Quantum Est 2flav

International Journal of Quantum Information
Vol. 15, No. 6 (2017) 1750045 (15 pages)

#.c World Scienti¯c Publishing Company
DOI: 10.1142/S0219749917500459
by UNIVERSITY OF MILAN on 06/21/23. Re-use and distribution is strictly not permitted, except for Open Access articles.
Quantum estimation in neutrino oscillations
Edson C. Nogueira*,§, Gustavo de Souza†,¶,

Adalberto D. Varizi*,|| and Marcos D. Sampaio‡,**
*Universidade Federal de Minas Gerais,
Int. J. Quantum Inform. 2017.15. Downloaded from www.worldscientific.com
Departamento de F{sica, ICEX,

P. O. Box 702, 30.161-970,
Belo Horizonte/MG, Brazil
†
Universidade Federal de Ouro Preto,
Departamento de Matem a tica, ICEB,
Campus Morro do Cruzeiro, s/n,
35400-000, Ouro Preto/MG, Brazil
‡
Universidade Federal do ABC, Departamento de F{sica,
CCNH, Campus Santo Andre, Bloco B, 8 andar,
09210-170, Santo Andre/SP, Brazil
§
edsoncezar16@gmail.com
¶
gdesouza@iceb.ufop.br
||
adalbertovarizi@gmail.com
**msampaio@fisica.ufmg.br
Received 18 March 2017

Revised 26 June 2017
Accepted 10 August 2017
Published 14 September 2017
In this work, we analyze two-°avor neutrino oscillations within the framework of quantum es-
timation theory (QET). We compute the quantum Fischer information (QFI) for the mixing angle
and show that mass measurements are the ones that achieve optimal precision. We also study
the Fischer information (FI) associated with °avor measurements and show that they are opti-
mized at speci¯c neutrino times-of-°ight. Therefore, although the usual population measurement
does not realize the precision limit set by the QFI, it can in principle be implemented with the best
possible sensitivity to . We investigate how these quanti¯ers relate to the single-particle, mode
entanglement in neutrino oscillations. We demonstrate that this form of entanglement does not
enhance either of them. In particular, our results show that in single-particle settings, entan-
glement is not directly connected with the enhancement of precision in metrological tasks.
Keywords: Neutrino oscillations; QFI; entanglement.
PACS Number(s): 03.65.Ta, 03.65.Ud, 14.60.Pq
||
Corresponding author.
1750045-1
E. C. Nogueira et al.
1. Introduction
Neutrinos have been a unique source of fundamental discoveries in Physics since
their introduction by Pauli and the development of the Fermi theory of beta decay.1,2
In the Standard Model of elementary particles, they are massless,3 but the phe-
nomena of neutrino oscillations — proposed for the ¯rst time in 1957 by B. Ponte-
corvo4 and con¯rmed some decades later by the series of experiments that culminated
in Ref. 5 — show that they must have mass.
The theory of neutrino mixing was developed in analogy with that of quark
mixing. After the initial e®orts of Pontecorvo,4 Maki, Nakagawa and Sakata gave
relevant contributions6,7 and the theory was completed in the 70s by Bilenky and
others.1,8 The point is that the mismatch between °avor and mass eigenstates is
described by a mixing matrix U which contains some parameters — mixing angles,

neutrino masses and CP violating phases — that encompass all the information
about oscillation phenomena. Hence, a precise knowledge about the values of these
parameters is essential to the comprehension of neutrino mixing and future devel-
opments in this ¯eld.
However, these parameters do not directly correspond to a physical observable of
the neutrino system. Therefore, in order to measure them, one must infer their values
indirectly from the measurement of other observables. For example, in neutrino
oscillations experiments, population measurements are performed and the values of
the mixing angles are estimated based on their outcomes. Although some parameters
of the mixing matrix have been measured over the last years,9,10 enhancing the
precision on their values, i.e. reducing uncertainty intervals, remains a fundamental
task, and experiments are being designed to do this in the following few years.11
From these considerations, we see that it is pertinent to ask the following ques-
tions: what is the optimal precision we can achieve through population measurements
in neutrino oscillation experiments? What is the ultimate precision limit allowed
by quantum theory? Is it possible to determine the observable that we should
measure in order to obtain it? These questions are properly addressed in the
framework of quantum estimation theory (QET),12,13 where we can determine the
optimal precision that can possibly be achieved in unbiased estimation with any
(generalized) measurement scheme through the so-called quantum Fischer informa-
tion (QFI), and sometimes, also the measurement scheme that permits such precision
to be achieved. It necessarily corresponds from a mathematical viewpoint to the
positive operator valued measure (POVM) generated by projectors over the eigen-
spaces of a certain operator — the so-called symmetric logarithmic derivative
(SLD).14 However, direct measurement of these projectors may sometimes be
(technologically) unfeasible.
Another point that deserves attention is entanglement. Recent works (see Ref. 14
for a summary) have shown, for example, that the use of entangled probe states
in parameter estimation tasks make it frequently possible to reach the so-called
Heisenberg scaling, which means that the variance of the estimator of the desired
1750045-2
parameter scales as varðÞ N12 , N being the number of trials. This is better than the
usual shot-noise scaling varðÞ N1 : it gives smaller measurement uncertainty for
fewer trials. This type of result would be very interesting in contexts such as the
present one of neutrino oscillations, where the trial events, i.e. neutrino detections,
are di±cult.
More importantly, these papers show that the use of entangled probe states
implies on the enhancement of the maximum precision limit achievable, optimizing
the value of the QFI. Thus, entanglement can be explored as a resource leading to
higher performance in parameter estimation tasks. However, these results were
obtained in the context of multiparticle (usually two-particle) probe systems. This
makes it hard to apply them to neutrino studies: since neutrinos hardly interact with
each other in oscillations experiments, usable entangled neutrino pairs would be very
hard to obtain in this context.
In the present work, we investigate the role of the single-particle, mode entan-
glement in neutrino oscillations of Blasone et al.15 in the estimation of the two-°avor
mixing angle. Due to the \hardly interacting" character of neutrinos mentioned
above, this is the natural form of entanglement to consider for them (also referred to
in the literature as occupation number entanglement16). To begin with, we consider
the two-°avor neutrino oscillations within the framework of QET and analyze
the estimation of the mixing angle which characterizes them. We will take
neutrinos produced in a de¯nite °avor and two types of model for the ensuing
neutrino oscillations:
(1) the standard plane-wave model;

(2) a \decoherence model" which takes into account the essential feature of wave-
packet models, which is the dynamical suppression of coherences between
di®erent neutrino mass states and the corresponding damping of the °avor
oscillation probability.17
In both cases, we will calculate the QFI and the population measurement Fischer
Information (PMFI), i.e. the Fischer information (FI) for °avor detection, which
quanti¯es the sensibility of this commonly used protocol for the estimation of and
the maximum precision achievable with it. We will see that for model (2), direct
measurement of projectors onto eigenstates of the SLD corresponds to direct neutrino
mass measurement. Moreover, we will show that for both models, the PMFI is op-
timized and equals the QFI at speci¯c neutrino times-of-°ight. Therefore, although
the usual °avor measurement is not the one which achieves the optimal precision as
set by the QFI, it can in principle be implemented with the best possible sensitivity to
the desired parameter .
Next, we use the results obtained to address how the single-particle, mode en-
tanglement relates to the QFI and the FI for direct °avor detection. We show that
this entanglement does not enhance either of these quanti¯ers. In particular, our
1750045-3
results demonstrate that at least in single-particle settings, entanglement is not di-

rectly connected with the optimal precision in metrological tasks.
This work is organized as follows. In Sec. 2, we review the basics of QET and set
our notation for two-°avor neutrino oscillations in the plane-wave approach. We also
discuss occupation number entanglement in neutrino oscillations (following Blasone
et al.15). Next, we consider in Sec. 3 the QFI and SLD for model (1) above. We will see
that in this case, there is no clear physical interpretation for the projectors onto
eigenspaces of the SLD in terms of the typical direct measurements performed over
neutrinos in oscillation experiments. In Sec. 4, we show that this is no longer the case
when one takes into account the dynamical suppression of quantum coherence terms
between the di®erent mass eigenstates that compose the initial state. This is the main
section of the present work since this e®ect always occurs to some extent in oscillation
experiments. We model it by introducing in the plane-wave model a Markovian
decoherence term. It will be given by a Lindblad operator that projects into one of the
mass eigenstates, thus damping the coherences between them. Solving the associated
Lindblad master equation, we show that the QFI is the same found for model (1), but
that now the projectors onto the invariant subspaces of the SLD that allow for
optimal precision correspond to direct neutrino mass measurement. The quantitative
relationship between entanglement and the Fischer Information will be analyzed for
each type of model in the section dedicated to it. We also calculate the PMFI in each
case and characterize its optimization. Finally, Section 5 summarizes the results and
closes with a few concluding remarks.
2. Quantum Estimation Theory and Neutrino Oscillations

2.1. Quantum estimation theory
Estimation theory is necessary when we want to know the value of a parameter of a
physical system that does not directly correspond to a physical observable.17 In this
case, we make some measurements on the system and infer the value of the parameter
based on the measurement outcomes. These problems are properly treated in the
framework of QET.12 Basically, for a given set of measurements outcomes , we look
^
for an estimator ðÞ ^
that maps every possible outcome to a value ¼ ðÞ of the
parameter we want to estimate. A result of estimation theory is that the variance of
any estimator ^ is bounded from below according to the Cramer–Rao inequality13
^ 1
VarðÞ ; ð1Þ
MF ðÞ
where M denotes the number of measurements and F ðÞ is the FI:
X 1 @pðxjÞ 2
F ðÞ : ð2Þ
x
pðxjÞ @
1750045-4
The sum runs over all possible measurement outcomes and pðxjÞ is the condi-
tional probability of obtaining the outcome x given that the value of the parameter
we want to estimate is .
In quantum mechanics, every information property of a given measurement can be
completely described by a set fx; x g of measurement outcomes x and non-negative
operators x , such that the probability to get the value x is Tr½x . Here, denotes
the density matrix of the system of interest.19 Such sets are called POVMs. When
the parameter has the value , the system is described by a density matrix and
we have
pðxjÞ ¼ Tr½ x : ð3Þ
We de¯ne the symmetric logarithmic derivative (SLD) L by the equation

1
@ f ; L g: ð4Þ
2
By substituting Eqs. (3) and (4) into (2), the FI reads
X ½ReTrð x L Þ 2
F ðÞ ¼ : ð5Þ
x
Tr½ x
From the above equation, we ¯nally get18,20

X ½ReTrð x L Þ 2
F ðÞ ¼
x
Tr½ x

Tr½ L 2
x
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Tr½ x
" pffiffiffiffiffipffiffiffiffiffiffi # 2
x pffiffiffiffiffiffi pffiffiffiffiffi

¼ Tr pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi x L
Tr½ x
X
Tr½x L L
x
¼ Tr½ L 2 : ð6Þ
Hence, we conclude that the FI is bounded by the so-called QFI HðÞ, de¯ned by
HðÞ Tr½ L 2 ¼ Tr½ð@ ÞL : ð7Þ
The QFI determines the highest possible precision in the unbiased estimation of a
parameter allowed by quantum mechanics. To reach that precision, one needs to
saturate inequality (6). It can be shown that this happens for the POVM consisting of
the projectors onto the invariant subspaces of L .18 Moreover, one can show that the
SLD and the QFI can be written explicitly as
Xh m jð@ Þj n i
L ¼ 2 j m ih n j; ð8Þ
n;m
m þ n
1750045-5
and
X jh n jð@ Þj m ij
2
HðÞ ¼ 2 ; ð9Þ
n;m
n þ m
P
where the sum extends over all terms with n þ m 6¼ 0 and ¼ n n j n ih n j
(see, e.g. Ref. 12). This solves the problem from a mathematical viewpoint, although
direct measurement of these projectors may be technologically unfeasible.
2.2. Two-°avor neutrino oscillations and mode entanglement

Neutrino oscillations is the phenomenon by which neutrinos produced with a well-
de¯ned °avor change this property as they propagate in free space.7,10 This happens
because they are massive and the mass eigenstates do not coincide with the °avor
eigenstates, the latter being a superposition of the former. Recall that in the standard
plane-wave model with only two neutrino °avors (e.g. e and ), which is a relevant
approximation for several practical neutrino oscillation scenarios,21,22 this mixing is
described by a single parameter (mixing angle) through the relations1,10
!
j ei cosðÞ sinðÞ j 1i
¼ ; ð10Þ
sinðÞ cosðÞ j 2i
where Hj i i ¼ Ei j i i; i ¼ 1; 2, with H being the Hamiltonian. We will assume in the

sequence that m2 > m1 .
Neutrinos are always produced in a °avor eigenstate,10 which we will take from
now on as an electron neutrino. Moreover, we will always work in the mass basis, in
which the evolution of such a neutrino density matrix is
0 1
2 ðÞ
1
B cos sinð2Þe i
C
¼@1 2 A; ð11Þ
sinð2Þe i sin 2 ðÞ
2
ðm 2 m 2 Þ
where t t 22E 1 and t is proper time. In the experiments with ultra-rela-
tivistic neutrinos, t can be expressed in terms of neutrino time-of-°ight, i.e. the
distance traveled by the neutrino between the source and detector.
Finally, let us quickly discuss the issue of entanglement in neutrino oscillations. As
mentioned in Sec. 1, the natural form of entanglement to consider in this context is
single-particle, mode entanglement. The general idea is very simple, and we refer the
interested reader to Blasone et al.15,22 for details. First, note that we can look at the
neutrino state space H as the two-qubit Hilbert space H1 H2 spanned by fj1i1
j0i2 ; j0i1 j1i2 g by means of the unitary equivalence de¯ned on the mass basis by
j 1 i 7! j1i1 j0i2 and j 2 i 7! j0i1 j1i2 . Then, observe that in this two-qubit re-
presentation,22 there is a bipartition of the space of quantum states available, relative
to which entanglement can be considered. Correspondingly, a neutrino state which is
entangled as a two qubit state is said to be mode entangled. This type of single-
1750045-6
particle entanglement is well studied in the QIT literature,16 and several recent works
investigate its properties in the context of neutrino oscillations (e.g. Refs. 17, 23–27).
This is the approach we will use in the sequence for entanglement in neutrino
oscillations.
3. Mixing Angle Estimation: Plane-Wave Model

Using the explicit form of in Eq. (11) in Eqs. (8) and (9), the SLD reads

sinð2Þ cosð2Þe i
L ¼ 2 ; ð12Þ
cosð2Þe i sinð2Þ
and the QFI is
HðÞ ¼ 4: ð13Þ
In this case, the operator L has the eigenvalues 2 with corresponding
eigenvectors
!
1 ð1
sinð2ÞÞ 1=2
j 2i ¼ pffiffiffi : ð14Þ
2 e i ð1 sinð2ÞÞ 1=2
The corresponding projectors to be measured in order to saturate the QFI are

1 1
sinð2Þ cosð2Þe i
j 2ih 2 j ¼ : ð15Þ
2 cosð2Þe i 1 sinð2Þ
However, these projectors have no clear interpretation in terms of the typical direct
measurements performed over neutrinos in oscillation experiments. In fact, neutrino
interactions can be of two types28: charged-current processes, which project the
neutrino in a de¯nite °avor state, or neutral-current processes, which are insensitive
to °avor. Thus, the type relevant for oscillation experiments is the former. In such
processes, the neutrino is absorbed and a charged lepton is emitted (electron, muon,
tau) which directly reveals the neutrino °avor state at the interaction. Apart from
these °avor measurements, charged-current reactions also contain information of
kinematic nature (energies, momenta) that in principle allow the direct measurement
of a neutrino's mass state at the interaction event. But due to the extreme smallness
of these masses, direct neutrino mass measurements in this type of experiment are
currently unfeasible29 (the current approaches are based on decay processes, see
Ref. 30).
Therefore, the SLD above cannot be used to determine an optimal measurement
protocol for the mixing angle in practice. However, we stress that this does not render
the SLD useless from a theoretical viewpoint: it is necessary to compute the FI
associated to speci¯c generalized measurements (POVMs) and quantitatively
address how they deviate from optimality.
1750045-7
For the sake of comparison, let us compute then the FI associated with °avor
measurements, which is what one actually detects in neutrino oscillation experi-
ments. Despite the fact that Eq. (3) is always correct, we stress that the proper way
to compute the FI is by Eq. (5) (using the result for the SLD) and not by Eq. (2). This
is because in this case, the POVM, and consequently the measurement outcomes,
depend directly on the parameter we want to estimate, and in Eq. (2), the derivative
must be taken at ¯xed values of the measurement outcomes. On the other hand, this
is exactly what is done in Eq. (5). That being said, the POVM is
80 1 0 19
> 1 1 >
< cos 2 ðÞ sinð2Þ sin 2 ðÞ sinð2Þ =
B 2 C B 2 C
@1 A; @ 1 A ; ð16Þ
>
: >
;
sinð2Þ sin 2 ðÞ sinð2Þ cos 2 ðÞ
2 2
which gives the °avor FI

4cos 2 ð2Þsin 2 2
Fflavor ðÞ ¼ : ð17Þ
1 sin 2 ð2Þsin 2 2
In Fig. 1, we contrast the behaviors of the FI associated to °avor measurements

and of the QFI for a few values of the mixing angle . We can see that the FI
for population measurement is optimized periodically, where it becomes equal to
the QFI.
To ¯nish this section, we discuss the role played by mode entanglement. We
already know that the QFI is constant, and therefore, it de¯nitively cannot be
enhanced by entanglement in the probe (neutrino) state. In order to compare the
behaviors of the FI and of the amount of quantum entanglement in the state ðtÞ,
recall that entanglement in pure bipartite systems is completely characterized by
Fig. 1. (Color online) Comparison between the FI associated with °avor measurements for the experi-
mental value of the mixing angle 12 (see Ref. 9) (dashed), for ¼
8 (dotted) and the QFI (solid).
1750045-8
Fig. 2. (Color online) Contrast between the °avor FI (dashed) and the scaled von Neumann entropy
(solid), with the mixing angle set at the experimental value.9
either the von Neumann entropy of one of the reduced system states, or else by any
of its monotones.31,32 In the present work, we will simply rescale the von Neumann
entropy so that it has the same maximum value as the QFI. As we can see in Fig. 2,
entanglement also does not contribute here to enhance the precision of °avor
measurements, since the local minima of entanglement match the local maxima of
the FI.
4. Mixing Angle Estimation: Decoherence Model

We now consider a model with decoherence of the neutrino oscillations, i.e., a model
where the interference terms between the mass eigenstates that compose the initial
de¯nite °avor neutrino are dynamically suppressed. This is the case relevant for
the neutrino oscillation experiments, where this decoherence e®ect is observed in
practice.
To this aim, we introduce a Markovian term in the dynamics of the neutrinos. The
reason for this is twofold. First, given that neutrinos interact so rarely, it is reason-
able to assume that the characteristic time of correlations between the neutrinos and
any environment is su±ciently short in comparison with the typical time of their
evolution, which is the starting point for a Markovian quantum dynamics.33,34
Moreover, in more appropriate models of neutrino oscillations such as the wave
packet model,10,35 the cause of decoherence in neutrino oscillations is the separation
of the wave packets of di®erent mass neutrinos due to their distinct group velocities.
This is equivalent to be able to identify which mass neutrino is arriving at the
detector. Therefore, although not derived from an underlying microscopic theory
of the interaction of the neutrinos, we can incorporate these considerations in the
1750045-9
plane-wave model by inserting a Lindbladian

pffiffiffi
A ¼ j 1 ih 1 j; ð18Þ
where > 0 is a decoherence parameter, and solving the resulting Markovian
equation in Lindblad form,34,36
d 1
ðtÞ ¼ i½H; ðtÞ þ AðtÞA † fðtÞ; A † Ag: ð19Þ
dt 2
The solution is
0 1
1
sinð2Þe ði 2 Þt C

B cos 2 ðÞ
ðtÞ ¼ @ 1 2 A: ð20Þ
sinð2Þe ði þ 2 Þt

sin 2 ðÞ
2
This time, observe that the eigenvalues of are
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 cos 2 ð2Þ þ e t sin 2 ð2Þ
¼ ; ð21Þ
2
the corresponding eigenvectors being
0 1
1 @ A;
j i ¼ pffiffiffi ½ð1 þ 2 Þ 1=2
1 1=2 ð22Þ
2ð1 þ 2 Þ 1=4 e i t ½ð1 þ 2 Þ 1=2
1 1=2

with tanð2Þe 2 t .
The matrix elements of @ with respect to this basis are
8
>
> sinð2Þð1 e t Þ
>
> h jð@ Þj i ¼
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
< 1 þ tan 2 ð2Þe t
ð23Þ
>
> cosecð2Þj tanð2Þe 2 t j

>
>h jð@ Þ j i ¼ p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi :
: þ
1 þ tan 2 ð2Þe t
Inserting them into Eq. (9), we obtain

H ðÞ ¼ 4: ð24Þ
The subscript is to remind us that in this case, we are dealing with a decoherence
term in the dynamics. We see that despite that, the QFI remains the same as the one
found for the plane-wave model in Sec. 3.
Although the QFI is not a®ected by the decoherence term, the situation becomes
very di®erent when one considers the SLD. In the present case, using Eqs. (21)–(23)
in Eq. (8), we ¯nd

tanðÞ 0
L ¼ 2 : ð25Þ
0 cotðÞ
1750045-10
This result shows that here, the optimal measurement given by the SLD for the
estimation of the mixing angle corresponds to direct mass measurement.
The result for the SLD in Eq. (25) is surprising in the sense that it does not depend
on t > 0. However, it is easy to check its validity. In fact, direct substitution shows
that the right-hand side of Eq. (25) validates the SLD de¯ning Eq. (4). Moreover, this
equation has a unique solution when is positive (see, e.g. Ref. 37), which is the case
here (as one can see from Eq. (21)).

Continuing, we look at the behavior of the FI associated with the °avor mea-
surement. Using the above SLD, a calculation in the same spirit as in Sec. 3 gives

ðÞ 4cos 2 ð2Þ½1 e 2 t cosð tÞ
F flavor ðÞ ¼ : ð26Þ
2 sin 2 ð2Þ½1 e 2 t cosð tÞ
We note that in the ! 0 limit of the above expression, one obtains again relation
(17). Another interesting feature is that the FI tends to a residual FI
4cos 2 ð2Þ
Fres ¼ : ð27Þ
1 þ cos 2 ð2Þ
This is the FI after decoherence has taken place, which is well below the upper limit
given by the QFI at the experimental value of the mixing angle. In Fig. 3, we contrast
the behaviors of the FI for di®erent . For values of this parameter within the
physically relevant range , we see that the FI for population measurement is still
optimized periodically. This time, however, it never reaches the maximum value
established by the QFI. The local maxima of the QFI reduce at each period.
Finally, let us analyze entanglement in this model. This time, we can no longer
rely on the von Neumann entropy as in Sec. 3, since the system will be in a mixed
state for every t > 0. However, given that the system is formally equivalent to a pair
Fig. 3. (Color online) Plots of the QFI (solid) and the FI associated with the °avor measurement

at the experimental value9 of the mixing angle. Decoherence parameter taken as 0 (dashed), 10 (dotted),
(dot-dashed) and 10 (double-dot-dashed).
1750045-11
Fig. 4. (Color online) Comparison between the °avor FI (dashed) and the logarithmic negativity (solid)
for di®erent values of : ¼ =10 (top) and ¼ (bottom).
of qubits, we can use the logarithmic negativity 31,38 as a proper quanti¯er of entan-
glement. Considering the QFI, since it is again constant, we conclude it is also not
enhanced here by mode entanglement. In Fig. 4, we do the same comparison between
entanglement and the FI as we did in the previous section for di®erent values of the
decoherence parameter . Again, the local maxima of the FI match the local minima
of the entanglement, showing that it also does not contribute to the sensibility of the
population measurement protocol.
5. Conclusions
We investigated estimation of the mixing parameter characterizing two-°avor
neutrino oscillations from a QET perspective. For neutrinos initially in a de¯nite
°avor state, we considered two types of models to describe the ensuing °avor oscil-
lations: (1) the standard plane-wave approach; (2) a model which takes into account
1750045-12
the damping of quantum coherences between the di®erent mass states that compose
the °avor state. Since this type of \decoherence e®ect" always occurs to some extent
in experiments, the results obtained in the second case are the ones to relate to the
physical applications. In this case, we demonstrated (Sec. 4) that the measurement
scheme which realizes the optimal precision limit established by the QFI for unbiased
estimation of can be determined to be the direct mass measurement.
Interestingly, direct mass measurements are optimal in the plane-wave model as

well. Since the decoherence parameter does not appear in Eq. (25), one could expect
these to remain optimal in the limit of zero decoherence. This is in fact the case, as
can be seen by calculating the FI in the plane-wave model for the POVM fj 1 ih 1 j;
j 2 ih 2 jg and verifying that it also equals HðÞ ¼ 4. Thus, there are at least two
distinct ways of saturating the QFI in the model without decoherence.

As we discussed in Sec. 3, direct detection of neutrino masses in oscillation
experiments is not within the reach of current technology. But we also analyzed here
the FI associated to the population measurement protocol, which is the one employed
to estimate . For both types of models, we saw (Secs. 3 and 4) that this FI is
optimized periodically. Equivalently, this means that its local maxima occur at
speci¯c neutrino times-of-°ight. Therefore, although the usual °avor measurement is
not the one which realizes the QFI, it can in principle be implemented with the best
possible sensitivity to the desired parameter .
We also investigated how the single-particle, mode entanglement in the oscillating
neutrino system relates to the QFI and the FI for direct °avor detection (Secs. 3
and 4). We showed that this entanglement does not enhance either of these quan-
ti¯ers. In fact, quite to the contrary, we found in all the cases considered that while
the QFI does not change regardless of the entanglement variation in time, the FI for
population measurement has its local maxima simultaneously with the entangle-
ment's local minima. In particular, this demonstrates that in single-particle settings,
the presence of entanglement correlations in probe states does not imply on precision
enhancement in estimations tasks.
We have studied the two-°avor scenario because we often ¯nd simple analytic
expressions for all quantities of interest because it enables us to understand many
aspects of neutrino oscillations that certainly will be present in more sophisticated
models, and also because of its applicability in several experimental oscillations
settings. Nevertheless, it should be interesting to make an analysis similar to the one
presented here for three-°avor oscillations. This would allow one to examine the
richer case of single-particle multipartite entanglement and obtain the limits of
precision in quantities that are the focus of future neutrino oscillations experiments,
like the CP phase CP and the mixing angle 13 .11
Acknowledgments
E. N. would like to acknowledge the ¯nancial support from the Brazilian institution
CAPES (High-Level Sta® Improvement Coordination). G.S. and M.S. would like to
1750045-13
thank the Departamento de Ciências Exatas e Tecnológicas/UESC — Ilheus for the

hospitality and the ¯nancial support during the development of this work. M. S. also
acknowledges the ¯nancial support from the Brazilian institutions CNPq and
FAPEMIG (Fundação de Amparo à Pesquisa do Estado de Minas Gerais).
References
1. C. Giunti and C. W. Kim, Fundamentals of Neutrino Physics and Astrophysics (Oxford
University Press, Oxford, England, 2007).
€ Physik 88(3) (1934) 161.
2. E. Fermi, Zeitschrift f ur
3. M. Maggiore, A Modern Introduction to Quantum Field Theory, Oxford master series in
physics (Oxford University Press, Oxford, England, 2005).
4. B. Pontecorvo, Zh. Eksp. Teor. Fiz. 33 (1957) 549.

5. Q. R. Ahmad and SNO Collaboration, Phys. Rev. Lett. 89 (2002) 011301.
6. Z. Maki, M. Nakagawa and S. Sakata, Shoichi, Prog. Theoret. Phys. 28(5) (1962) 870.
7. V. Gribov and B. Pontecorvo, Phys. Lett. B 28(7) (1969) 493.
8. S. M. Bilenky and B. Pontecorvo, Phys. Lett. B 61 (1976) 248.
9. K. A. Olive, Rev. Particle Phys. Chin. Phys. C 38(9) (2014) 090001.
10. D. Samitz, Particle Oscillations, Entanglement and Decoherence (Bachelorarbeit,
Universität Wien, 2012).
11. R. Acciari et al., Long-Baseline Neutrino Facility (LBNF) and Deep Underground
Neutrino Experiment (DUNE), (2015), arXiv:1512.06148.
12. C. W. Helstrom, Quantum Detection and Estimation Theory, Mathematics in Science and
Engineering: A series of monographs and textbooks (Academic Press, New York, NY,
1976).
13. H. Cramer, Mathematical Methods of Statistics, Princeton landmarks in mathematics and
physics (Princeton University Press, Princeton, NJ, 1999).
14. G. Tóth and I. Apellaniz, J. Phys. A: Math. Theoret. 47 (2014) 424006.
15. M. Blasone, F. Dell'Anno, S. De Siena, M. Di Mauro and F. Illuminati, Phys. Rev. D 77
(2008) 096002.
16. M. O. T. Cunha, J. A. Dunningham and V. Vedral, Proc. Roy. Soc. Lond. A: Math. Phys.
Eng. Sci. 463(2085) (2007) 2277.
17. V. A. S. V. Bittencourt, C. J. Villas Boas and A. E. Bernardini, EPL 108 (2014)
50005.
18. M. G. A. Paris, International Journal of Quantum Information 07(01) (2009) 125.
19. K. Jacobs, Quantum Measurement Theory and its Applications (Cambridge University
Press, Cambridge, England, 2014).
20. S. L. Braunstein and C. M. Caves, Phys. Rev. Lett. 72 (1994) 3439.
21. B. J. P. Jones, Sterile neutrinos in cold climates, PhD thesis, MIT (2015).
22. M. Blasone, F. Dell'Anno, S. De Siena and F. Illuminati, EPL 85 (2009) 50002.
23. M. Blasone, F. Dell'Anno, S. De Siena and F. Illuminati, EPL 106 (2014) 30002.
24. M. Blasone, F. Dell'Anno, S. De Siena and F. Illuminati, Adv. High Energy Phys. 2014
(2014) 359168.
25. M. Blasone, F. Dell'Anno, S. De Siena, M. Di Mauro and F. Illuminati, EPL 112 (2015)
20007.
26. S. Banerjee, A. K. Alok, R. Srikanth and B. C. Hiesmayr, Eur. Phys. J. C 75 (2015)
487.
27. A. K. Alok, S. Banerjee and S. U. Sankar, Nuclear Physics B 909 (2016) 65.
1750045-14
28. F. Boehm and P. Vogel, Physics of Massive Neutrinos (Cambridge University Press,
Cambridge, England, 1992).
29. C. Weinheimer, Hyper¯ne Interact. 215(1–3) (2013) 85.
30. S. Mertens, J. Phys.: Conf. Ser. 718 (2016) 022013.
31. L. Amico, R. Fazio, A. Osterloh and V. Vedral, Rev. Mod. Phys. 80(2) (2008) 517.
32. R. Horodecki, P. Horodecki, M. Horodecki and K. Horodecki, Rev. Mod. Phys. 81(2)
(2009) 865.
33. H. P. Breuer and F. Petruccione, The Theory of Open Quantum Systems (Oxford
University Press, Oxford, England, 2002).
34. S. Haroche and J. M. Raimond, Exploring the Quantum: Atoms, Cavities, and Photons
(Oxford Graduate Texts, Oxford University Press, Oxford, England, 2006).
35. C. Giunti, C. W. Kim and U. W. Lee, Phys. Rev. D 44 (1991) 3635.
36. G. Lindblad, Comm. Math. Phys. 48 (1976) 119.
37. W. J. Rugh, Linear Systems Theory, Prentice-Hall Information and Systems Sciences
Series (Prentice Hall, Upper Saddle River, NJ, 1993).
38. M. B. Plenio, Phys. Rev. Lett. 95 (2005) 090503.
1750045-15

Quantum Est 2flav

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Quantum Est 2flav

Uploaded by

Copyright:

Available Formats

International Journal of Quantum Information

Vol. 15, No. 6 (2017) 1750045 (15 pages)

Quantum estimation in neutrino oscillations

Edson C. Nogueira*,§, Gustavo de Souza†,¶,

Departamento de F{sica, ICEX,

Received 18 March 2017

Keywords: Neutrino oscillations; QFI; entanglement.

PACS Number(s): 03.65.Ta, 03.65.Ud, 14.60.Pq

described by a mixing matrix U which contains some parameters — mixing angles,

(1) the standard plane-wave model;

results demonstrate that at least in single-particle settings, entanglement is not di-

2. Quantum Estimation Theory and Neutrino Oscillations

We de¯ne the symmetric logarithmic derivative (SLD) L by the equation

From the above equation, we ¯nally get18,20

HðÞ Tr½ L 2  ¼ Tr½ð@ ÞL : ð7Þ

2.2. Two-°avor neutrino oscillations and mode entanglement

where Hj  i i ¼ Ei j  i i; i ¼ 1; 2, with H being the Hamiltonian. We will assume in the

3. Mixing Angle Estimation: Plane-Wave Model

The corresponding projectors to be measured in order to saturate the QFI are

which gives the °avor FI

In Fig. 1, we contrast the behaviors of the FI associated to °avor measurements

4. Mixing Angle Estimation: Decoherence Model

plane-wave model by inserting a Lindbladian

Inserting them into Eq. (9), we obtain

here (as one can see from Eq. (21)).

Interestingly, direct mass measurements are optimal in the plane-wave model as

distinct ways of saturating the QFI in the model without decoherence.

thank the Departamento de Ciências Exatas e Tecnológicas/UESC — Ilheus for the

4. B. Pontecorvo, Zh. Eksp. Teor. Fiz. 33 (1957) 549.

You might also like

HðÞ Tr½ L 2 ¼ Tr½ð@ ÞL : ð7Þ

where Hj i i ¼ Ei j i i; i ¼ 1; 2, with H being the Hamiltonian. We will assume in the