus on the time interval τ [0, T/4] ∈ (such that ωcλτ [0, E(τ) = W (τ).
π/2]), ∈where the stored energy monotonically increases
from 0 to its maximum value ¯hω c . Note that ρ is
diagonalized under basis e{|, g ⟩ |. ⟩Thus,
} according to Eq. 1 E(τ )
(10), the criterion for obtaining non-zero ergotropy is the W [N = 4]
moment that population inversion occurs, where the W [N = 2]
excited state population becomes larger than the ground W [N = 1]
state population (⟨ e| ρ(τ) | e⟩ > ⟨g| ρ(τ) |g⟩ ). The time-
dependence ( of the ergotropy can then be derived as
[h¯ωc ]
E, W

0 2 if 0 ≤ τ < T
8T
,
W (τ) = ¯hωc (2 T (18)
(ωcλτ) − 1) if 8 ≤τ ≤ 4 .
sin
One can observe that in the duration τ [0, T/8], al- 0
∈
though the stored energy E monotonically increases,
0 π/4 π/2
there is no extractable work, W = 0, for the battery
because ρ remains a passive state during this period. τ [ωc−1λ−1]
We now consider the scenario involving N chargers.
When the selective measurements satisfy Eq. (7) and FIG. 2. The stored energy E and average ergotropy W
Eq. (9), the unnormalized post-measurement states (both in units of ¯hω c ) on time τ (in units of 1/ωc) for λ =
can be written as (see Appendix A for detailed 0.05. The black solid curve plots the stored energy while the
derivations): dashed curves plot the average ergotropy. From bottom to
1 top, the blue, red and green dashed curves show the results
σk=1(τ) = c |e⟩ ⟨e| + cos (ω cλτ) |g⟩ ⟨g| ,
sin2 (ω λτ) 2
for N = 1,
N
1 N = 2 and N = 4, respectively.
σ (τ) = sin2 (ω λτ) |e⟩ ⟨e| .
k̸=1 c We now extend our scope of discussion into the
N (19) ultrastrong–coupling regime [52], where the rotating–
 7

0.8 (a) 0.8 (c)

E, W k /= 1
[h¯ωc]
E, W

[h¯ωc]
0.0 0.0
0 0.63 1.00 π/2 0 π/4 π/2
—1 —1
τ [ωc λ ] τ [ωc—1λ—1]
0.8 (b) (d)
1.0

P
E, Wk=1
[h¯ωc]

0.0 0.5
0 0.63 1.00 π/2 0 π/4 π/2
—1 —1
τ [ωc λ ] τ [ωc—1λ—1]
 8

E(ø) N=1 N=2 N=7 N=1

FIG. 3. (a) The stored energy E and average ergotropy W (both in units of ¯hω c ) as functions of time τ (in units of 1/ωc)
for λ = 0.5. (b) The stored energy E and the average ergotropy contributed by states of k = 1 Wk=1. Two brown vertical
lines at τ = 0.63, 1.00 indicate the inversion points TN for N = 1(blue) and N = 2(red). Here, Wk=1 is 0 when N =
7(green) and N = ∞(magenta). (c) The stored energy E and the average ergotropy contributed by states of k ̸= 1 Wk̸=1 on
time τ . Except for N = 1 (blue), these states contribute average ergotropy when τ > 0. (d) The dashed curves plot the
change in average purity P against time τ . The cutoff photon number is set to 9 in the above results.

wave approximation is no longer valid. In Fig. 3 (a), we contribution of k̸ = 1, since the corresponding average
present the dynamics of the average ergotropy and the er- gotropy becomes non-zero as soon as the charging
stored energy. We can still observe the quick charging process begins when N > 2. However, in contrast to the
effect, a delay of the inversion point, and an enhance- previous results with the rotating–wave approximation,
ment in average ergotropy as N increases. As shown the aver- age ergotropy cannot reach the upper bound
in Figs. 3 (b) and (c), we further present the individual even in the asymptotic limit, implying that the post-
contributions of the average ergotropy from the outputs measurement states are not pure.
k = 1 and k ̸= 1, which are respectively defined by In Fig. 3(d), we present the average purityP associated
W k=1 = pk=1W (ρk=1(τ)) with the post-measurement states, which is defined by

and " #
Σ σk 2
Σ (21) Tr[σk]Tr . (22)
W k̸=1 P= Tr[σ ]
= p W (ρ
k
(τ)).
k k k
k̸=1

We can observe that the inversion points come from the We can observe the overall average purity increasing as
contribution of k = 1. Further, we can observe that its N increases, thereby leading to the enhancement of
contribution decreases as N increases. When N ≥ 7, aver- age ergotropy. However, the average purity cannot
the contribution vanishes for the entire charging period. reach unity even in the asymptotic limit. Thus, the stored
In Appendix A, we provide analytical analysis, showing en- ergy cannot be fully converted into extractable work.
that the decrease in average ergotropy can also attribute In Figs. 4 (a) and (b), we compare the maximum average
to the Zeno-like state freezing effect. More specifically, ergotropy with the average purity with respect to the
we demonstrate that the excited state population cou- pling strength λ, where the maximum average
decreases when N increases. In the asymptotic limie ergotropy is defined by
(N → ∞),
the state can be frozen in the ground state. We further
W max = max W (τ). (23)
observe that the quick charging effect originates from the τ
∈[0,T/4]
 9

We can observe that as the coupling strength increases,

charger [53]. Therefore, the position-dependent QB-
the average purity drops, hence leading to a decrease in
charger Hamiltonian can be written as
the maximum average ergotropy.
Hj = HQ + HC + Hj′ , (24)
j

1.0 (a)
where HC = h¯ ω c aˆ † aˆ . Here, H′ represents the
interaction
Hamiltonian when QB is located at the position rj. Its
explicit form is given by
[N = 1]
j c rj x
[N = 10] H′ = h¯ ω λ cos πL σˆ aˆ + aˆ † , (25)
max
W

[N = 100] where L denotes the width of the charger and λ

[N = 1] repre- sents the maximal QB-charger coupling
strength, which can be achieved when rj = 0 or rj = L.
Thus, the total Hamiltonian for this scenario can be
0.5 expressed by
N
0.05 0.25 0.50 Htot =
Σ
|j⟩ ⟨j|D ⊗ Hj. (26)
∏ j=1

(b) Similar to the previous consideration, we prepare the

1.0 total system in the following initial state:

1 Σ
N
[N = 1] |j⟩D ⊗ |g⟩Q ⊗ |1⟩C , (27)
P

|ψ(0)⟩DQC = √
[N = 10] N j=1

[N = 100]
[N = 1]
 10
and allow it to evolve according to the total Hamiltonian
Htot, namely

0.6
1 Σ
N
0.05 0.25 0.50 |ψ(τ)⟩DQC = √ |j⟩D ⊗ |ϕj(τ)⟩QC , (28)
∏ N j=1

where |ϕj(τ)⟩QC, in this case, is defined as

FIG. 4. (a) Maximum average ergotropy W max defined in
Eq. (23) with respect to different coupling strength λ. The τ
blue, red, green and magenta dashed lines represent the re- i
|ϕj(τ)⟩QC = exp − H¯j h |g⟩Q ⊗ |1⟩C . (29)
sults for N = 1, N = 10, N = 100, and N = ∞, respectively. We also consider the projectors defined in Eq. (7) to char-
(b)The average purity P on dimensionless coupling acterize the measurements performed by using MPBS2,
strength λ. The blue, red, green and magenta dashed lines such that the corresponding post-measurement states
represent the results for N = 1, N = 10, N = 100, and N = reads
∞, re- spectively. The cutoff photon number is set to 9 in N
the above Σ
results.
Pk |ψ(τ)⟩ DQC = |ξk⟩D ⊗ k,j |ϕj(τ)⟩QC , (30)

c
j=1

III. SINGLE-CHARGER PROTOCOL where the coefficient ck,j is given by

1
In this section, we discuss the single-charger protocol ck,j = √
⟨ξk|j⟩. (31)
as shown in Fig. 1(b), where the quantum battery can N
enter and interact with a single cavity (charger) C at We now switch to the interaction picture, such that
different positions. The charging protocol resembles the the interaction Hamiltonian associated with the position
one in Sec. II, albeit with two minor adjustments. First, rj can be expressed as
we only use a single charger C instead of multiple ones. i
h¯
i
h¯
Second, the N dimensional qudit D now acts as a quan- Hj′ (τ) =e (HQ+H rj
C )τ
Hj′ e− (HQ+HC (32)
tum control that decides the QB’s position of entry into )τ
= cos π HI′ (τ),
C. More specifically, a path labeled by j guides the QB L
to a designated position denoted as rj inside the cavity. where HI (τ) is the position-independent part and reads
′

In this scenario, the QB experiences a varying cou- H′ (τ) = h¯ ω λe

i
(HQ+HC )τ
σˆ (aˆ +aˆ † )e −
i
(HQ+HC )τ
pling strength with respect to the position inside that . (33) c
h¯ h¯
I x
 11

We can now characterize the time evolution with the

propagator Uj,I (τ, 0) in terms of the Dyson series, namely We can observe that in the Dyson series, all the odd
∫ terms vanish, leading to the phenomenon of destructive
τ
ˆ −i ′ ′ ′
inter- ference. This effect originates from the complete
Uj,I (τ, 0) =T Hj (t out-of- phase nature of the coupling strengths for the
exp )dt two posi-
¯ 0 tions. As a direct consequence, the QB remains in the
∞ h ground state g throughout the entire process. Anal-
Σ 1 cos rj n ogously, the post-measurement state for k = 2 can be
π | ⟩
=
i −
n=0 n! ¯h L written as
∫τ ∫τ Σ
× dt1 · · · dtnTˆ H′I (t1) · · · H′ I (tn), 2
0 0 c2,j U (τ, 0) |g⟩Q ⊗ |1⟩C
(34) j,I
where Tˆ is the time-ordering operator with t1 > t2 > j=1

· · > tn. The post-measurement state of the QB and the 1

·
charger in Eq. (30) can then be expressed by = [U1,I (τ, 0) − U2,I (τ, 0)] |g⟩Q ⊗ |1⟩C
2∞ n
Σ Σ 1 i r1
N Σ
N = − π
cos

ck,j |ϕj(τ)⟩QC = ck,jUj,I (τ, 0) |g⟩Q ⊗ |1⟩C . (35) n! ¯h L

j=1 n odd
∫τ ∫ τ
ˆ
j=1 × ′ ′

Therefore, one can observe that the total evolution (in- dt1 · · · dtnT HI (t1) · · · HI (tn) |g⟩Q ⊗ |1⟩C .
cluding the qudit, the QB, and the charger) is described 0 0
(40)
by a linear combination of the position-dependent prop-
In this case, all the even terms in the Dyson series vanish,
agators, which characterizes the collective quantum in-
implying the QB is in the excited state |e⟩ for all τ > 0
terference effect among different positions [45].
(with zero probability of obtaining the output k = 2 at
We now show that two superposed trajectories (posi-
tions) can lead to the saturation of the ergotropy to its τ = 0). Because the post-measurement QB states for
upper bound with an appropriate adjustment of the col- these two outputs are pure states (i.e, the average pu-
lective interference effect. More specifically, we consider rity is one), we can conclude that the stored energy can
that the two positions satisfy r1 + r2 = L, such that be fully converted to the extractable work, i.e., E = W ,
throughout the whole charging process. Note that the
presented analysis does not rely on the rotating wave
r2 r1 r1 approximation, thus indicating the saturation of the er-
cos π L = cos π − π L = − cos π L . (36)
gotropy to its upper limit generally holds for all regimes
Therefore, the coupling strengths share the same magni- of the QB-charger coupling strength.
tude but are completely out of phase. Furthermore, we
consider
1
|ξk=1⟩ = √ (|1⟩D + | IV. IMPLEMENTATION ON
2⟩D), 2 QUANTUM DEVICES
(37)
1
|ξk=2⟩ = √ (|1⟩D − | In this section, we provide circuit models for the pro-
2⟩D), 2
and, according to Eq. (31), the coefficients in this case posed charging protocols and perform proof-of-concept
are experiments on the quantum processors provided by
IBMQ and IonQ, which involves two superposed trajec-
1 tories (N = 2).
c1,1 = = = 2,2 = . (38)
−c 2 The quantum circuit for the multiple-charging setup is
c1,2 c2,1
Therefore, the time evolution of the post-measurement described by Fig. 5 (a). The circuit consists of 4 qubits,
state of the QB and the charger for k = 1 is given by representing the control qubit D, the quantum battery
2 Q, and the two charging cavities C1 and C2, respec-
Σ
c tively. The circuit can be divided into three parts: state
1,j Uj,I (τ, 0) |g⟩Q ⊗ |1⟩C preparation, charging process, and measurements on the
j=1
control qubit and quantum battery. In the state prepa-
1 ration part, the qubits are prepared in the initial state
= [U1,I (τ, 0) + U2,I (τ, 0)] |g⟩Q ⊗ | specified in Eq. (4) using single-qubit gates. The charg-
1⟩2C
∞
Σ 1 ing process involves the utilization of two controlled-
cos r1 n
= − π unitary gates to simulate the simultaneous charging of
i
 12
n! ¯h L
n even the qubit by the two cavities through Jaynes-Cummings
∫ τ ∫ τ
interactions. In Fig. 5 (b), we present the decomposi-
× dt1 · · · dtnTˆH′I(t1) · · · H′ (t
I n) |g⟩ Q⊗ |1⟩ C .
0 0
tion of the controlled–unitaries into bit-flip (X) gates,
(39) controlled-z gates (CZ), and Ising coupling gates [XX(θ)
 13

(a) State (b)

Preparation Charging process(multiple-chargers) Measurement
D : |+ih+ X

D Q C1
C1 : |1ih1 T
exp[—iHQC1 ⌧/~] r Z
Q : |0ih0 C2
exp[—iHQC2 ⌧/~] (d)
C2 : |1ih1 T XX [✓(⌧)]
(c) r
H H
exp[—iHQC1 ⌧/~] XX [—✓(⌧)] YY [—✓(⌧)] XX [✓(⌧)] YY [✓(⌧)] H Rz [✓(⌧)] H

FIG. 5. (a) Quantum circuit for multiple-chargers protocol. Here, D, Q, C1, C2 represents the control qubit, battery qubit, first
charger and second charger, respectively. (b) Decomposition of a controlled unitary in (a). (c) Qubit configuration used on
ibmq algiers. (d) Decomposition of a XX gate into CNOT gates.

and Y Y (θ)], defined as follows: Charging process(single-charger)

X = σˆ x ,
D
CZ = |0⟩ ⟨0| ⊗ 11 + |1⟩ ⟨1| ⊗ σˆ z , Q
XX (θ) = cos(θ/2)11 ⊗ 11 − i sin(θ/2)σˆx ⊗ (41)
σˆ x , Y Y (θ) = cos(θ/2)11 ⊗ 11 − i C XX [2✓(⌧)] YY [2✓(⌧)]
sin(θ/2)σˆy ⊗ σˆ y .

Here, we map the charging time τ into the angle θ FIG. 6. Charging process of the single-charger setup, where
using the following relation the coupling strengths of the two superposed trajectories have
the same magnitude but are out of phase.
θ(τ) = ωcλτ/2. (42)

Finally, in the measurement part, we measure the con- average of 1000 experiment repetitions. The experimen-
trol qubit D in the x-direction, aligned with the projec- tal results demonstrate a notable increase in average er-
tors described by Eq. (37). Furthermore, as indicated gotropy, aligning with the theoretical predictions. Fur-
in Eq. (19), the QB’s post-measurement states are di- thermore, we can observe that the deviation between
agonalized under the energy eigenstates. Consequently, the experimental and theoretical results is correlated by
we can only measure Q in the z-direction to determine the circuit size, primarily determined by the number of
the stored energy as well as the ergotropy. We utilize two- qubit gates involved. Therefore, comparing the
the ibmq algiers and IonQ-Aria 1 devices. Note that the results from the IonQ and IBMQ devices, we find that the
qubit configuration for ibmq algiers is illustrated in Fig. 5 ex- perimental data from the IonQ device exhibit a closer
(c), while the qubits in IonQ-Aria 1 are fully connected. match to the theoretical curves compared to those from
Also, since the Ising coupling gates are not native gates the IBMQ device. In addition, the errors associated with
for the IBMQ device, we need to further decompose the single-charger protocol are smaller in magnitude
them into CNOT gates, which is shown in Fig. 5 (d). than those of the multiple-chargers protocol.
Conse- quently, the circuits for the IBMQ and IonQ
devices con- sist of 20 and 12 two-qubit gates,
respectively. Here, we use the Hadamard (H) gate and V. SUMMARY AND OUTLOOK
the rotation-z [Rz(θ)] gate, which are defined by
In this work, we utilize superposition of trajectories
H= to propose two charging protocols for quantum batteries
,
2 1 11 −11 (QBs), leading to improved storage of extractable work.
The first protocol, called the multiple-charger protocol,
(43)
Rz(θ) = exp(−iσˆ z θ/2). allows for simultaneous interaction between the QB and
multiple chargers. Leveraging the interference effect, we
For the single-charger setup, the circuit model for the observe an increase in extractable work, attributed to
charging process is presented in Fig. 6, which consists of the Zeno-like state freezing effect. The second protocol,
6 and 4 two-qubit gates in the circuits for the IBMQ and re- ferred to as the single-charger protocol, enables the
the IonQ devices, respectively. QB to interact with a single charger from multiple
Figures 7 and 8 illustrate the results obtained from positions concurrently. We demonstrate that the
the two protocols, with each data point representing the collective inter- ference effect generated by this setup
allows the QB to
 14

1 nite causal order [54–59], which allows for the control of

E [Sim.]
quantum operation ordering through a quantum switch.
E [IBMQ Exp.]
Based on this framework, we could consider the scenar-
E [IonQ Exp.]
ios, where the ordering of the charging process becomes
W [Sim.]
W [IBMQ Exp.]
indefinite [60]. This exploration could shed new light on
W [IonQ Exp.]
the potential benefits and implications of incorporating
[h¯ωc]

indefinite causal order into our proposed protocols, fur-

E, W

ther advancing the field of quantum battery charging.

0 ACKNOWLEDGEMENT
0 π/4 π/2
τ [ωc−1λ−1] We acknowledge the NTU-IBM Q Hub and the IBM
quantum experience for providing us a platform to im-
FIG. 7. The stored energy E and average ergotropy W
plement the experiment. This work is supported by the
(both in units of ¯hω c ) on time τ (in units of 1/ωc) National Center for Theoretical Sciences and National
for λ = 0.05, N = 2. The black solid and dashed curves rep- Science and Technology Council, Taiwan, Grant Nos.
resent the stored energy and average ergotorpy predicted
MOST 111-2123-M-006-001and NSTC 111-2627-M-006-
by numerical simulations. The green circles and blue “x”s 008.
repre- sent experimental results performed on ibmq algiers.
The red triangles and magenta diamonds represent
experimental re- sults performed on IonQ Aria 1. Each data Appendix A: Effects of the choice of projectors on
point is obtained after averaging 1000 experimental the post-measurement quantum battery states
repetitions.
Here, we prove that the choice of projectors for
ξ| k̸=1 ξ⟩ k̸⟨=1 D is irrelevant to the post-measurement
1 E, W [Sim.] states| σk̸=1(τ) as long as they are orthonormal to
E [IBMQ Exp.] |ξk=1 ⟩ ⟨ξk=1 D.
E [IonQ Exp.] Let| us start from Eq. (8), which can be expanded into
W [IBMQ Exp.]
W [IonQ Exp.] σk(τ) =
[h¯ωc]

1 N h i
E, W

Σ (A1)
N ⟨ξk|j⟩ ⟨f|ξk⟩ TrC
j,f =1 |ϕj(τ)⟩ ⟨ϕf (τ)|QC

00 We switch to the interaction picture, where the inter-

π/4 π/2 action Hamiltonian reads
′ i
(HQ+HC )τ ′ − i (HQ+HC )τ
HQC j (τ) = e H QCj h¯ j . (A2)
τ [ωc−1λ−1]
j

h¯

The time-dependent part of post-measurement states can

FIG. 8. The stored energy E and average ergotropy W be generally expressed by
(both in units of ¯hω c ) on time τ (in units of 1/ωc) with r1 = ∞
0.1l and r2 = 0.9l. The black curve represents the stored Σ
|ϕj (τ)⟩QC = αn(τ) |nj⟩ , (A3)
energy and average ergotropy predicted by numerical n=0
simulations. The green circle and blue “x”s represent the
experimental results performed on ibmq algiers. The red with
triangles and magenta N
diamonds represent experimental results performed on IonQ n √
j
Aria 1. Each data point is obtained after averaging 1000 |nj⟩ = |e⟩ ) aˆj O|1⟩Cm .
(aˆ†n! (A4)
(σˆx ) ⊗ n
experimental repetitions.
m=1
 15
achieve the maximum extractable work throughout the Note that in the following analysis, we show the explicit
entire charging period. Moreover, we investigate the cir- expression of the time-dependent coefficient αn(τ) does
cuit implementations utilizing IonQ and IBMQ devices, not affect the result. Thus, we will keep them unspecified.
Now, wehcan show that i
providing experimental data that further support the en-
hanced extractable work, thus validating our theoretical TrC |ϕj (τ)⟩ ⟨ϕf (τ)|QC
predictions. (Σ (A5)
∞ |α (τ)|2σn |e⟩ ⟨e| σn for j = f
As a possible future direction, we could extend our = n=0
2
n x x
charging protocols to a related framework called indefi- |α1(τ)| |g⟩ ⟨g| , for j ̸= f.
 16

Inserting this into Eq. (A1), we can obtain Recall that for the case of k = 1, we require the
Σ
Σ 1 projectors to be orthonormal to̸ that of k = 1, i.e.,
2
σ (τ) = ∞ | (τ)| σn |e⟩ ⟨e| σn |⟨ξ |j⟩ ⟨ξk̸=1|ξk=1⟩ = 0, implying that the ket |ξk̸=1⟩ can be
2 N
α |
k n x x k expressed by
n=0 j=f
N (A6) N
Σ Σ Σ
1 N N
+ |α (τ)|2|g⟩ ⟨g| ⟨ξ |j⟩ ⟨f|ξ ⟩ .
1 k |ξk̸=1⟩ = βk,j |j⟩D ,
N k j̸=f f =1 j=1
N (A9)
For the case of k = 1, according to Eq. (7), we can Σ
obtain with β k,j = 0 ∀ k ̸= 1.
j=1
1
⟨ξk=1|j⟩ = √ ∀ j = 1, · · · , N. (A7)
N
Therefore, the corresponding post-measurement state Therefore, the post-measurement states can be written
can be written as as
∞
∞ Σ
Σ 1 |αn(τ)| σ
2 x n|e⟩ ⟨e| σxn
σk=1(τ) = 1 σk̸=1(τ) =
2
|αn(τ)| σ xn|e⟩ ⟨e| σ xn
N N n=0 (A10)
n=0
N (A8) 1
−1 2
2

+ |α1 (τ)| |g⟩ ⟨g| . — |α1(τ)| |g⟩ ⟨g| .

N N
Note that in the asymptotic limit (N→ ), we observe This concludes the proof that the post-measurement
σk=1(τ) ∝g | g⟩ ⟨, indicating the Zeno-like
∞ state freezing states are independent of the explicit expression of the
effect [45].
| coefficients βk,j associated with the projectors.

[1] R. Uzdin, A. Levy, and R. Kosloff, Equivalence of quan- localized quantum batteries, Phys. Rev. B 100, 115142
tum heat machines, and quantum-thermodynamic signa- (2019).
tures, Phys. Rev. X 5, 031044 (2015). [12] A. Crescente, M. Carrega, M. Sasetti, and D. Ferraro,
[2] W. Niedenzu, V. Mukherjee, A. Ghosh, A. G. Kofman, Charging and energy fluctuations of a driven quantum
and G. Kurizki, Quantum engine efficiency bound battery, New J. Phys. 22, 063057 (2020).
beyond the second law of thermodynamics, Nat. [13] A. Crescente, M. Carrega, M. Sassetti, and D. Ferraro,
Commun. 9, 165 (2018). Ultrafast charging in a two-photon dicke quantum bat-
[3] R. Horodecki, P. Horodecki, M. Horodecki, and tery, Phys. Rev. B 102, 245407 (2020).
K. Horodecki, Quantum entanglement, Rev. Mod. Phys. [14] S. Ghosh, T. Chanda, S. Mal, and A. Sen(De), Fast
81, 865 (2009). charging of a quantum battery assisted by noise,
[4] D. Bruß, Characterizing entanglement, Journal of Math- Phys. Rev. A 104, 032207 (2021).
ematical Physics 43, 4237 (2002). [15] A. C. Santos, B. i. e. i. f. m. c. C¸ akmak, S. Campbell,
[5] A. Streltsov, G. Adesso, and M. B. Plenio, Colloquium: and
Quantum coherence as a resource, Rev. Mod. Phys. 89, N. T. Zinner, Stable adiabatic quantum batteries,
041003 (2017). Phys. Rev. E 100, 032107 (2019).
[6] O. Nairz, M. Arndt, and A. Zeilinger, Quantum interfer- [16] F. Pirmoradian and K. Mølmer, Aging of a quantum bat-
ence experiments with large molecules, American Journal tery, Phys. Rev. A 100, 043833 (2019).
of Physics 71, 319 (2003). [17] J. Liu, D. Segal, and G. Hanna, Loss-free excitonic
[7] F. C. Binder, S. Vinjanampathy, K. Modi, and J. Goold, quan- tum battery, J. Phys. Chem. C 123, 18303
Quantacell: powerful charging of quantum batteries, New (2019).
J. Phys. 17, 075015 (2015). [18] J. Q. Quach and W. J. Munro, Using dark states to
[8] F. Campaioli, F. A. Pollock, F. C. Binder, L. C´eleri, charge and stabilize open quantum batteries, Phys. Rev.
J. Goold, S. Vinjanampathy, and K. Modi, Enhancing the Applied 14, 024092 (2020).
charging power of quantum batteries, Phys. Rev. Lett. [19] S. Gherardini, F. Campaioli, F. Caruso, and F. C. Binder,
118, 150601 (2017). Stabilizing open quantum batteries by sequential mea-
[9] D. Ferraro, M. Campisi, G. M. Andolina, V. Pellegrini, surements, Phys. Rev. Res. 2, 013095 (2020).
and M. Polini, High-power collective charging of a solid- [20] D. Rosa, D. Rossini, G. M. Andolina, M. Polini, and
state quantum battery, Phys. Rev. Lett. 120, 117702 M. Carrega, Ultra-stable charging of fast-scrambling syk
(2018). quantum batteries, J. High Energy Phys. 2020 (11), 67.
[10] T. P. Le, J. Levinsen, K. Modi, M. M. Parish, and F. A. [21] R. Alicki and M. Fannes, Entanglement boost for ex-
Pollock, Spin-chain model of a many-body quantum bat- tractable work from ensembles of quantum batteries,
tery, Phys. Rev. A 97, 022106 (2018). Phys. Rev. E 87, 042123 (2013).
[11] D. Rossini, G. M. Andolina, and M. Polini, Many-body
 17
[22] G. Francica, J. Goold, F. Plastina, and M. Paternos-
tro, Daemonic ergotropy: enhanced work extraction from
quantum correlations, npj Quantum Inf. 3, 12 (2017).
 18

[23] G. Manzano, F. Plastina, and R. Zambrini, Optimal work

extraction and thermodynamics of quantum measure- Commun. 10, 1630 (2019).
ments and correlations, Phys. Rev. Lett. 121, 120602 [42] G. Rubino, L. A. Rozema, D. Ebler, H. Kristj´ansson,
(2018). S. Salek, P. Allard Gu´erin, A. A. Abbott, C. Branciard,
[24] G. M. Andolina, M. Keck, A. Mari, M. Campisi, V. Gio- i. c. v. Brukner, G. Chiribella, and P. Walther, Experi-
vannetti, and M. Polini, Extractable work, the role of mental quantum communication enhancement by super-
correlations, and asymptotic freedom in quantum bat- posing trajectories, Phys. Rev. Res. 3, 013093 (2021).
teries, Phys. Rev. Lett. 122, 047702 (2019). [43] J. Foo, S. Onoe, R. B. Mann, and M. Zych, Thermal-
[25] F. Barra, Dissipative charging of a quantum battery, ity, causality, and the quantum-controlled unruh–dewitt
Phys. Rev. Lett. 122, 210601 (2019). detector, Phys. Rev. Res. 3, 043056 (2021).
[26] G. Francica, F. C. Binder, G. Guarnieri, M. T. Mitchi- [44] F.-J. Chan, Y.-T. Huang, J.-D. Lin, H.-Y. Ku, J.-S.
son, J. Goold, and F. Plastina, Quantum coherence Chen, H.-B. Chen, and Y.-N. Chen, Maxwell’s two-
and ergotropy, Phys. Rev. Lett. 125, 180603 (2020). demon engine under pure dephasing noise, Phys.
[27] F. H. Kamin, F. T. Tabesh, S. Salimi, F. Kheirandish, Rev. A 106, 052201 (2022).
and A. C. Santos, Non-Markovian effects on charging [45] J.-D. Lin, C.-Y. Huang, N. Lambert, G.-Y. Chen,
and self-discharging process of quantum batteries, New F. Nori, and Y.-N. Chen, Space-time dual quantum Zeno
J. Phys. 22, 083007 (2020). effect: Interferometric engineering of open quantum sys-
[28] J. Monsel, M. Fellous-Asiani, B. Huard, and A. tem dynamics, Phys. Rev. Research 4, 033143 (2022).
Auff`eves, The energetic cost of work extraction, [46] J.-D. Lin and Y.-N. Chen, Boosting entanglement growth
Phys. Rev. Lett. 124, 130601 (2020). of many-body localization by superpositions of disorder,
[29] S. Ghosh, T. Chanda, and A. Sen(De), Enhancement 2304.08849 (2023).
in the performance of a quantum battery by ordered [47] H.-Y. Ku, K.-Y. Lee, P.-R. Lai, J.-D. Lin, and Y.-
and disordered interactions, Phys. Rev. A 101, 032115 N. Chen, Coherent activation of a steerability-breaking
(2020). channel, Phys. Rev. A 107, 042415 (2023).
[30] K. Mei and J. Fang, Superabsorption-a method to im- [48] K.-Y. Lee, J.-D. Lin, A. Miranowicz, F. Nori, H.-Y. Ku,
prove absorbing boundary conditions (electromagnetic and Y.-N. Chen, Steering-enhanced quantum metrology
waves), IEEE Transactions on Antennas and Propaga- using superpositions of noisy phase shifts, Phys. Rev.
tion 40, 1001 (1992). Res. 5, 013103 (2023).
[31] K. D. B. Higgins, S. C. Benjamin, T. M. Stace, G. J. Mil- [49] B. Misra and E. G. Sudarshan, The Zeno’s paradox in
burn, B. W. Lovett, and E. M. Gauger, quantum theory, J. Math. Phys. 18, 756 (1977).
Superabsorption of light via quantum engineering, Nat. [50] H.-P. Breuer and F. Petruccione, The theory of open
Commun. 5, 4705 (2014). quantum systems (Oxford University Press on Demand,
[32] J. Q. Quach, K. E. McGhee, L. Ganzer, D. M. Rouse, 2002).
B. W. Lovett, E. M. Gauger, J. Keeling, G. Cerullo, [51] D. Meschede, H. Walther, and G. Mu¨ller, One-atom
D. G. Lidzey, and T. Virgili, Superabsorption in an or- maser, Phys. Rev. Lett. 54, 551 (1985).
ganic microcavity: Toward a quantum battery, Sci. Adv. [52] A. F. Kockum, A. Miranowicz, S. D. Liberato, S.
8, eabk3160 (2022). Savasta, and F. Nori, Ultrastrong coupling between light
[33] J. Joshi and T. S. Mahesh, Experimental investigation and mat- ter, Nat. Rev. Phys 1, 19 (2019).
of a quantum battery using star-topology nmr spin [53] W. E. Shanks, D. L. Underwood, and A. A. Houck,
systems, Phys. Rev. A 106, 042601 (2022). A scanning transmon qubit for strong coupling circuit
[34] R. H. Dicke, Coherence in spontaneous radiation pro- quantum electrodynamics, Nat. Commun. 4, 1991 (2013).
cesses, Phys. Rev. 93, 99 (1954). [54] G. Rubino, L. A. Rozema, A. Feix, M. Arau´jo, J. M.
[35] D. Yang, S.-h. Oh, J. Han, G. Son, J. Kim, J. Kim, Zeuner, L. M. Procopio, Cˇ aslav Brukner, and P.
M. Lee, and K. An, Realization of superabsorption by Walther, Experimental verification of an indefinite
causal order,
time reversal of superradiance, Nat. Photon. 15, 272
Sci. Adv. 3, e1602589 (2017).
(2021).
[55] K. Goswami, C. Giarmatzi, M. Kewming, F. Costa,
[36] Y. Ueki, S. Kamimura, Y. Matsuzaki, K. Yoshida, and
C. Branciard, J. Romero, and A. G. White, Indefinite
Y. Tokura, Quantum battery based on superabsorption,
causal order in a quantum switch, Phys. Rev. Lett.
J. Phys. Soc. Jpn. 91, 124002 (2022).
121, 090503 (2018).
[37] G. Chiribella and H. Kristjansson, Quantum Shannon
[56] D. Ebler, S. Salek, and G. Chiribella, Enhanced com-
theory with superposition of trajectories, Proc. R. Soc.
munication with the assistance of indefinite causal
A 475, 20180903 (2019).
order, Phys. Rev. Lett. 120, 120502 (2018).
[38] J. Foo, S. Onoe, and M. Zych, Unruh-DeWitt detectors
[57] X. Zhao, Y. Yang, and G. Chiribella, Quantum
in quantum superpositions of trajectories, Phys. Rev. D
metrology with indefinite causal order, Phys. Rev. Lett.
102, 085013 (2020).
124, 190503 (2020).
[39] H. Kristj´ansson, G. Chiribella, S. Salek, D. Ebler, and
[58] N. Loizeau and A. Grinbaum, Channel capacity enhance-
M. Wilson, Resource theories of communication, New J.
ment with indefinite causal order, Phys. Rev. A 101,
Phys. 22 (2020).
012340 (2020).
[40] Q. Duprey and A. Matzkin, Proposal to observe path
[59] G. Chiribella, M. Banik, S. S. Bhattacharya, T. Guha,
superpositions in a double-slit setup, Phys. Rev. A 105,
M. Alimuddin, A. Roy, S. Saha, S. Agrawal, and G. Kar,
052231 (2022).
Indefinite causal order enables perfect quantum commu-
[41] F. Ghafari, N. Tischler, C. Di Franco, J. Thompson,
nication with zero capacity channels, New J. Phys. 23,
M. Gu, and G. J. Pryde, Interfering trajectories in exper-
033039 (2021).
imental quantum-enhanced stochastic simulation, Nat.
[60] K. Simonov, G. Francica, G. Guarnieri, and M. Pater-
nostro, Work extraction from coherently activated maps
 19

via quantum switch, Phys. Rev. A 105, 032217 (2022).