Harvesting entanglement from the gravitational vacuum T. Rick Perche1 2 3 Boris Ragula1and Eduardo Mart ın-Mart ınez1 2 3 1Department of Applied Mathematics University of Waterloo Waterloo Ontario N2L 3G1 Canada

2025-05-06 0 0 1.37MB 53 页 10玖币
侵权投诉
Harvesting entanglement from the gravitational vacuum
T. Rick Perche,1, 2, 3, Boris Ragula,1, and Eduardo Mart´ın-Mart´ınez1, 2, 3,
1Department of Applied Mathematics, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada
2Perimeter Institute for Theoretical Physics, Waterloo, Ontario, N2L 2Y5, Canada
3Institute for Quantum Computing, University of Waterloo, Waterloo, Ontario, N2L 3G1, Canada
We study how quantum systems can harvest entanglement from the quantum degrees of free-
dom of the gravitational field. Concretely, we describe in detail the interaction of non-relativistic
quantum systems with linearized quantum gravity, and explore how two spacelike separated probes
can harvest entanglement from the gravitational field in this context. We provide estimates for the
harvested entanglement for realistic probes which can be experimentally relevant in the future, since
entanglement harvesting experiments can provide evidence for the existence of quantum degrees of
freedom of gravity.
I. INTRODUCTION
Arguably, the most important unsolved question in
theoretical physics is how to give a description for the
gravitational interaction that is consistent with our un-
derstanding of quantum matter. It is well known that the
coupling of classical gravity and quantum matter is theo-
retically inconsistent [1, 2], and as such we either need a
quantum description for gravity or a complete reformula-
tion of quantum theory. However, as of today, there is no
experimental confirmation of quantum behaviour of grav-
ity. Promising experimental setups, such as gravity me-
diated entanglement experiments, have been proposed to
attempt to verify quantum properties of the gravitational
field [3–5]. Despite its promise, there is plenty of debate
regarding which quantum properties of the gravitational
interaction can be confirmed by such experiments [6–12].
The core of this debate lies on how to identify genuinely
quantum degrees of freedom for the gravitational field.
One of the most remarkable differences between theo-
ries for classical and quantum fields is their lowest energy
state. While in a classical field theory, the ground state
corresponds to a zero-valued field, the vacuum state of
a quantum field theory is, arguably, not truly ‘empty’.
This gives rise to non-trivial statistics for local measure-
ments (see e.g. [13–16]). Moreover, it is well known that
the vacuum state of a quantum field contains quantum
correlations between different spacetime regions. This is
true even if these regions are spacelike separated [17, 18].
This fact is a fundamental feature of quantum field the-
ory in both flat and curved spacetimes [19, 20], and is
instrumental to our understanding of phenomena such
as the renormalization of the stress-energy tensor [20],
area laws in quantum field theories [21–25] and black
hole evaporation [26–31].
This vacuum entanglement can actually be detected:
localized probes can become entangled with each other
trickperche@perimeterinstitute.ca
bragula@uwaterloo.ca
emartinmartinez@uwaterloo.ca
through the interaction with the field, even when they are
spacelike separated through their interaction. This is the
idea behind the protocol of entanglement harvesting [32–
34]. In recent years, the protocol has been extensively
studied in many different scenarios [35–45], when probes
are coupled to different field operators [46, 47] and in
different spacetimes [48–53].
Entanglement harvesting from spacelike separated re-
gions is only possible from a field with quantum degrees
of freedom: a classical field cannot contain entanglement
that can be extracted. This fact can be used to decide
whether a field is classical or quantum. In fact, it has
been argued that an entanglement harvesting protocol
for the gravitational field can be used to witness quan-
tum gravity (see, e.g., [12, 54]). The main goal of this
manuscript is to perform a detailed study of this setup,
and to quantify the theoretical amount of entanglement
that could be extracted from a weak gravitational quan-
tum field.
Previous studies of entanglement harvesting which
take gravity degrees of freedom into consideration typi-
cally only couple to a scalar quantum field [55, 56]. That
is, the effect of gravity in the protocol is indirect, so that
the detectors are still coupled to the scalar field in a clas-
sical background spacetime. However, to the authors’
knowledge no previous work has considered entanglement
harvesting directly from a quantum gravitational field.
This manuscript is organized as follows. In Section II
we review the protocol of entanglement harvesting us-
ing two spacelike separated probes. In Section III we
review the formalism of linearized quantum gravity, and
describe how non-relativistic quantum systems couple to
a weak gravitational field. The protocol of entanglement
harvesting from the gravitational field is described in Sec-
tion IV. We present our first examples of entanglement
harvesting from the gravitational field in Section V. In
Section VI we compare the results found for the grav-
itational field with scalar model analogues. In Section
VII we study atoms coupled to quantum gravity in the
linearized regime. The conclusions of our work can be
found in Section VIII.
arXiv:2210.14921v2 [quant-ph] 18 Nov 2023
2
II. EXTRACTING ENTANGLEMENT FROM
QUANTUM FIELDS: THE ENTANGLEMENT
HARVESTING PROTOCOL
The goal of this section is to provide a brief review
of the protocol of entanglement harvesting from a scalar
field, so that we can later present a model for entangle-
ment harvesting from the gravitational field. In Subsec-
tion II A we review the formalism for an Unruh-DeWitt
(UDW) particle detector interacting with a real scalar
quantum field. In Subsection II B, we review the proto-
col of entanglement harvesting from a scalar field.
A. Particle detector models
In this section we review the UDW detector model [57–
59], where a localized two-level quantum system inter-
acts with a relativistic quantum field. This model has
been extensively used in the literature in order to probe
many features of different quantum field theories. Among
its applications are the ability to probe the Unruh ef-
fect [57, 60, 61] and Hawking radiation [57, 62], probe the
spacetime topology and geometry [53, 63, 64], describe
communication protocols in quantum field theory [65–
68], and, most relevant for this work, harvest entangle-
ment from a quantum field [35–56].
We consider a massless scalar quantum field in
(3 + 1)-dimensional Minkowski spacetime. We may ex-
press the field in terms of a plane-wave mode expansion
as
ˆ
ϕ(x) = 1
(2π)3
2Zd3k
p2|k|ˆakeik·x+ ˆa
keik·x,(1)
where we use inertial coordinates xµ= (t, x), so that
kµ= (|k|,k). The operators ˆa
k, ˆakrepresent creation
and annihilation operators, respectively, for a field mode
of momentum k. The creation and annihilation operators
satisfy the canonical commutation relations
hˆak,ˆa
ki=δ(3)(kk).(2)
In order to define the Hilbert space associated with the
quantum field, we define the vacuum state by ˆak|0= 0
for all k. The Hilbert space is then built from repeated
applications of the creation operators ˆa
kon |0, and is
usually referred to as the Fock space.
As a first approach, we will consider our detector to
be modelled by a two-level system. This qubit detector
follows an inertial trajectory, zµ(t) = (t, x0), along which
it can interact with the quantum field. Moreover, the free
Hamiltonian of the detector is given by
ˆ
HD= Ωˆσ+ˆσ,(3)
where Ω is the (proper) energy gap between the two lev-
els. We will denote the ground and excited states of the
detector by |gand |e, respectively. Then, the ladder op-
erators for our detector are ˆσ+=|eg|, and ˆσ=|ge|.
The associated interaction between the detector and the
field is prescribed by the following Hamiltonian in the
interaction picture
ˆ
HI(t) = λχ(t)ˆµ(t)ˆ
ϕ(t, x0),(4)
where λis the coupling strength, χ(t) is the switching
function, which determines the duration of the interac-
tion, and ˆµ(t) is the time evolved monopole moment of
the detector, explicitly given by
ˆµ(t) = eiΩtˆσ++eiΩtˆσ.(5)
We can relax the idealization of the detector being a
pointlike system. To do so, we implement a smearing
function such that the localization of the detector is de-
fined by a function f(x) centered at x0. In most physical
setups, the smearing function f(x) is given in terms of
the wavefunctions of the ground and excited state of the
system (see, e.g., [38, 69]). In fact, if the system has a
position degree of freedom ˆ
xand a canonically conjugate
momentum operator ˆ
p, one can write the interaction with
the field as [70]
ˆ
HI(t) = λχ(t)ˆ
ϕ(t, ˆ
x).(6)
Assuming the energy levels of the free Hamiltonian of the
detector to be discrete, the interaction Hamiltonian can
then be expanded in terms of the system’s wavefunctions
as [38, 69]
ˆ
HI(t) = λχ(t)
Zd3xˆ
ϕ(t, x)|xx|t(7)
=λX
n,m
χ(t)
Zd3xψ
n(x)ψm(x)eiΩnmtˆ
ϕ(t, x)|nm|,
where |xx|tdenotes the operator |xx|in the interac-
tion picture, ψn(x) = x|nis the eigenfunction associ-
ated with the energy eigenvalue Enand Ωnm =EnEm.
Restricting the interaction to two levels, the wave-
functions of the ground and excited states are simply
ψg(x) = x|gand ψe(x) = x|e.
We then notice that the diagonal terms of ˆ
HI(t) com-
mute with the detector’s free Hamiltonian. This means
that these terms would not be able to produce detector
excitations at leading order in λ, and would simply shift
the energy level of the system. That is, these terms have
no effect on the detector dynamics at leading order, so
that they can be neglected. The interaction Hamiltonian
in the interaction picture then reads
ˆ
H(t) =λχ(t)
Zd3xf(x)eiΩtˆσ++f(x)eiΩtˆσˆ
ϕ(x),
(8)
where f(x) = ψe(x)ψ
g(x), and Ω = EeEgis the energy
gap of the detector.
3
In order to write the time evolution of our system in
terms of an integral in spacetime, we define the interac-
tion Hamiltonian density as
ˆ
H(x) = λχ(t)f(x)eiΩtˆσ++f(x)eiΩtˆσˆ
ϕ(x).(9)
The relationship between the interaction Hamiltonian
and the interaction Hamiltonian density in Minkowski
spacetime is explicitly given by
ˆ
H(t) = Zd3xˆ
H(x).(10)
To study the interaction between the quantum field
and the detector, we will consider the detector to start
in the ground state and the quantum field to start in its
vacuum state. That is, we may write the initial density
operator of the detector-field system as
ˆρ0=|gg|⊗|00|.(11)
To obtain the final state of the detector, we first time-
evolve the density operator ˆρ0. The time evolution opera-
tor in the interaction picture is given by the time ordered
exponential1
ˆ
UI=Texp iZd4xˆ
HI(x),(12)
where d4x= dtd3x. Working perturbatively to second
order in λ, we may write
ˆ
UI=1+U(1)
I+U(2)
I+O(λ3),(13)
where, in terms of the Hamiltonian density, we have
ˆ
U(0)
I=1,(14)
ˆ
U(1)
I=iλZd4xˆ
HI(x),(15)
ˆ
U(2)
I=λ2Zd4xd4xˆ
HI(x)ˆ
HI(x)θ(tt),(16)
and θ(t) denotes the Heaviside theta function and im-
plements the time ordering operation. Using the density
operator in Eq. (11), the time evolved density operator
may be written as a power expansion in λ,
ˆρ= ˆρ(0) + ˆρ(1) + ˆρ(2) +O(λ3),(17)
where
ˆρ(0) = ˆρ0,(18)
ˆρ(1) =ˆ
U(1)
Iˆρ0+ ˆρ0ˆ
U(1)
I,(19)
ˆρ(2) =ˆ
U(2)
Iˆρ0+ˆ
U(1)
Iˆρ0ˆ
U(1)
I+ ˆρ0ˆ
U(2)
I.(20)
1Notice that for smeared detectors the notion of time-ordering can
be ambiguous but we are going to work in regimes where this is
not an issue. See [71] for details.
When a quantum field is measured, one does not have di-
rect access to the state of the field (only to the detector’s
degrees of freedom). To find the detector’s state after
the interaction, we compute the partial trace of ˆρwith
respect to the field’s Hilbert space. It is then possible
to find an expression (up to second order in λ) for the
probability of the detector transitioning from the ground
to the excited state:
L=λ2
Zd4xd4xχ(t)χ(t)f(x)f
(x)eiΩ(tt)
ˆ
ϕ(x)ˆ
ϕ(x)0,
(21)
where
ˆ
ϕ(x)ˆ
ϕ(x)0=1
(2π)3Zd3k
2|k|eik·(xx)(22)
is the two-point Wightman function of the quantum field
in the vacuum state. Given the two-point correlator in
Eq. (22), and defining the Fourier transforms of f(x)
and χ(t) as
˜
f(k) = Zd3xf(x)eik·x,(23)
˜χ(ω) = Zdt χ(t)eiωt,(24)
we can write the excitation probability, L, as an integral
over momenta
L=λ2
(2π)3Zd3k
2|k||˜χ(Ω + |k|)|2|˜
f(k)|2.(25)
B. Entanglement harvesting protocol
In this subsection, we will review the protocol of entan-
glement harvesting using UDW detectors. Understand-
ing entanglement in QFT is a nontrivial task, given that
most of our notions of entanglement depend on a tensor
product decomposition of the Hilbert space where the
theory is defined. However, in QFT, the Hilbert space of
the theory cannot easily be expressed as a tensor prod-
uct decomposition associated to localized regions of space
(see, e.g., [72] for the massive field case). This renders
most of our methods for discussing entanglement in non-
relativistic quantum mechanics unsuitable for localized
states in QFT [25]. However, one can still use other
methods in order to quantify the field’s quantum corre-
lations. For instance, one can quantify the entanglement
between probes that couple locally to the quantum field,
and use these to infer the entanglement properties of the
field. This phenomenology is generally called entangle-
ment harvesting.
In order to study the entanglement harvesting proto-
col, we consider two approximately spacelike separated
UDW detectors locally interacting with a quantum field.
The interaction regions chosen later in this paper are
approximately spacelike separated in the sense that the
4
effect of communication between the two regions is neg-
ligible for entanglement harvesting. See e.g., [73, 74], for
details. We consider the probes to be spacelike separated
throughout their interaction, so that they cannot com-
municate with each other. As a result, any entanglement
acquired by the probes must have come from the quan-
tum field itself. We will label our two detectors Aand
B, where the detectors are inertial and comoving, so that
their centres of mass undergo trajectories with constant
spatial coordinates, xa=x0, and xb=x0+L, with
Lbeing the spatial separation vector between the detec-
tors. Each detector will interact with the field according
to the interaction Hamiltonian density (9):
ˆ
HI,i(x) = λiχi(t)f
i(x)eiΩitˆσ+
i+fi(x)eiΩitˆσ
iˆ
ϕ(x),
(26)
for I∈ {A,B}. Here, we have that λiis the coupling
strength, fi(x) is the smearing function localized around
the trajectory, χi(t) is the switching function, Ωiis the
energy gap, ˆσ+
i=|eigi|, ˆσ
i=|giei|are the ladder
operators of detector I, and |ga,|ea,|gb,|ebdenote
the ground and excited states of the detectors.
We are interested in the case where both detectors are
initially in the ground state and the quantum field is
initially in its vacuum state. Thus, the resulting initial
density operator of the detectors-field system is given by
ˆρ0=|gaga|⊗|gbgb|⊗|00|.(27)
The final interaction Hamiltonian density will then be
the sum of the individual Hamiltonian densities
ˆ
HI(x) = ˆ
HI,a(x) + ˆ
HI,b(x).(28)
The unitary time evolution operator in the interaction
picture will be given by Eq. (12) with the interaction
Hamiltonian density of Eq. (28). After applying the uni-
tary time evolution operator and tracing out the quan-
tum field’s degrees of freedom, we obtain the following
density matrix of the two-detector system at order O(λ2)
in the basis {|gAgB,|gAeB,|eAgB,|eAeB⟩}
ˆρd=
1− Laa − Lbb 0 0 M
0Lbb Lba 0
0Lab Laa 0
M0 0 0
.(29)
Here, the Laa and Lbb terms are the excitation probabil-
ity of detector Aand B, respectively. The M,Lab and
Lba terms capture non-local correlations acquired by the
detectors. In the case of two qubit detectors coupled to a
scalar quantum field, we can find expressions for Lij and
M. Explicitly,
Lij =λiλjZd4xd4xˆ
ϕ(x)ˆ
ϕ(x)0ei(Ωitjt)
×χi(t)χj(t)fi(x)f
j(x),(30)
M=λaλbZd4xd4xˆ
ϕ(x)ˆ
ϕ(x)0θ(tt)
×ei(Ωata+Ωbt
b)χa(t)χb(t)fa(x)fb(x)
+ei(Ωat
a+ΩbtB)χa(t)χb(t)fa(x)fb(x).
(31)
We wish to quantify the entanglement between the
two detectors after they have both locally interacted
with the quantum field. There are multiple entangle-
ment measures that can be used for this purpose. In this
manuscript, we use the negativity, which is a faithful en-
tanglement measure for a system of two qubits [75, 76]
(see, e.g.,[53] for the specific discussion in the context of
entanglement harvesting). The negativity of ˆρdis defined
as N=Pλi<0λi, where the λi’s are the eigenvalues
of the partial transpose of ˆρd. To leading order in λ, the
negativity is then found to be
N= max 0,r|M|2(Laa − Lbb)2
4Laa +Lbb
2!.
(32)
In particular, in the special case where the excitation
probabilities of the detectors are the same, we have that
Laa =Lbb =L, and hence the negativity becomes
N= max(0,|M| − L).(33)
This happens, for instance, in the case of identical inertial
detectors in Minkowski spacetime. In fact, from now on,
we will assume that the detectors are identical, so that
λa=λb=λ, Ωa= Ωb= Ω, and that the interaction
happens simultaneously in their frames, which implies
χa(t) = χb(t) = χ(t), and the smearings are identical
modulo a spatial translation.
Given that the quantum field is assumed to be in its
vacuum state, we substitute Eq. (22) into Eqs. (30) and
(31), and we can write the transition probability and the
non-local term Min terms of the Fourier transform of
the smearing and switching functions as follows
Lij =λiλj
(2π)3Zd3k
2|k|˜χ(Ω + |k|)˜χ(Ω + |k|)˜
fi(k)˜
f
j(k),
(34)
M=λaλb
(2π)3
Zd3k
2|k|Q(|k|,Ω)( ˜
fa(k)˜
fb(k)+ ˜
fb(k)˜
fa(k)),
(35)
where
Q(|k|,Ω) = Zdtdtχ(t)χ(t)ei(Ω+|k|)tei(Ω−|k|)tθ(tt),
(36)
5
and where we have defined the Fourier transform of f(x)
and χ(t) as Eqs. (23) and (24).
The protocol of entanglement harvesting with two
UDW detectors coupled linearly to a scalar quantum
field has been extensively studied in the literature (see
e.g. [37, 39, 47, 51, 52, 56, 73, 77–79]), and many prop-
erties of the protocol are well understood by now.
It is important to mention that there are two ways in
which two probes can become entangled via an interac-
tion with a field [74]. The first way only relies on com-
munication via the field, and can be achieved through
an interaction of a quantum system with a classical field
(see, e.g., [12]). Two probes can only become entangled
in this way if they are in causal contact, and the role
played by the field in this case is a mere mediator that al-
lows for an exchange of information between the probes.
The second way in which detectors can become entan-
gled after an interaction with a field is by extracting (or
harvesting) entanglement previously present in the field.
This entangling procedure can only be performed when
the probes are coupled to a quantum field, which has
its own local quantum degrees of freedom. Moreover,
this protocol for extracting entanglement from a quan-
tum field can be achieved even if the detectors are not
within causal contact, and we will refer to this as entan-
glement harvesting to distinguish it from entanglement
generated by communication.
When one couples two causally connected probes to a
quantum field, these probes become entangled both via
communication and via extracting entanglement from the
field. This makes it a hard task to identify which part
of the entanglement acquired by the detectors was ex-
tracted from the field and which part is due to commu-
nication [74]. For this reason, here we will be concerned
with scenarios where the probes are mostly spacelike sep-
arated and we make sure that the entanglement acquired
by the detectors is primarily extracted from the field.
Experimentally, this kind of setup can be used to test
whether a field possesses quantum degrees of freedom:
only a quantum field allows spacelike separated locally
coupled detectors to become entangled.
III. LINEARIZED QUANTUM GRAVITY AND
MATTER: A GRAVITATIONAL DETECTOR
MODEL
The goal of this section is to summarize the description
of the interaction of a non-relativistic quantum system
with linearized quantum gravity. First, we review the
quantization of linearized perturbations of the metric in
Subsection III A. Then, in Subsection III B we describe
the coupling of a particle detector with the quantized
perturbations of the gravitational field.
A. Quantum fluctuations in a Minkowski
background
Einstein’s theory of general relativity describes space-
time as a four dimensional manifold with a Lorentzian
metric gµν that satisfies Einstein’s equations:
Gµν = 8π2
pTµν ,(37)
where Gµν is the Einstein tensor, Tµν is the stress-energy
momentum tensor and pis the Planck length (recall that
in natural units p=G).
Einstein’s equations are highly non-linear due to the
dependence of Gµν on the metric. In order to describe
small fluctuations of the gravitational field around a
given background, we can consider perturbations γµν of
the spacetime metric, so that the effective metric be-
comes gµν +γµν and Eq. (37) defines an equation of mo-
tion for γµν . Considering |γµν | ≪ 1 one can then expand
these equations to linear order in γµν to obtain a linear
equation of motion for the metric perturbations. This
approach is commonly called linearized gravity, and the
resulting theory for the field γµν is invariant under gauge
transformations of the form γµν 7−γµν +µξν+νξµ,
where ξµis the infinitesimal generator of such transfor-
mations.
In this manuscript we will be concerned with perturba-
tions around Minkowski spacetime, without the presence
of matter (Tµν = 0). In this case, the equations of motion
for γµν simplify, so that in inertial coordinates (t, x, y, z)
the first order variation of the Einstein tensor can be
written as
G(1)
µν (γ) (38)
=(µαγν)α1
2γµν 1
2µνγ1
2ηµν (αβγαβ γ),
where G(1)(γ) denotes the Einstein tensor to linear or-
der in γµν ,=ααis the D’Alembertian operator,
γ=ηµν γµν and ηµν is the Minkowski metric, which we
use to lower and raise indices. In the absence of matter,
the linearized Einstein’s equations for the gravitational
perturbations then read
G(1)
µν (γ)=0,(39)
which defines equations of motion for the linearized met-
ric perturbations whose solutions are typically called
gravitational waves. The theory for γµν can also be
thought of as a theory for a tensor field associated with
the action
S[γ] = 1
16π2
pZd4x(40)
1
2αγµν αγµν +1
2λγλγβγαγαβ +νγναµγµα.
The action above can be obtained by evaluating the met-
ric in the Einstein-Hilbert action at ηµν +γµν , and only
considering the leading order terms in γµν . In partic-
ular, extremizing this action with respect to γµν yields
摘要:

HarvestingentanglementfromthegravitationalvacuumT.RickPerche,1,2,3,∗BorisRagula,1,†andEduardoMart´ın-Mart´ınez1,2,3,‡1DepartmentofAppliedMathematics,UniversityofWaterloo,Waterloo,Ontario,N2L3G1,Canada2PerimeterInstituteforTheoreticalPhysics,Waterloo,Ontario,N2L2Y5,Canada3InstituteforQuantumComputing...

展开>> 收起<<
Harvesting entanglement from the gravitational vacuum T. Rick Perche1 2 3 Boris Ragula1and Eduardo Mart ın-Mart ınez1 2 3 1Department of Applied Mathematics University of Waterloo Waterloo Ontario N2L 3G1 Canada.pdf

共53页,预览5页

还剩页未读, 继续阅读

声明:本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知玖贝云文库,我们立即给予删除!
分类:图书资源 价格:10玖币 属性:53 页 大小:1.37MB 格式:PDF 时间:2025-05-06

开通VIP享超值会员特权

  • 多端同步记录
  • 高速下载文档
  • 免费文档工具
  • 分享文档赚钱
  • 每日登录抽奖
  • 优质衍生服务
/ 53
客服
关注