EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH (CERN)

CERN-EP-2022-280
2023/01/16

CMS-HIN-21-006

K0S and Λ (Λ) two-particle femtoscopic

√ correlations in PbPb
collisions at sNN = 5.02 TeV
arXiv:2301.05290v1 [nucl-ex] 12 Jan 2023

The CMS Collaboration

Abstract

Two-particle correlations are presented for K 0S , Λ, and Λ strange hadrons as a func-

tion of relative momentum in lead-lead collisions at a nucleon-nucleon center-of-mass
energy of 5.02 TeV. The dataset corresponds to an integrated luminosity of 0.607 nb−1
and was collected using the CMS detector at the CERN LHC. These correlations are
sensitive to quantum statistics and to final-state interactions between the particles.
The source size extracted from the K 0S K 0S correlations is found to decrease from 4 to
1 fm in going from central to peripheral collisions. Strong interaction scattering pa-
rameters (i.e., scattering length and effective range) are determined from the ΛK 0S and
ΛΛ (including their charge conjugates) correlations using the Lednický–Lyuboshitz
model and are compared to theoretical and other experimental results.

Submitted to Physics Letters B

© 2023 CERN for the benefit of the CMS Collaboration. CC-BY-4.0 license
 1

1 Introduction
Two-particle correlations in relative momentum, so-called femtoscopic correlations, arising
from relativistic heavy ion collisions provide a powerful tool for studying both the quark-gluon
plasma (QGP) that is created in the collisions, and the subsequent interactions of the emitted
particles [1]. All two-particle correlations are affected by final-state interaction (FSI) effects,
and correlations of identical particles are also sensitive to the constraints of quantum statistics
(QS). The correlations among the neutral K 0S , Λ, and Λ particles, collectively referred to as V0
particles, are of special interest. First, they can be used to determine the space-time extent of
the QGP. In addition, information can be extracted about the strong-interaction scattering pa-
rameters, i.e., the scattering length and the effective range, that is impossible to obtain from
currently achievable scattering experiments [2–6]. Because of their relatively heavy mass and
the absence of a Coulomb interaction, femtoscopy based on K 0S particles supplements the more
commonly studied pion and charged kaon pairs [7]. The results from ΛΛ correlation studies
can help constrain baryon-baryon and, more specifically, hyperon-hyperon interaction models
that are used, for example, in modeling the composition of neutron stars [8–10].
Regarding the scattering parameters, of particular interest is establishing whether the interac-
tion between two Λ particles allows for the existence of the H-dibaryon, a bound state with
quantum numbers I = 0, J P = 0+ , S = −2. In 1977, R. L. Jaffe predicted the existence of such a
six-quark (uuddss) state having a mass about 81 MeV below the threshold of twice the Λ mass
by considering the strong attraction resulting from color magnetic interactions [11]. Although
a double hypernucleus, ΛΛ6 He, was subsequently observed in the NAGARA event from the
E313 hybrid emulsion experiment at KEK [12, 13], the observed ΛΛ binding energy was not
consistent with the conjectured H-dibaryon [14]. A study of ΛΛ correlations may provide ad-
ditional information on whether the baryon-baryon interaction can lead to the formation of the
conjectured H-dibaryon.
Recently, the ALICE Collaboration reported on√ ΛK correlations in lead-lead (PbPb) collisions
at a center-of-mass energy per nucleon pair of sNN = 2.76 TeV [15]. According to their find-
ings, the strong force is repulsive in ΛK + interactions, yet attractive in ΛK − interactions. For
the ΛK 0S pairs, the uncertainty of the ALICE results does not permit a definite conclusion
on whether the associated strong interaction is repulsive or attractive. A more precise mea-
surement of ΛK 0S correlations should improve our understanding of the strong interaction in
baryon-meson systems.
This Letter presents K 0S K 0S , ΛK 0S , √
and ΛΛ femtoscopic correlations as a function of relative
momentum in PbPb collisions at sNN = 5.02 TeV, using data recorded by the CMS experi-
ment during the 2018 LHC run. The K 0S K 0S correlations are measured in six centrality intervals
within the 0–60% range, where centrality refers to the percentage of the total inelastic hadronic
nucleus-nucleus cross section [16], and 0% corresponds to the maximum overlap of the col-
liding nuclei. The K 0S K 0S , ΛK 0S , and ΛΛ correlations are measured in an integrated centrality
range (0–80%), with the ΛΛ femtoscopic correlation measured in PbPb collisions at the LHC
for the first time. The source size and strong interaction parameters are determined using the
Lednický–Lyuboshitz (LL) model [17]. Unless otherwise indicated, all measurements include
the charge conjugate states, so ΛK 0S and ΛΛ include Λ K 0S and Λ Λ , respectively. Tabulated
results are provided in the HEPData record for this analysis [18].
2

2 Experimental setup and data sample

The central feature of the CMS apparatus is a superconducting solenoid of 6 m internal diam-
eter, providing a magnetic field of 3.8 T. Within the solenoid volume there is a silicon pixel
and strip tracker, a lead tungstate crystal electromagnetic calorimeter, and a brass and scintil-
lator hadron calorimeter, each composed of a barrel and two endcap sections. The silicon pixel
detector [19] is composed of 1856 silicon pixel modules distributed in four 54 cm long bar-
rel layers at radii of 2.9–16.0 cm plus three pairs of endcap disks covering radii of 4.5–16.1 cm
at longitudinal distances of 31–51 cm from the origin. The 15 148 silicon strip module are ar-
ranged in 10 barrel layers at radii of 20–116 cm plus 3 pairs of small and 9 pairs of large endcap
disk layers. Charged particles of pseudorapidity |η | < 3 are reconstructed with the combined
system. For particles with transverse momentum of 1 < pT < 10 GeV, the track resolutions
are typically 1.5% in pT and 20–75 µm in the transverse impact parameter [20]. The barrel and
endcap detectors are extended to the forward region with two calorimeters which use steel as
the absorber and quartz fibers as the sensitive material. These hadron forward (HF) calorime-
ters are located 11.2 m from the interaction region, one on each side, and provide coverage in
the range 3.0 < |η | < 5.2. These detectors are segmented into multiple 0.175×0.175 (∆η ×∆φ)
“towers”, where φ is azimuthal angle in radians. Muons are measured in gas-ionization detec-
tors embedded in the steel flux-return yoke outside the solenoid. Events of interest are selected
using a two-tiered trigger system. The first level, composed of custom hardware processors,
uses information from the calorimeters and muon detectors to select events at a rate of around
100 kHz within a fixed latency of about 4 µs [21]. The second level, known as the high-level
trigger, consists of a farm of processors running a version of the full event reconstruction soft-
ware optimized for fast processing, and reduces the event rate to around 1 kHz before data
storage [22]. A more detailed description of the CMS detector, together with a definition of the
coordinate system used and the relevant kinematic variables, can be found in Ref. [23].
With an integrated luminosity of 0.607 nb−1 [24, 25], this analysis uses 4.27 × 109 minimum bias
events that are triggered by requiring signals above the readout threshold of 3 GeV in each of
the HF calorimeters [22]. Background events due to beam-gas interactions and non-hadronic
collisions are filtered offline by applying the procedure described in Ref. [26]. The events used
in this analysis are required to have at least one primary interaction vertex determined using
two or more tracks [27] within a distance of 15 cm from the center of the nominal interaction
point along the beam axis and to have at least two calorimeter towers in each HF detector with
energy deposits of more than 4 GeV per tower. The shapes of the clusters in the pixel detector
are required to be compatible with those expected in PbPb collisions in order to suppress the
contamination from events with multiple collisions [28]. The combined trigger and offline
selection efficiency for inelastic events is greater than 95%. The event centrality is obtained from
the transverse energy deposited in both HF calorimeters, using the methodology described in
Ref. [29]. The analysis makes use of a minimum bias Monte Carlo PbPb sample, based on the
HYDJET 1.9 [30] event generator with a full detector simulation using G EANT 4 [31].

3 Reconstruction of K 0S and Λ candidates

The K 0S and Λ candidates, denoted as V0 candidates, used in this study are reconstructed as in
previous CMS analyses [32–34]. The V0 candidates are found by combining oppositely charged
tracks that pass criteria based on the “loose” selection discussed in Ref. [27]. The charged tracks
are assumed to be π + π − in K 0S reconstruction and π − p in Λ reconstruction. For the latter, the
higher momentum track is assumed to be a proton since the proton carries nearly all of the mo-
mentum in the Λ decay. Each of the oppositely charged tracks must have hits in at least three
 3

layers of the silicon tracker, and both tracks must have transverse and longitudinal impact pa-
rameter significances (defined as the parameter value divided by its uncertainty) with respect
to the primary vertex greater than 1. The two tracks are fitted to a common vertex and the χ2
per degree of freedom (dof) from the fit must be less than 7. The distance of closest approach
between the two tracks is required to be less than 1 cm. As a consequence of the long lifetime of
K 0S and Λ particles, the significance of the V0 decay length, which is the three-dimensional dis-
tance between the primary and V0 vertices divided by its uncertainty, is required to be greater
than 2.5 to reduce combinatorial background contributions. To remove K 0S candidates misiden-
tified as Λ particles and vice versa, the Λ (K 0S ) candidates must have a corresponding ππ (pπ )
mass more than 14 (7) MeV (corresponding to approximately 3 times the average resolution)
away from the world-average value [35] of the K 0S (Λ) mass. The angle θ between the V0 mo-
mentum vector and the vector connecting the primary and V0 vertices is required to satisfy
cos θ > 0.999. This reduces the contribution from nuclear interactions, random combinations
of tracks, and Λ particles originating from weak decays of Ξ and Ω particles.
Further selection of V0 candidates is performed with a boosted decision tree (BDT) [36]. The se-
lection is optimized separately for K 0S and Λ candidates. The discriminating variables include:
the collision centrality, the V0 candidate pT and rapidity (y), the distance of closest approach
of the track pair, the three-dimensional decay length and significance, cos θ, and the V0 ver-
tex fit χ2 . The included variables related to the V0 daughters are pT , uncertainty in pT , η, the
number of hits in the silicon tracker, the number of pixel detector layers with hits, and the
transverse and longitudinal impact parameter significances with respect to the primary ver-
tex. The BDT training is performed using the simulated minimum bias sample separated into
the signal and background subsamples using the generator-level information. The K 0S mesons
are selected with 1 < pT < 8.5 GeV and |y| < 1, while the Λ baryons are required to have
1.8 < pT < 8.5 GeV and |y| < 1. The minimum pT and maximum y requirements are used to
reduce background while the maximum pT requirement is to reduce contributions from jets.
The combined V0 reconstruction and selection efficiencies are strongly dependent on the cen-
trality of the event and the pT of the V0 . Integrating over the selected pT ranges, the combined
efficiencies from the most central to peripheral PbPb collisions are 1–3% for K 0S and 1–2% for
Λ. The V0 reconstruction algorithm does not prevent a track from being used for more than
one V0 . While this is normally an infrequent occurrence, selecting pairs of V0 particles close
together in phase space makes it a significant contribution. To resolve this problem, for each
correlation measurement, a check of each pair of V0 candidates is performed and if two V0 can-
didates are found to share one or both daughter tracks, one of the V0 candidates is randomly
selected to be removed from the event.
Fits to the invariant mass spectrum are performed using a sum of three Gaussian functions
with a common mean to describe the signal distribution and a fourth-order polynomial to de-
scribe the background. These empirical functions were chosen to provide a good description of
the data. Peak and sideband invariant mass regions are defined to select events dominated by
signal and background, respectively. Defining σ as the average resolution based on the Gaus-
sian sum, the peak regions are selected to be within ±2σ from the nominal V0 mass and are
given by 486 < M(π + π − ) < 509 MeV and 1111.5 < M(pπ − ) < 1120.4 MeV for K 0S and Λ
candidates, respectively. The sideband regions are selected to be more than 4σ from the nom-
inal V0 mass and are given by 23.5 < |M(π + π − ) − 497.5| < 62.5 MeV for K 0S candidates and
1080 < M(pπ − ) < 1107.5 MeV together with 1124.2 < M(pπ − ) < 1160 MeV for Λ candidates.
Examples of invariant mass distributions for π + π − and pπ − pairs, and their corresponding
fits in the 0–80% centrality range, are shown in Fig. 1.
4

CMS PbPb, sNN = 5.02 TeV (0.607 nb -1) CMS PbPb, sNN = 5.02 TeV (0.607 nb -1)
108
0
KS Data Λ+Λ Data
Fit Fit
Candidates / (0.5 MeV)

Candidates / (0.5 MeV)

Centrality: 0-80 % Centrality: 0-80 %
Background 106 Background
1 < p < 8.5 GeV Peak region 1.8 < p < 8.5 GeV Peak region
T T
107 |y| < 1 Sideband region |y| < 1 Sideband region

105
6
10

0.45 0.5 0.55 1.08 1.1 1.12 1.14 1.16

π+π− invariant mass (GeV) pπ−+ pπ+ invariant mass (GeV)
Figure 1: The invariant mass of K 0S (left) and Λ (right), and their corresponding fits in the 0–80%
centrality range. The circles are the data, and the fit is shown with a solid (red) line for the total
fit, and a dashed (green) line for the background fit. The vertical dashed-dotted (pink) lines
indicate the peak region and the vertical dashed (blue) lines indicate the sideband regions.

4 Analysis method
The two-particle correlation is constructed as

Aobs (qinv )
Cobs (qinv ) = N , (1)
Bobs (qinv )

where Cobs (qinv ) is the observed normalized pair yield, corrected for detector effects, as a func-
tion of the invariant relative momentum qinv , defined as [1]
q
qinv = − Qµ Qµ ,
(k1 − k2 )µ Pµ µ (2)
Qµ = (k 1 − k 2 )µ − P ,
Pµ Pµ

where P = k1 + k2 , and k1 and k2 are the four momenta of the V0 particles. Note that for two
particles of the same mass, the second term of Qµ is zero.
The distribution Aobs (qinv ) is the signal distribution that contains femtoscopic correlations
formed by pairing the selected V0 particles from a given event. The reference distribution
Bobs (qinv ) is used to correct for phase space effects, largely removing artifacts due to detector
non-uniformities in the Aobs (qinv ) distribution. The Bobs (qinv ) distribution is constructed by
mixing the V0 particles from different events [37]. In this procedure, the V0 particle from one
event is paired with V0 particles from 30 different events. To ensure that the 30 events used
in the mixing are similar to the signal events, the centrality and primary vertex of each mixed
event must be within 5% and 2 cm, respectively, of those in the corresponding signal event.
The normalization factor N is the ratio of the number of pairs in the reference distribution
to that in the signal distribution. Because of the background in the peak region of the invari-
ant mass distributions, the measured signal distribution ( Aobs (qinv )) contains contributions
from signal-signal ( Ass (qinv )), signal-background ( Asb (qinv )), and background-background
( Abb (qinv )) correlations. The measured Aobs (qinv ) distribution can be written as

Aobs (qinv ) = f ss Ass (qinv ) + f sb Asb (qinv ) + f bb Abb (qinv ). (3)

 5

The distributions Asb (qinv ) and Abb (qinv ) are obtained from the peak-sideband and sideband-
sideband combinations, respectively. The small amount of background (signal) contamination
in the signal (sideband) region has a negligible effect on the shape of Asb (qinv ) ( Abb (qinv )). All
distributions, Aobs (qinv ), Asb (qinv ), and Abb (qinv ) are normalized to unity. The parameters, f ss ,
f sb , and f bb are the signal-signal, signal-background, and background-background fractions,
extracted using an invariant mass fit based on combinatorial analyses with

(2s )
f ss = ,
(s+2 b)
(b2) (4)
f bb = , and
(s+2 b)
f sb = 1 − f ss
− f bb ,
n ( n −1)
where (n2 ) = 2 is the binomial coefficient, which returns the number of ways that a pair can
be chosen from n objects. The quantities s and b in the binomial coefficients are the number of
signal and background particles, respectively, obtained by integrating the appropriate function
from the fit to the invariant mass distribution.
Once we have all the distributions ( Aobs (qinv ), Asb (qinv ), and Abb (qinv )) and the parameters
( f ss , f sb , and f bb ), the Ass (qinv ) distribution can be extracted using Eq. (3), with
h i
Ass (qinv ) = Aobs (qinv ) − f sb ( Asb (qinv )) − f bb ( Abb (qinv )) / f ss . (5)

The same procedure is followed for the reference distribution. After extracting the Ass (qinv )
and Bss (qinv ) distributions, the correlation distribution is calculated as

Ass (qinv )
Css (qinv ) = N . (6)
Bss (qinv )

While the Css (qinv ) distribution is corrected for detector effects and non-V0 backgrounds, it
still includes non-femtoscopic background correlations, such as those associated with elliptic
flow [38], minijets [7], resonance decays [7], and energy-momentum conservation [39]. The
non-femtoscopic background contribution is modeled using an empirically determined double
Gaussian function   
2 2 2 2
Ω(qinv ) = N 1 + α1 e−qinv R1 1 − α2 e−qinv R2 , (7)

where N, α1 , α2 , R1 , and R2 are fit parameters. This function was selected for its reproduction
of the distributions in both real data at high qinv and simulated data that do not include the
correlations being measured.
Fits are performed to the Css (qinv ) distributions to extract the source size and strong interaction
scattering parameters. As the V0 particles are neutral, the Coulomb interaction is absent. How-
ever, the correlations are sensitive to QS and FSI effects, with s-wave interactions assumed to
dominate for the small relative momenta of the particle pairs analyzed. The correlation distri-
bution for all pairs (K 0S K 0S , ΛK 0S , and ΛΛ) is interpreted in the LL model. This model relates the
two-particle correlation function to the source size and also takes into account FSI effects [17].
The general correlation function is
h  i
Ctotal (qinv ) = 1 + λ CQS (qinv ) + CFSI (qinv ) Ω(qinv ), (8)
6

where CQS (qinv ) is the QS function and CFSI (qinv ) is the FSI function. The parameter λ is re-
ferred to as the incoherence parameter. In the absence of FSI effects, λ equals unity for a
perfectly incoherent Gaussian source. Effects such as resonance decay violate the incoherent
source assumption and can lead to deviations of the λ parameter from the unity. Its value
can also be affected by non-Gaussian components of the correlations function and by the FSI
between particles.
Neglecting CP violation, the K 0S K 0S system can be written as

1 0 0 
|K 0S K 0S i = |K K i + |K 0 K 0 i + |K 0 K 0 i + |K 0 K 0 i . (9)
2
It can be shown [17, 40] that the resulting correlations follow Bose–Einstein quantum statistics,
with
2 2
CQS (qinv ) = e(−qinv Rinv ) , (10)
where the source radius Rinv reflects the size of the region over which the particles are emitted.
The FSI for the K 0S K 0S correlations is modeled by [17, 40]
" 2 #
1  f (k)  4< f ( k ) 2= f ( k )
CFSI (qinv ) = +√ F (q R ) − F (q R ) , (11)
2  Rinv  πRinv 1 inv inv Rinv 2 inv inv

where
k = qinv /2,
Z z
1 2 2
F1 (z) = e−z e x dx, and
2 0 (12)
− z 2
1−e
F2 (z) = .
z
The function f (k ) is the K 0 K 0 s-wave scattering amplitude, with real and imaginary parts < f (k )
and = f (k), respectively. This amplitude is dominated by the near-threshold s-wave isoscalar
resonance f 0 (980) and the s-wave isovector resonance a 0 (980) , with the total scattering ampli-
tude given by an average of these contributions: f (k) = ( f f (980) (k ) + f a (980) (k ))/2. The indi-
0 0
vidual resonance amplitudes depend on the resonance mass mr , with r = f 0 (980) or a 0 (980) ,
the kaon mass mK , and the resonance couplings γr (γr0 ) to the K 0 K 0 (ππ for f 0 (980) and π 0 η
for a 0 (980) channels. Then, f r (k) = γr / m2r − ζ − iγr k − iγr0 k0r , where ζ = 4(m2K + k2 ) and
 

k0r denotes the momentum in the second (ππ or π 0 η) decay channel with the corresponding
partial width Γ0 = γr0 k0r /mr (more details can be found in Ref. [40]). The scattering amplitude
is calculated using the resonance mass and the coupling parameters from Refs. [41–44], taken
from row C of Table 1 of Ref. [40].
For the correlations involving Λ baryons, the CQS (qinv ) and CFSI (qinv ) functions are [17]
2 2
CQS (qinv ) = αe(−qinv Rinv ) , and
" #
1 | f (k )|2 1 d0
2< f ( k ) = f (k)
CFSI (qinv ) = (1 + α) 1− √ +√ F (q R ) − F (q R ) ,
2 R2inv 2 π Rinv πRinv 1 inv inv Rinv 2 inv inv
(13)

where α = −1/2 for ΛΛ correlations for two identical fermions and α = 0 for ΛK 0S corre-
lations as there are no QS effects for non-identical particles [17]. The scattering amplitude
 7

f (k ) is parameterized by a complex scattering length ( f 0 ) and an effective range (d0 ) with

f (k ) = [1/ f 0 + d0 k2 /2 − ik]−1 [17]. The imaginary part of f 0 is responsible for inelastic pro-
cesses (annihilation). For an attractive interaction that is not strong enough to produce a bound
state, the real part of f 0 is positive, while a repulsive interaction corresponds to a negative < f 0
of the order of the range of the repulsive potential. In the presence of a bound state, < f 0 is also
negative, but with a much larger magnitude. The femtoscopic sign convention and notation
for the scattering length differ from those used in nuclear physics, where the corresponding
scattering length a0 = − f 0 . As the ΛK 0S and ΛΛ correlations each have only one spin state that
contributes to the s-wave scattering, Eq. (13) suffices to describe the FSI effects.
Fits to the correlation distribution of all the pairs were performed using Eq. (8) with the non-
femtoscopic background parameters (N, α1 , α2 , R1 , and R2 ) treated as free parameters. For
K 0S K 0S correlations, the parameters of interest are Rinv and λ, with the scattering amplitude
based on previous measurements [41–44]. The ΛK 0S and ΛΛ correlations include additional
parameters: d0 , < f 0 , and = f 0 . The = f 0 term for ΛΛ correlations is set to zero since there are no
baryon-baryon annihilation processes.
Histograms of the correlation distributions are generated in the range 0 < qinv < 6 GeV with
0.02 GeV wide bins for the K 0S K 0S and ΛK 0S correlations and 0.04 GeV wide bins for the ΛΛ cor-
relations. The fits exclude the first qinv bin to avoid a potential bias from the method used to
address cases where V0 candidates share daughter tracks. Studies using simulated events in-
dicate that only this first bin is affected by this remediation. Least-square fits are performed to
the experimental data with the uncertainties in the fit parameters calculated using the MINOS
technique [45]. Examples of correlation measurements and their fits and corresponding χ2 /dof
values, are presented in Figs. 2 and 3. The K 0S K 0S correlations, shown in Fig. 2, are independently
fitted for each of the six centrality bins with 0 < kT < 2.5 GeV, where kT ≡ |~pT,1 + ~pT,2 |/2 is
the average transverse momentum of the particle pair. While the LL model assumes a Gaus-
sian source function, results from charged-particle correlations have demonstrated that this
assumption breaks down for peripheral collisions [46]. This is likely the cause of the poor fit
at low qinv for centralities above 40%. The ΛK 0S (left) and ΛΛ (right) correlations, shown in
Fig. 3, involve fewer events and, therefore, only a single fit is performed for each, with the data
integrated over the centrality range 0–80% and with no restriction on kT .

5 Systematic uncertainties
The systematic uncertainties for the fit parameters are based on the changes found in the pa-
rameter values after individually varying each of the analysis criteria, as discussed below. In
cases with more than one variation for a single source, the maximum deviation from the nom-
inal value is used. The total systematic uncertainty is obtained by adding the uncertainties
from each source in quadrature. The BDT discriminant is varied so as to adjust the signal-to-
background ratio, with the signal yield changing by ±15% in the process. The nominal method
to account for V0 candidates sharing daughter tracks is to remove one of the V0 candidates at
random, which is then not used by any pair. Two alternative approaches are used, one in which
both V0 candidates are removed and another in which, for events with multiple V0 candidate
pair combinations, only the pairs in which the two particles share a daughter are removed.
The systematic uncertainties related to V0 signal and background modeling are investigated by
varying the background shape from a fourth- to a third-order polynomial and the signal shape
from a sum of three Gaussian functions to a sum of two or four Gaussian functions. An alterna-
2
tive non-femtoscopic background function Ω(qinv ) = N (1 + Be−|qinv /σ| )(1 + eqinv ) is used to
assess the uncertainty associated with the choice of the non-femtoscopic background function.
8

1 < p < 8.5 GeV Data

T Full fit PbPb, sNN = 5.02 TeV (0.607 nb-1)
CMS 0 < kT < 2.5 GeV Nonfemto
χ2: 287 KSKS
0 0
0-10% χ2: 323 0
KSKS 10-20%
0 χ2: 312 0
KSKS 20-30%
0
dof: 292 dof: 292 dof: 292
2 1.4 χ2: 25 2 χ2: 18 2 χ2: 13
C (q )

C (q )

C (q )
# bins: 19 1.5 # bins: 19 1.5 # bins: 19
Css(q )

Css(q )

Css(q )
inv

inv

inv
1.2
inv

inv

inv
ss

ss

ss
1
1.5 1.5 1 1.5 1
0.80 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4
q (GeV) q (GeV) q (GeV)
inv inv inv

1 1 1

0 χ2: 306 2 KSKS 4 30-40%6 0 χ : 334 2

0 0 2
KSKS 440-50% 6 0 χ : 353 2
0 0 2 0 0
KSKS 450-60% 6
dof: 292 q (GeV) dof:292 q (GeV) dof: 292 q (GeV)
inv inv 2.5 inv
2 2 χ2: 19 2 χ2: 43 2 χ2: 40
2
C (q )

C (q )

C (q )
# bins: 19 1.5 # bins: 19 # bins: 19
inv

inv

inv
Css(q )

Css(q )

Css(q )
inv

inv

inv
1.5
ss

ss

ss
1.5
1.5 1 1.5 1 1.5 1
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4
q (GeV) q (GeV) q (GeV)
inv inv inv

1 1 1

0 2 4 60 2 4 60 2 4 6
q (GeV) q (GeV) q (GeV)
inv inv inv

Figure 2: The correlation distributions and fits for K 0S K 0S pairs in different centrality ranges,
starting from 0–10% centrality to 50–60% centrality, with 0 < kT < 2.5 GeV. In each plot, the
red circles are the data, the blue solid line is the fit using Eq. (8), and the green dotted line is
the non-femtoscopic background from Eq. (7). The χ2 and dof values are for the full qinv range.
The insert plots show the data and the fit for the qinv < 0.4 GeV region, with the χ2 and number
of bins evaluated in that region.

CMS PbPb, sNN = 5.02 TeV (0.607 nb-1) CMS PbPb, sNN = 5.02 TeV (0.607 nb-1)
1.3 1.6
ΛKS⊕ΛKS
Data 0 0 Data
ΛΛ⊕ΛΛ
Full fit Full fit
1.2 0-80% 1.4 Nonfemto 0-80%
Non-femto
1.8 < pΛ / Λ < 8.5 GeV 1.8 < p < 8.5 GeV
χ2: 349 T χ2: 124 T
dof: 289 KS
0 1.2 dof: 140 all kT
1.1 1<p < 8.5 GeV
Css(q )

Css(q )

T
inv

inv

all kT 1
1
1.1 0.8 1
C (q )

C (q )

0.9
inv

inv

1 0.8
χ2 : 21 0.6 χ2 : 5
ss

ss

0.9 0.6
# bins: 19 # bins: 9
0.8 0.4
0.8 0.4
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4
q (GeV) q (GeV)
inv inv
0.7
0 1 2 3 4 5 6 0 1 2 3 4 5 6
q (GeV) q (GeV)
inv inv

Figure 3: The correlation distributions and fits for (left) and ΛΛ (right) pairs with 0–80% ΛK 0S
centrality and no restriction on kT . In each plot, the red circles are the data, the blue solid line is
the fit using Eq. (8), and the green dotted line is the non-femtoscopic background from Eq. (7).
The χ2 and dof values are for the full qinv range. The insert plots show the data and the fit for
the qinv < 0.4 GeV region, with the χ2 and number of bins evaluated in that region.
 9

The selection requirements used to construct the mixed event sample are varied to require
centrality matching of 3 and 7% instead of the nominal 5% and the primary vertex position
matching with 1 and 3 cm instead of the nominal 2 cm. The effect of the centrality resolution
has been checked and found to be negligible. The peak region requirement is changed from
<2.0σ to <1.5σ and <2.5σ and the sideband region selection from >4.0σ to >3.5σ and >4.5σ.
The upper limit of the qinv fit ranges is changed by ±1 GeV and the lower limit is changed to
include the first bin. At low pT , the tracking efficiency is strongly dependent on pT . There-
fore, the simulated sample is used to explore possible effects of the tracking efficiency. Based
on these studies, it is found that the V0 reconstruction efficiency for the detection of two V0
particles is well described by taking the product of the efficiency for each V0 . It is also found
that the fit results are only weakly affected by the V0 reconstruction efficiency. This is under-
stood as a consequence of the signal and reference samples being similarly affected by the V0
efficiency. Differences in the Monte Carlo and experimental pT spectra could influence the can-
cellation of efficiency-dependent effects in the signal and background correlations. Therefore,
a systematic uncertainty for the efficiency is assessed by comparing the results for the default
simulated sample to one in which the V0 pT distribution is reweighted to match the data. For
the K 0S K 0S correlations, an additional systematic uncertainty is found from varying the mass and
coupling parameters for the f 0 (980) and a 0 (980) resonances by using rows A, B, and D of Table
1 of Ref. [40]. The systematic uncertainties are summarized in Table 1.

Table 1: Summary of absolute systematic uncertainties in K 0S K 0S , ΛK 0S and ΛΛ correlation mea-

surements. The values for Rinv , d0 , < f 0 , and = f 0 are in fm.

K 0S K 0S ΛK 0S ΛΛ
Uncertainty source
Rinv λ Rinv d0 < f0 = f0 λ Rinv d0 < f 0 λ
BDT cut 0.04–0.18 0.01–0.04 0.19 0.75 0.10 0.07 0.03 0.06 0.43 0.05 0.31
Duplicate V0 removal 0.06–0.40 0.01–0.08 0.35 0.92 0.10 0.19 0.11 0.01 1.14 0.05 0.14
Mass fit function 0.00–0.01 0.00–0.01 0.09 0.05 0.01 0.03 0.03 0.02 0.04 0.01 0.02
Non-femtoscopic func. 0.02–0.16 0.01–0.12 0.02 0.17 0.05 0.07 0.03 0.02 1.02 0.14 0.93
Reference sample 0.03–0.08 0.01–0.05 0.22 0.48 0.12 0.12 0.03 0.10 1.12 0.20 0.76
Peak region 0.00–0.07 0.01–0.02 0.43 0.10 0.05 0.17 0.08 0.22 1.21 0.08 0.35
Sideband region 0.00–0.03 0.00–0.01 0.02 0.02 0.01 0.01 0.00 0.01 0.03 0.01 0.04
Fitting range 0.01–0.11 0.01–0.04 0.20 0.18 0.03 0.08 0.04 0.04 1.79 0.20 0.60
Efficiency 0.03–0.03 0.01–0.01 0.08 0.29 0.06 0.09 0.03 0.02 0.05 0.03 0.04
f 0 (980) /a 0 (980) param. 0.07–0.39 0.03–0.05 — — — — — — — — —
Total uncertainty 0.29–0.47 0.08–0.16 0.69 1.34 0.21 0.32 0.16 0.25 2.91 0.34 1.43

6 Results
The size of the particle emitting source Rinv and the λ parameter extracted from the K 0S K 0S cor-
relations for 0 < kT < 2.5 GeV are shown as a function of centrality in Fig. 4. It is observed that
the Rinv value decreases from central to peripheral events, as expected from a simple geometric
√ Rinv = 3.30 ± 0.10 (stat) ±
picture of the collisions. Over the full centrality range of 0–80%,
0.37 (syst) fm. The transverse mass can be calculated as mT = (minv /2)2 + k2T , where minv
is the invariant mass of the two-particle system [15]. The average hmT i is evaluated from the
transverse mass distribution using two-particle pairs with qinv < 0.4 GeV, accounting for back-
ground using the binomial analysis as done for the qinv
√ distributions. Our results for Rinv agree
0 0
with the ALICE K S K S results from PbPb collisions at sNN = 2.76 TeV at a similar mT value [47].
The λ parameter is seen to decrease from about 0.45 to 0.25 as the collisions become more pe-
10

ripheral. This decrease could arise from a relative increase in the contribution from resonance
decays or a source function that becomes increasingly non-Gaussian as the collisions become
more peripheral. The assumption of a Gaussian source function in the LL model may also be
responsible for the relatively poor fits at low qinv for the most peripheral collisions, as seen in
Fig. 2.

CMS PbPb, sNN = 5.02 TeV (0.607 nb -1) CMS PbPb, sNN = 5.02 TeV (0.607 nb -1)
5 1
0
KSKS
0 1 < p < 8.5 GeV 0
KSKS
0
T
4 0.8 0 < kT < 2.5 GeV

3 0.6
Rinv(fm)

λ
2 0.4

1 < p < 8.5 GeV

1 T 0.2
0 < kT < 2.5 GeV

0 0
0 10 20 30 40 50 60 0 10 20 30 40 50 60
Centrality (%) Centrality (%)
Figure 4: The Rinv (left) and λ parameter (right) as a function of centrality. For each data point,
the line and shaded area indicate the statistical and systematic uncertainty, respectively.

Table 2 includes the extracted Rinv and λ parameters as well as hmT i for K 0S K 0S , ΛK 0S , and ΛΛ
combinations in the 0–80% centrality range. A significant decrease is seen in Rinv as the hmT i
increases. Qualitatively similar results have been found, both for a given pair type in bins of
mT and when comparing multiple pair types [1]. Because of the different minimum pT require-
ments for K 0S and Λ particles, the variation in hmT i includes both h pT i and particle mass differ-
ences. The anticorrelation of Rinv and hmT i has been interpreted as indicating the presence of
an expanding source [1].
Table 2 also includes the strong interaction scattering parameters d0 , < f 0 , and = f 0 obtained
from the ΛK 0S and ΛΛ correlations. Figure 5 shows d0 and = f 0 versus < f 0 in the left and
right panels, respectively, with the current results shown as red stars and squares for ΛK 0S and
ΛΛ, respectively. The displayed uncertainties are one-dimensional and are not based on a
two-dimensional contour.
Table 2: Extracted values of the Rinv , < f 0 , = f 0 , d0 , λ, and hmT i parameters from the K 0S K 0S , ΛK 0S ,
and ΛΛ combinations in the 0–80% centrality range. The first and second uncertainties are
statistical and systematic, respectively.

Parameter K 0S K 0S ΛK 0S ΛΛ
Rinv (fm) 3.30 ± 0.10 ± 0.37 2.1+ 1.4
−0.5 ± 0.7 1.3+ 0.4
−0.2 ± 0.3
< f 0 (fm) — −0.76+ 0.29
−0.19 ± 0.21 0.74+ 0.59
−0.16 ± 0.34
= f 0 (fm) — −0.07+ 0.48
−0.11 ± 0.32 —
d0 (fm) — 2.3+ 0.7
−0.5 ± 1.3 4.2+ 5.7
−2.1 ± 2.9
λ 0.38 ± 0.02 ± 0.08 0.34+ 0.41
−0.12 ± 0.16 1.5+ 1.2
−1.1 ± 1.4
hmT i (GeV) 1.53 2.09 2.60
 11

CMS CMS
15 1
AA collisions AA collisions
10
CMS (5.02 TeV): ΛK0S
5 0.5 ALICE (2.76 TeV): ΛK0S
d0 (fm)

ℑ f0 (fm)
0

−5 CMS (5.02 TeV): ΛΛ 0

CMS (5.02 TeV): ΛKS
0
PRC 91, 024916: ΛΛ
−10 STAR (200 GeV): ΛΛ PRC 66, 024007: ΛΛ
ALICE (2.76 TeV): ΛK0S NPA 707, 491: ΛΛ
−15 −0.5
−2 −1.5 −1 −0.5 0 0.5 1 1.5 2 −1.5 −1 −0.5 0 0.5
ℜ f0 (fm) ℜ f0 (fm)
Figure 5: The measured values of d0 versus < f 0 (left) and = f 0 versus < f 0 (right) from this
analysis along with other measurements and predictions as described in the text. For each
data point, the lines and the boxes indicate the (one-dimensional) statistical and systematic
uncertainties, respectively.

The negative value of < f 0 observed for the ΛK 0S correlations, combined with its relatively small
magnitude, suggests a repulsive ΛK 0S interaction. The uncertainty associated with the = f 0
value for the ΛK 0S correlations prevents any claim concerning possible inelastic processes. The
value of < f 0 found for ΛK 0S correlations differs from that reported
√ by the ALICE Collaboration
(teal diamonds) [15], which is also for PbPb collisions but at sNN = 2.76 TeV. The uncertainties
are too large to determine if d0 and = f 0 also differ between the two results.
The positive < f 0 value obtained for the ΛΛ correlations suggests an attractive interaction
that is not strong enough to produce a bound state such as the H-dibaryon [9, 48]. This re-
sult
√ disagrees with the finding from the STAR Collaboration in gold-gold (AuAu) collisions at
sNN = 200 GeV (blue circle). The negative < f 0 value of −1.10 ± 0.37 (stat)+ 0.68
−0.08 (syst) fm found
by STAR, combined with its magnitude, imply a repulsive interaction. It is noted, however,
that a theoretical study of the STAR data which considers collective flow and feed-down effects
(shown as a shaded region at d0 ≈ 5 fm, < f 0 ≈ 0.9 fm) suggests that these data are consistent
with the ΛΛ interaction being attractive [9]. An exclusion plot by the ALICE Collaboration
√ the ΛΛ scattering parameters obtained using
for √ the ΛΛ correlations from pp collisions at
s = 7 and 13 TeV, as well as pPb collisions at sNN = 5.02 TeV, also suggests an attractive
interaction [10]. In addition, our results are consistent with two theoretical calculations (black
triangles) that reproduce the ΛΛ binding energy of ΛΛ6 He, as extracted from the NAGARA
event [49, 50].

7 Summary
The K 0S K 0S , ΛK 0S , and ΛΛ femtoscopic correlations are
√ studied using lead-lead (PbPb) collision
data at a center-of-mass energy per nucleon pair of sNN = 5.02 TeV, collected by the CMS Col-
laboration. This is the first report on ΛΛ correlations in PbPb collisions at the CERN LHC. The
source size Rinv and the incoherence parameter λ were extracted for K 0S K 0S correlations in six
centrality bins covering the 0–60% range. The value of Rinv decreases from 4 to 1 fm going from
central to peripheral collisions and agrees with results from the ALICE Collaboration at a sim-
ilar transverse mass. Along with the Rinv and λ parameters, the strong interaction scattering
parameters, i.e., the complex scattering length and effective range, were extracted from ΛK 0S
and ΛΛ correlations in the 0–80% centrality range. These scattering parameters indicate that
12

the ΛK 0S interaction is repulsive and that the ΛΛ interaction is attractive. The scattering param-
eters obtained from ΛK 0S correlations differ from those reported by the ALICE Collaboration.
The positive real scattering length obtained from the ΛΛ correlation disfavors the existence
of a bound H-dibaryon state. The ΛΛ scattering parameters help to constrain baryon-baryon
and, more specifically, hyperon-hyperon interaction models. These measurements provide an
additional input to understand the nature of the strong interaction between pairs of strange
hadrons.

References
[1] M. A. Lisa, S. Pratt, R. Soltz, and U. Wiedemann, “Femtoscopy in relativistic heavy ion
collisions: Two decades of progress”, Ann. Rev. Nucl. Part. Sci. 55 (2005) 357,
doi:10.1146/annurev.nucl.55.090704.151533, arXiv:nucl-ex/0505014.

[2] V. G. J. Stoks, R. A. M. Klomp, M. C. M. Rentmeester, and J. J. de Swart, “Partial-wave

analysis of all nucleon-nucleon scattering data below 350 MeV”, Phys. Rev. C 48 (1993)
792, doi:10.1103/PhysRevC.48.792.

[3] J. J. de Swart and C. Dullemond, “Effective range theory and the low energy
hyperon-nucleon interactions”, Anna. Phys. 19 (1962) 458,
doi:10.1016/0003-4916(62)90185-9.

[4] R. Engelmann, H. Filthuth, V. Hepp, and E. Kluge, “Inelastic Σ − p-interactions at low

momenta”, Phys. Lett. 21 (1966) 587, doi:10.1016/0031-9163(66)91310-2.

[5] F. Eisele et al., “Elastic Σ ± p scattering at low energies”, Phys. Lett. B 37 (1971) 204,
doi:10.1016/0370-2693(71)90053-0.

[6] B. Sechi-Zorn, B. Kehoe, J. Twitty, and R. A. Burnstein, “Low-energy Λ-proton elastic

scattering”, Phys. Rev. 175 (1968) 1735, doi:10.1103/PhysRev.175.1735.

[7] CMS
√ Collaboration, “Bose–Einstein correlations in pp, pPb, and PbPb collisions at
sNN = 0.9–7 TeV”, Phys. Rev. C 97 (2018) 064912,
doi:10.1103/PhysRevC.97.064912, arXiv:1712.07198.

[8] J. Schaffner-Bielich, M. Hanauske, H. Stöcker, and W. Greiner, “Phase transition to

hyperon matter in neutron stars”, Phys. Rev. Lett. 89 (2002) 171101,
doi:10.1103/PhysRevLett.89.171101, arXiv:astro-ph/0005490.

[9] K. Morita, T. Furumoto, and A. Ohnishi, “ΛΛ interaction from relativistic heavy-ion
collisions”, Phys. Rev. C 91 (2015) 024916, doi:10.1103/PhysRevC.91.024916,
arXiv:1408.6682.

[10] ALICE Collaboration, “Study of the Λ-Λ interaction with femtoscopy correlations in pp
and pPb collisions at the LHC”, Phys. Lett. B 797 (2019) 134822,
doi:10.1016/j.physletb.2019.134822, arXiv:1905.07209.

[11] R. L. Jaffe, “Perhaps a stable dihyperon”, Phys. Rev. Lett. 38 (1977) 195,
doi:10.1103/PhysRevLett.38.195.

[12] H. Takahashi et al., “Observation of a ΛΛ6 He double hypernucleus”, Phys. Rev. Lett. 87
(2001) 212502, doi:10.1103/PhysRevLett.87.212502.
References 13

[13] K. Nakazawa and H. Takahashi, “Experimental study of double-Λ hypernuclei with

nuclear emulsion”, Prog. Theor. Phys. Supplement 185 (2010) 335,
doi:10.1143/PTPS.185.335.

[14] Belle Collaboration, “Search for an H-dibaryon with a mass near 2mΛ in Υ (1S) and
Υ (2S) decays”, Phys. Rev. Lett. 110 (2013) 222002,
doi:10.1103/PhysRevLett.110.222002, arXiv:1302.4028.
√
[15] ALICE Collaboration, “ΛK femtoscopy in Pb-Pb collisions at sNN = 2.76 TeV”, Phys.
Rev. C 103 (2021) 055201, doi:10.1103/PhysRevC.103.055201,
arXiv:2005.11124.

[16] C. Loizides, J. Kamin, and D. d’Enterria, “Improved Monte Carlo Glauber predictions at
present and future nuclear colliders”, Phys. Rev. C 97 (2018) 054910,
doi:10.1103/PhysRevC.97.054910, arXiv:1710.07098.

[17] R. Lednický and V. L. Lyuboshitz, “Final state interaction effect on pairing correlations
between particles with small relative momenta”, Sov. J. Nucl. Phys. 35 (1982) 770.

[18] HEPData record for this analysis, 2022. doi:10.17182/hepdata.133573.

[19] Tracker Group of the CMS Collaboration, “The CMS phase-1 pixel detector upgrade”,
JINST 16 (2021) P02027, doi:10.1088/1748-0221/16/02/P02027,
arXiv:2012.14304.

[20] CMS Collaboration, “Track impact parameter resolution for the full pseudorapidity
coverage in the 2017 dataset with the CMS phase-1 pixel detector”, CMS Detector
Performance Note CMS-DP-2020-049, 2020.

[21] CMS
√ Collaboration, “Performance of the CMS level-1 trigger in proton-proton collisions
at s = 13 TeV”, JINST 15 (2020) P10017, doi:10.1088/1748-0221/15/10/P10017,
arXiv:2006.10165.

[22] CMS Collaboration, “The CMS trigger system”, JINST 12 (2017) 01020,
doi:10.1088/1748-0221/12/01/P01020, arXiv:1609.02366.

[23] CMS Collaboration, “The CMS experiment at the CERN LHC”, JINST 3 (2008) S08004,
doi:10.1088/1748-0221/3/08/S08004.

[24] CMS
√ Collaboration, “Precision luminosity measurement in proton-proton collisions at
s = 13 TeV in 2015 and 2016 at CMS”, Eur. Phys. J. C 81 (2021) 800,
doi:10.1140/epjc/s10052-021-09538-2, arXiv:2104.01927.

[25] CMS
√ Collaboration, “CMS luminosity measurement using nucleus-nucleus collisions at
sNN = 5.02 TeV in 2018”, CMS Physics Analysis Summary CMS-PAS-LUM-18-001, 2022.

[26] CMS Collaboration,

√ “Charged-particle nuclear modification factors in PbPb and pPb
collisions at sNN = 5.02 TeV”, JHEP 04 (2017) 039, doi:10.1007/JHEP04(2017)039,
arXiv:1611.01664.

[27] CMS Collaboration, “Description and performance of track and primary-vertex

reconstruction with the CMS tracker”, JINST 9 (2014) P10009,
doi:10.1088/1748-0221/9/10/p10009, arXiv:1405.6569.
14

[28] CMS Collaboration, “Transverse-momentum

√ and pseudorapidity distributions of
charged hadrons in pp collisions at s = 0.9 and 2.36 TeV”, JHEP 02 (2010) 041,
doi:10.1007/JHEP02(2010)041, arXiv:1002.0621.
[29] CMS Collaboration,
√ “Observation and studies of jet quenching in PbPb collisions at
nucleon-nucleon sNN = 2.76 TeV”, Phys. Rev. C 84 (2011) 024906,
doi:10.1103/PhysRevC.84.024906, arXiv:1102.1957.
[30] C. Gale, S. Jeon, and B. Schenke, “Hydrodynamic modeling of heavy ion collisions”, Int.
J. Mod. Phys. A 28 (2013) 1340011, doi:10.1142/S0217751X13400113,
arXiv:1301.5893.
[31] GEANT4 Collaboration, “G EANT 4—a simulation toolkit”, Nucl. Instrum. Meth. A 506
(2003) 250, doi:10.1016/S0168-9002(03)01368-8.
[32] CMS Collaboration, “Strange hadron collectivity in pPb and PbPb collisions”, 2022.
arXiv:2205.00080.
√
[33] CMS Collaboration, “Strange particle production in pp collisions at s = 0.9 and 7 TeV”,
JHEP 05 (2011) 064, doi:10.1007/JHEP05(2011)064, arXiv:1102.4282.
[34] CMS Collaboration, “Long-range two-particle correlations of strange hadrons with
charged particles in pPb and PbPb collisions at LHC energies”, Phys. Lett. B 742 (2015)
200, doi:10.1016/j.physletb.2015.01.034, arXiv:1409.3392.
[35] Particle Data Group Collaboration, “Review of particle physics”, Prog. Theor. Exp. Phys.
2022 (2022) 083C01, doi:10.1093/ptep/ptac097.
[36] H. Voss, A. Höcker, J. Stelzer, and F. Tegenfeldt, “TMVA, the toolkit for multivariate data
analysis with ROOT”, in XIth International Workshop on Advanced Computing and Analysis
Techniques in Physics Research (ACAT), p. 40. 2007. arXiv:physics/0703039.
doi:10.22323/1.050.0040.
[37] G. I. Kopylov, “Like particle correlations as a tool to study the multiple production
mechanism”, Phys. Lett. B 50 (1974) 472, doi:10.1016/0370-2693(74)90263-9.
[38] A. Kisiel, “Non-identical particle correlation analysis in the presence of non-femtoscopic
correlations”, Acta Phys. Polon. B 48 (2017) 717, doi:10.5506/APhysPolB.48.717.
[39] ALICE Collaboration,
√ “pp, p-Λ, and Λ-Λ correlations studied via femtoscopy in pp
reactions at s = 7 TeV”, Phys. Rev. C 99 (2019) 024001,
doi:10.1103/PhysRevC.99.024001, arXiv:1805.12455.
[40] STAR
√ Collaboration, “Neutral kaon interferometry in Au+Au collisions at
sNN = 200 GeV”, Phys. Rev. C 74 (2006) 054902,
doi:10.1103/PhysRevC.74.054902, arXiv:nucl-ex/0608012.
[41] A. D. Martin and E. N. Ozmutlu, “Analyses of KK production and scalar mesons”, Nucl.
Phys. B 158 (1979) 520, doi:10.1016/0550-3213(79)90180-9.
[42] A. Antonelli, “Radiative φ decays”, eConfC 020620 (2002) THAT06,
doi:10.48550/ARXIV.HEP-EX/0209069, arXiv:hep-ex/0209069.
[43] N. N. Achasov and V. V. Gubin, “Analysis of the nature of the φ → γπη and φ → γπ 0 π 0
decays”, Phys. Rev. D 63 (2001) 094007, doi:10.1103/PhysRevD.63.094007,
arXiv:hep-ph/0101024.
References 15

[44] N. N. Achasov and A. V. Kiselev, “New analysis of the KLOE data on the φ → η π 0 γ
decay”, Phys. Rev. D 68 (2003) 014006, doi:10.1103/PhysRevD.68.014006,
arXiv:hep-ph/0212153.

[45] F. James and M. Roos, “Minuit: A system for function minimization and analysis of the
parameter errors and correlations”, Comput. Phys. Commun. 10 (1975) 343,
doi:10.1016/0010-4655(75)90039-9.

[46] PHENIX
√ Collaboration, “Lévy-stable two-pion Bose–Einstein correlations in
sNN = 200 GeV Au+Au collisions”, Phys. Rev. C 97 (2018) 064911,
doi:10.1103/PhysRevC.97.064911, arXiv:1709.05649.

[47] ALICE Collaboration,

√ “One-dimensional pion, kaon, and proton femtoscopy in Pb-Pb
collisions at sNN = 2.76 TeV”, Phys. Rev. C 92 (2015) 054908,
doi:10.1103/PhysRevC.92.054908, arXiv:1506.07884.
√
[48] STAR Collaboration, “ΛΛ correlation function in Au+Au collisions at sNN = 200 GeV”,
Phys. Rev. Lett. 114 (2015) 022301, doi:10.1103/PhysRevLett.114.022301,
arXiv:1408.4360.

[49] E. Hiyama et al., “Four-body cluster structure of A = 7–10 double-Λ hypernuclei”, Phys.
Rev. C 66 (2002) 024007, doi:10.1103/physrevc.66.024007,
arXiv:nucl-th/0204059.

[50] I. N. Filikhin and A. Gal, “Faddeev–Yakubovsky calculations for light ΛΛ hypernuclei”,

Nucl. Phys. A 707 (2002) 491, doi:10.1016/s0375-9474(02)01008-4,
arXiv:nucl-th/0203036.