arX

iv:1

010.

5616

v1 [

nucl

-ex]

27

Oct

201

0

Study of some two-body non-mesonic decays

of 4ΛHe and 5

ΛHe

FINUDA Collaboration

M. Agnello a,b, L. Benussi c, M. Bertani c, H.C. Bhang d,

G. Bonomi e,f, E. Botta g,b, M. Bregant h, T. Bressani g,b,S. Bufalino b, L. Busso k,b, D. Calvo b, P. Camerini i,j,

B. Dalena ℓ, F. De Mori g,b, G. D’Erasmom,n, F.L. Fabbri c,

A. Feliciello b, A. Filippi b,1, E.M. Fiorem,n, A. Fontana f,H. Fujioka o, P. Genova f, P. Gianotti c, N. Grion j,

V. Lucherini c, S. Marcello g,b, N. Mirfakhrai p, F. Moia e,f,O. Morra q,b, T. Nagae o, H. Outa r, A. Pantaleo n,2,

V. Paticchio n, S. Piano j, R. Rui i,j, G. Simonettim,n,R. Wheadon b, A. Zenoni e,f

aDipartimento di Fisica, Politecnico di Torino, Corso Duca degli Abruzzi 24,

Torino, Italy

bINFN Sezione di Torino, via P. Giuria 1, Torino, Italy

cLaboratori Nazionali di Frascati dell’INFN, via. E. Fermi, 40, Frascati, Italy

dDepartment of Physics, Seoul National University, 151-742 Seoul, South Korea

eDipartimento di Ingegneria Meccanica e Industriale, Universita di Brescia, via

Branze 38, Brescia, Italy

fINFN Sezione di Pavia, via Bassi 6, Pavia, Italy

gDipartimento di Fisica Sperimentale, Universita di Torino, via P. Giuria 1,

Torino, Italy

hSUBATECH, Ecole des Mines de Nantes, Universite de Nantes, CNRS-IN2P3,

Nantes, France

iDipartimento di Fisica, Universita di Trieste, via Valerio 2, Trieste, Italy

jINFN Sezione di Trieste, via Valerio 2, Trieste, Italy

kDipartimento di Fisica Generale, Universita di Torino, via P. Giuria 1, Torino,

Italy

ℓCERN, CH-1211 Geneva 23, Switzerland

mDipartimento di Fisica Universita di Bari, via Amendola 173, Bari, Italy

nINFN Sezione di Bari, via Amendola 173, Bari, Italy

oDepartment of Physics, Kyoto University, Kitashirakawa, Kyoto 606-8502 Japan

pDepartment of Physics, Shahid Behesty University, 19834 Teheran, Iran

Preprint submitted to Elsevier Preprint 28 October 2010

qINAF-IFSI, Sezione di Torino, Corso Fiume 4, Torino, Italy

rRIKEN, Wako, Saitama 351-0198, Japan

Abstract

The Non-Mesonic (NM) decay of 4ΛHe and 5

ΛHe in two-body channels has beenstudied by the FINUDA Collaboration. Two-body NM decays of hypernuclei are rareand the existing observations and theoretical calculations are scarce. The 4

ΛHe →

d+ d, p+ t decay channels simultaneously observed by FINUDA on several nucleiare compared: the pt channel is the dominant one. The decay yields for the twodecay channels are assessed for the first time: they are (1.37 ± 0.37) × 10−5/K−

stop

and (8.3 ± 1.0) × 10−5/K−

stop, respectively. Due to the capability of FINUDA ofidentifying 5

ΛHe hypernuclei, a few5ΛHe → d+t decay events have also been observed.

The branching ratio for this decay channel has been measured for the first time:(3.0 ± 2.3)× 10−3.

Key words: Light Hypernuclei; Non Mesonic Weak DecayPACS: 21.80.+a, 25.80.Pw

1 Introduction

Two weak decay modes of Λ Hypernuclei are known, the mesonic and theNon-Mesonic (NM) one. The latter gives a unique tool to study the baryon-baryon weak interactions. Relatively little information about NM decays canbe inferred by alternative processes, such as nucleon scattering experimentsor parity violating nuclear transitions.

The NM decay mode, in which two or more nucleons are emitted after a Λ-Amultibaryon weak interaction in a nuclear medium, is dominant for medium-heavy nuclei. It is characterized by a large momentum transfer; in fact, theemitted nucleons have a momentum in the range 400-600 MeV/c and canescape the nucleus. In light hypernuclei the NM decay can also lead to finalstates composed by two bodies only:

4ΛHe→ d+ d (1)4ΛHe→ p+ t (2)

1 corresponding author. E-mail: filippi@to.infn.it; Phone: +39.011.6707323; Fax:+39.011.6707324.2 Deceased

2

4ΛHe→n+ 3He (3)5ΛHe→ d+ t. (4)

Some of them can be interpreted as two-nucleon induced decays. FINUDA re-cently determined the two-nucleon induced NM decay to be about 1/4 of thetotal NM one, for p-shell hypernuclei [1]. In addition, a suppressed branchingratio for the two-body NM channels is expected because of the large momen-tum transfer and of the possible two-step mechanisms involved.

The only existing calculation for the two body 4ΛHe decay rates was performed

by Rayet [2], which is based on a phenomenological non-relativistic matrixelements evaluation for the ΛN → N + N interaction. Branching ratios of0.015 for (1) and (2) and of 0.03 for (3) were predicted, with a spread withina 1.5 factor due to the dependence of the model on the nuclear density andon the Λ compression effect.

These evaluations are in rough agreement with the existing experimental ob-servations, which however are scarce and dated. They all belong to bubblechamber and emulsion experiments [3,4,5]. The full present database consistsof a few 4

ΛHe → n +3 He events [3,4], whose rate is 8-14% of all the 4ΛHe NM

decays, and one 4ΛHe → d + d event [4]. No 4

ΛHe → p + t decays were everobserved. Keyes et al. [5], in a liquid Helium bubble chamber experiment, re-port 1.8% as upper limit for the reactions (1) and (2) as compared to the π−

mesonic decay of 4ΛHe. For the 5

ΛHe → d + t decay only one event was everreported [3], and only one theoretical evaluation for the expected decay ratesof 5

ΛHe exists [6].

In this Letter the rates of the decay modes (1) and (2) for 4ΛHe hyperfragments

and (4) for 5ΛHe are presented. The data were collected by the FINUDA spec-

trometer operating at the DAΦNE φ-factory, Laboratori Nazionali di Frascati(LNF), Italy. FINUDA is a magnetic spectrometer designed for the study ofhypernuclei production and decay being induced by stopped negative kaonson targets of different composition. The apparatus features a large geometricalacceptance (∼ 2π sr) and an outstanding particle identification capability forcharged hadrons (98% and 94% for proton and deuteron, respectively). Dueto the apparatus transparency an excellent momentum resolution is achieved,which is ∆p/p ∼ 0.6% for negative pions of about 250 MeV/c, ∼ 1% for 500MeV/c protons, ∼ 2% for 500 MeV/c deuterons. FINUDA cannot give any in-formation about reaction (3), since 507 MeV/c 3He nuclei cannot be detectedby the apparatus.

The (1), (2) and (4) two-body NM decays of Hyperhelium isotopes present aclear back-to-back topology for the reconstructed tracks. Such a feature, alongwith the PID information, makes the full reconstruction of the hypernucleardecay products feasible.

3

FINUDA can directly identify 5ΛHe hypernuclei end indirectly detect 4

ΛHe hy-perfragments. The identification of the 5

ΛHe hypernucleus is performed by mea-suring negative pions from the K− 6Li → 6

ΛLi + π− and K− 7Li → 7ΛLi

∗

+ π−

reactions. The unstable hypernuclei decay strongly to 5ΛHe according to the

reactions 6ΛLi →

5ΛHe + p and 7

ΛLi∗→ 5

ΛHe + d. In these reactions, the protonand the deuteron momentum is below 100 MeV/c, which precludes their detec-tion. However, the formation pions have a momentum in a well defined rangewhich allows for a clear 5

ΛHe tagging. Conversely, the FINUDA spectrome-ter is unable to examine the K−

stop4He → 4

ΛHe + π− reaction, since a liquidHelium target cannot be arranged in it. However, 4

ΛHe studies are possiblevia 4

ΛHe hyperfragments which result from the K−

stopA interaction, for A ≥ 6.In solid targets the 4

ΛHe hyperfragment cannot escape the target volume andthus be tracked, as it was done in emulsion experiments [7,8,9]. In these ex-periments, the measured yield of hyperfragment production in K− absorptionat rest ranged from (4.5± 0.5)% [10] to the more recent values (5 ± 1)% [11]and (6.5 ± 0.2)% [7]. The fraction of the hyperfragment mesonic decay wasmeasured to be around 20%. Hyperfragments were in general observed morecopiously when produced by light emulsion nuclei, i.e up to 16O [11].

The signature of two-body 4ΛHe decays is particularly clean and simple. For

reaction (1), it consists of two 571.8 MeV/c monochromatic back-to-backdeuterons, for reaction (2), the two hadrons (p, t) are back-to-back emittedwith a momentum of 508 MeV/c.

2 Outline of the FINUDA experimental apparatus

A short description of the experimental set-up is given here for the sake ofclarity. FINUDA features a cylindrical geometry and is installed in one of theDAΦNE interaction regions, where φ(1020) mesons from the e+e− collisionsare produced. The charged kaons from φ(1020) → K++K− decay (B.R.=0.49)have a momentum of 127 MeV/c at most, and they slow down while cross-ing the internal region of the apparatus until they interact at rest in a setof targets. The apparatus consists of five position-sensitive layers, arrangedcoaxially around the beam axis. Four of them are also used for particle identi-fication through energy loss measurements. The tracking region is immersed ina uniform solenoidal magnetic field of 1 T. Three main sectors may be singledout in the detector layout:

• interaction/target region: located at the apparatus center, consisting of aBeryllium beam pipe, a 12 scintillator slab hodoscope (TOFINO) [12] usedfor trigger purposes and for charged kaons discrimination, an eight modulearray of double-side Si microstrip detectors (ISIM) [13] facing eight targettiles. The target set-up for the data used in the present analysis consisted of

4

two (90% enriched) 6Li (thickness: 4 mm), two natural isotopic composition7Li (4 mm), two 9Be (2 mm), one 13C (10 mm) and one D2O (liquid filledand mylar walled, 3 mm thick) targets.

• tracking region: consisting of ten Si microstrip modules (OSIM) [13] facingexternally the targets, two arrays of eight planar low-mass drift chambers(LMDC’s) [14], filled with a He-iC4H10 gas mixture, and a system of six Ar-C2H6 filled straw tube longitudinal-stereo layers [15]. The ISIM and OSIMmodules feature a spatial resolution better than σ ∼ 30 µm for both the(rφ) and z coordinates. LMDC’s provide a resolution σrφ ∼ 150 µm andσz ∼ 1 cm, while for the straw tubes system the resolution is σrφ ∼ 150 µmand σz ∼ 500 µm.

• outer scintillator array: a barrel of 72 thick slabs [16], used for first leveltrigger, time-of-flight measurements with an overall time resolution σ ∼ 800ps and neutron identification with a ∼ 10% efficiency.

The data used in the present analysis correspond to an integrated e+e− lumi-nosity of 966 pb−1, collected by FINUDA in the 2006-2007 period.

3 4ΛHe → d+ d and 4

ΛHe → p+ t decays

The 4ΛHe hyperfragment can be produced via the reaction K−

stopAZX → 4

ΛHe+π− +A−4

Z−2 X′, where X ′ is the recoiling nuclear system (bound or unbound),

and the pion momentum is larger than 220 MeV/c. The hyperfragment pro-duction can also occur on protons with the emission of a π0; however, this isundetectable by FINUDA.

The signature of a dd decay event consists of two high momentum back-to-back deuteron tracks. The main source of background is given by the 4

ΛHe →(pn)+(pn)π0 mesonic decay, whose frequency is almost comparable to the fullNM branch [17]. The mean momentum of the (pn) pairs in this three-bodydecay is around 150 MeV/c; therefore, these events can be easily discarded byapplying a cut on the missing mass distribution of the 4

ΛHe → d+ d at rest

decay.

The total inclusive dd raw collected sample consists of 272 ± 16 events, overall the FINUDA targets. Fig. 1 shows the momentum distribution of the twoobserved deuterons. In a pair, one of the deuterons is emitted from the targettoward the outer hemisphere, i.e. toward the tracking region: forward track,Fig. 1a). The second deuteron is emitted in the opposite direction, thus cross-ing twice ISIM, TOFINO, the beam pipe and one of the targets: backwardtrack, Fig. 1b). The forward tracks are required to be reconstructed by a mini-mum of three hits, one of which is necessarily located on the OSIM array. Themomentum resolution of backward deuterons is spoilt by the larger material

5

budget crossed by the particle, but at least two hits of the track are located onthe high resolution ISIM and OSIM detectors. From Montecarlo simulationsthe momentum resolution of forward deuterons in FINUDA is 3% FWHM (17MeV/c for 570 MeV/c particles), while for backward deuterons is 4% FWHM(corresponding to about 22 MeV/c). Both tracks were selected with a trackfitting procedure which required χ2 < 20 (C.L. ∼ 95%).

GeV/c0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8

entr

ies/

(7 M

eV/c

)

0

5

10

15

20

25

a)

forward deuteron momentum

GeV/c0.4 0.45 0.5 0.55 0.6 0.65 0.7 0.75 0.8

entr

ies/

(9 M

eV/c

)0

2

4

6

8

10

12

14

16

18

20

22

b)

backward deuteron momentum

Fig. 1. Momentum distributions for the forward a) and the backward b) deuteronsin the two deuteron semi-inclusive sample. The hatched histograms are fed by4ΛHe → d + d at rest decay events. The black histograms correspond to dd eventswith a high momentum π− (pπ− > 220 MeV/c) in coincidence.

The semi-inclusive momentum distributions are shown as open histograms inFig. 1. Fig. 2a) shows the distribution of the angle between the two deuterontracks, while Fig. 2b) displays the invariant mass distribution of the dd pairs.Events with a sharp back-to-back angular correlation (cosΘ < −0.995) arethen chosen. This cut eliminates possible in-flight 4

ΛHe → d+ d decays as wellas dd non-monochromatic pairs emitted in possible heavier hyperfragmentdecays (selection not shown in the pictures).

Θcos-1 -0.8 -0.6 -0.4 -0.2 0 0.2

entr

ies/

(0.2

)

0

10

20

30

40

50

60

70

cos(angle between deuterons)

a)

cos(angle between deuterons)

2GeV/c3.86 3.88 3.9 3.92 3.94 3.96 3.98 4 4.02 4.04

)2en

trie

s/(3

MeV

/c

0

2

4

6

8

10

12

14

16

(dd) invariant mass

b)

(dd) invariant mass

Fig. 2. Distribution of the angle between the two deuterons (a) and of the (dd) invari-ant mass (b). Open histogram: semi-inclusive dd sample. The hatched histogramsare fed by 4

ΛHe → d + d at rest decay events. The black histograms correspond todd events with a high momentum π− (pπ− > 220 MeV/c) in coincidence.

6

In the surviving dd sample, events are then selected with a total energy in therange 3-4 GeV and a total momentum less than 50 MeV/c. A last selectionrequires the missing mass for the 4

ΛHe → d + d decay at rest to lie in a 2σwindow (σ = 11 MeV/c2) across to the 4

ΛHe binding energy (−2.39 MeV). Thefinal selected events produce the cross-hatched distributions in Fig. 1 and 2.

By summing over all the targets, the total statistics available after the de-scribed selections is 31 ± 6 dd events. 14 ± 4 of them present an additionalπ− with momentum larger than 220 MeV/c. These events feed the black his-tograms in Fig. 1 and 2. They can be interpreted as exclusive events in whichboth the 4

ΛHe hyperfragment formation and its decay have been measured.The selected sample is affected by a background contribution consistent withzero. Side bin evaluations have been made selecting different missing massand momentum slices displaced one or two σ’s from the respective expectedvalues; in fact, no events were found matching the required event signature.

A target-by-target measurement of the number of 4ΛHe → d+ d decays at rest

per incident K−

stop can be performed. The results are reported, for differentnuclei, in Tab. 1. This yield is basically constant for lighter targets (up to9Be); for heavier targets, with just one selected event, an upper limit onlyis given at 90% confidence level. The systematic errors take into account thespread of the values measured in different targets of the same compositionand the effect of slight variations in the selection criteria. They also take intoaccount the uncertainties in assessing the number of stopped kaons (due to outof target and in-flight interactions, K−/K+ swap in the pattern recognitionprocedure, backtracking algorithm inefficiencies).

target dd Events Yield ×10−5/(K−

stop) pt Events Yield ×10−5/(K−

stop)

6Li 12± 3 3.0 ± 1.3stat ± 0.9sys 1± 1 < 16.8 (90%C.L.)

7Li 7± 3 2.4 ± 1.3stat ± 0.8sys 1± 1 < 14.3 (90%C.L.)

9Be 10± 3 3.3 ± 1.4stat ± 0.4sys 5± 2 14.9 ± 3.1stat ± 0.9sys

13C 1± 1 < 2.3 (90%C.L.) 1± 1 < 30.5 (90%C.L.)

16O 1± 1 < 2.7 (90%C.L.) 2± 1 10.4 ± 2.0stat ± 0.2sysTable 1Second and third column, number of events and yields, perK−

stop, of the4ΛHe → d+d

decay, for hyperfragments produced in targets of different nuclear composition (firstcolumn). Fourth and fifth column, number of events and yields per K−

stop of the4ΛHe → p+ t decay at rest.

An overall average yield can be deduced using, for the entries with n = 1 inTab. 1, the actual value with a statistical error as given by a Gaussian prop-agation. The resulting value is Y4

ΛHe→d+d = (1.37± 0.37)× 10−5/K−

stop, wherethe overall uncertainty accounts for statistical and systematic errors addedin quadrature. Assuming a hyperfragment production rate of about 5%/K−

stop

7

[11], and under the simple hypothesis that 4ΛHe’s are the most abundantly

produced hyperfragments, this yield corresponds to an upper limit for the4ΛHe → d+ d branching ratio of about 3× 10−4.

Concerning the 4ΛHe → p+t decay channel, tritons in FINUDA can be detected

only for momenta larger than 550 MeV/c, which exceeds the value expectedfor the two body 4

ΛHe → p + t decay at rest, 508 MeV/c. Tritons with lowermomenta cannot be reconstructed as they stop in the inner tracking layersof FINUDA. However, if they cross the I/OSIM layers they deposit a largeamount of energy, that can be used to tag such an event. As a consequence, the4ΛHe → p+ t decay can effectively be measured only through the informationon the proton and the negative formation pion momentum. The proton trackis selected with the same quality criteria described above for deuterons. Forthe pion the quality criteria are released, and only a loose cut is applied toeliminate contributions from Quasi-Free (QF) reactions due to the absorptionof the K− by just one nucleon, pπ− > 200 MeV/c. Events with a detectedneutron are rejected to discard possible contaminations due to baryon (Λ, Σ−)decays. A total sample of 297023±545 semi-inclusive (pπ−) events is collected.None of the protons selected in the proper momentum band (498-520 MeV/c)bears hits with large energy deposit on opposite silicon layers.

GeV/c0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65

entr

ies/

(3 M

eV/c

)

0

5

10

15

20

25 a)

coincidence-πproton momentum,

2GeV/c2.75 2.8 2.85 2.9 2.95

)2en

trie

s/(1

MeV

/c

0

5

10

15

20

25 b)

X’ missing massZ-2A-4 +

miss + p + t-π →X Z

A+ -K

Fig. 3. a) Open histogram: momentum distribution of a proton in events with a coin-cident π− with momentum larger than 200 MeV/c; the shadowed area correspondsto events for which the missing mass distribution of theK− A

ZX → π−ptmiss+A−4Z−2X

′

reaction matches the mass of a triton ±5 MeV/c2 (∼ 2σ wide interval). The linecorresponds to the nominal p momentum for the decay at rest 4

ΛHe → p+t. b) Miss-ing mass distribution of the 4

ΛHe → p+ t decay at rest for events featuring a protonand a π− with momentum larger than 200 MeV/c; the shadowed area correspondsto events with a missing mass of a triton ±5 MeV/c2 (line in the figure).

The main background source of the 4ΛHe → p + t decay channel is given

by the QF K−(np) → Σ− + p two nucleon absorption (K2N), which has alarge capture rate, 1.62%/K−

stop in 6Li [18], and an order of magnitude largerin heavier nuclei. The QF signature is similar to the studied decay, with a“quasi”-monochromatic proton of about 510 MeV/c momentum, and a fast

8

π−.

A study of the impact parameters and the angular distributions of the emittedproton and negative pion has been performed to remove the most sizeable frac-tion of K2N -QF events without loosing an excessive amount of the searchedsignal. An optimized cut in the distribution of the π− impact parameter, i.e.the distance between the track and the K− interaction vertex, rejects about79% of the Σ−p background, but it also reduces by 60% the searched signal.

The pπ− angle in the K−2N → Σ−+ p absorption displays a marked back-to-back trend. In the case of hypernuclear formation and decay one should expecta relatively enhanced emission of tracks on the same side with respect to thetarget; however, the selection of tracks emitted in the same hemisphere alsoenhances the contamination of Λ’s. A proper cut on the pπ− invariant masscan however remove this contribution. The selection of pπ− pairs emittedin the same hemisphere reduces the available sample of 53%, and the Σ−pcontamination is finally dropped to 12%. This criterion is therefore chosen.Then, a final target-by-target missing mass cut is applied to pick out eventsfeaturing a missing triton mass ± 5 MeV/c2.

The events surviving the mentioned selections fill the shadowed histogram ofFig. 3 a) and b). The open histogram in Fig. 3b) presents the missing massdistribution of the K− A

ZX → π− + d + tmiss +A−5Z−2 X

′ reaction, where X ′ istaken at rest. The line indicates the mass of the 3H nucleus, 2.82 GeV/c2.The selected events are compatible with this signal within the missing massresolution, 6 MeV/c2 at FWHM. The proton momentum for the selected eventsis shown in Fig. 3 a); this is a 15 MeV/c wide peak at about 508 MeV/c. Thisproton has a resolution of 0.7% (σ).

The final (still background inclusive) selected sample amounts to a total of44 ± 7 ptmissπ

− events, which belong to all the available targets. The finalcontamination on the selected sample due to the Σ−p source consists of afew 10−5/K−

stop. In the final sample the signal to background ratio amountsto S/N = 0.23, which corresponds to a 2.9σ statistical significance of theobserved signal. After background subtraction, an evaluation of the yield ofthe 4

ΛHe → p+ t decay at rest is done and the results are reported in columns5 and 6 of Tab. 1. In case of a single selected event only an upper limit, at 90%C.L., is given. The systematic uncertainties take into account the maximumspread of values obtained with small variation of the selection cuts and intargets of the same nuclear composition. They also take into account theuncertainty on the kaon normalization described in the previous section. The4ΛHe → p+ t decay at rest is favoured in heavier targets. Following the methoddescribed above for dd decays, the average value (8.3 ± 1.0) × 10−5/K−

stop isobtained. Assuming again a hyperfragment production of about 5%/K−

stop, anupper limit for the branching ratio of the 4

ΛHe → p+ t decay is assessed to be

9

∼ 1.7× 10−3.

The ratio between the decay yields in the dd and pt channels is also evaluatedtarget by target. The trend of the ratio as a function of A is reported in Fig. 4,in which the error bars take into account both statistic and systematic errors.The average mean value R(4ΛHe → d + d/4ΛHe → p + t) is (0.11 ± 0.04). In

A6 8 10 12 14 16

R(d

d/pt

)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Fig. 4. Ratio of the 4ΛHe → d + d to 4

ΛHe → p + t decay yields as a function of theatomic mass number A.

spite of the large errors, a clear dominance of the pt decay mode over the ddarises, in particular for nuclei with small A.

4 5ΛHe production and 5

ΛHe → d+ t decay in FINUDA

In FINUDA 5ΛHe hypernuclei may be produced via the K−

stop6,7Li interaction.

They can cleanly be observed by FINUDA in the formation π− spectra byrequiring a proton from the non-mesonic decay of the hypernucleus in coinci-dence [19].

The data set for the present analysis is obtained by selecting a negative pionin the momentum range 267-273 MeV/c along with a deuteron in the 570-630MeV/c momentum interval. Such a deuteron, coming from the 5

ΛHe → d + tdecay, has a well-defined momentum of 597 MeV/c. This precise selectionprevents further tritons from being detected. However, events with a largeenergy release on ISIM or OSIM modules have been found. One event withall the above features has been reconstructed for 6Li and two for 7Li. Nobackground events presenting the same topology and momenta right outsidethe above mentioned momentum ranges have been found.

With these figures the yield Y5

ΛHe→d+t = (2.6 ± 1.5st ± 1.1sys) × 10−5/K−

stop,which bears statistical significance 1.7σ, is found. Normalizing to the absolute

10

number of 5ΛHe hypernuclei [19], the branching ratio for this decay channel

is B.R.(5ΛHe → d+ t) = (3.0 ± 2.3) × 10−3. On the same data sample, anassessment of the one-proton induced NM decay rate (inclusive of FSI dis-tortion effects) has been determined, R(5ΛHe) = 0.28 ± 0.07 [19]. Taking intoaccount the missing terms due to one-neutron and two-nucleon induced de-cays, the measured dt branching ratio, about two orders of magnitude smaller,is roughly in agreement with the theoretical predictions [2]. The same conclu-sion can also be attained by deducing the NM branching ratio from the totaland mesonic widths of the 5

ΛHe decays [20].

Following the same analysis pattern, a search of the dtmiss decay of 5ΛHe hyper-

fragments is performed. A sample of 190±14 dtmissπ− events, still background

comprehensive, out of a total of 1056 ± 32 dπ− inclusive ones, is collected.Such dtmissπ

− events rely on a high-resolution π− (∆p/p ∼ 0.75%) in co-incidence with a deuteron in the required momentum range. A stringent χ2

from of the track fitting procedure is required, corresponding to C.L. ∼ 98%.The resolution for the deuteron momentum is 8 MeV/c (σ), corresponding to∆p/p ∼ 1.5%. The K− A

ZX → π− + d + tmiss +A−5Z−2 X

′ events, whose missingmass belongs to the interval [triton mass ± 12.5 MeV/c2], corresponding to1.2σ, are further analysized. Fig. 5 shows the deuteron momentum (a) and themissing mass spectrum (b) for the full dπ− sample (open histogram) and theselected events (shadowed areas). The background contribution in the π−dtmiss

GeV/c0.2 0.4 0.6 0.8 1 1.2

entr

ies/

(8 M

eV/c

)

0

2

4

6

8

10

a)

coincidence-πdeuteron momentum,

2GeV/c2.4 2.5 2.6 2.7 2.8 2.9 3

)2en

trie

s/(4

MeV

/c

0

1

2

3

4

5

6

7

8

9

b)

X’ missing massZ-2A-5 +

miss + d + t-π →X Z

A+ -K

Fig. 5. a) Open histogram: deuteron momentum distribution in coincidence with aπ− having a momentum larger than 200 MeV/c. The shadowed area corresponds toevents for which the K− A

ZX → π− + d+ tmiss +A−5Z−2 X

′ missing mass is within themass of a triton ±12.5 MeV/c2. The line indicates the d momentum for the decayat rest 5

ΛHe → d+ t. b) Missing mass distribution for the 5ΛHe → d+ t decay at rest.

The shadowed area corresponds to the events for which the missing mass is withinthe mass of a triton (line in the figure) ±12.5 MeV/c2.

sample is sizeable and mostly due to the K2N → Σ− + p QF reaction, wherethe p is misidentified as d. From side bin evaluation the S/N ratio is about20%, being averaged over all the available targets. Tab. 2 reports the decayrate per K−

stop of5ΛHe hyperfragments to the dt channel. The systematic uncer-

11

tainty takes into account the spread of the values due to small changes in theselection criteria, the different counting in targets of the same nucleus type,as well as the varying factors in the kaon normalization.

target Number of events Yield×10−4/(K−

stop)

6Li 4± 2 1.83 ± 0.93stat ± 0.12sys

7Li 5± 2 1.12 ± 0.51stat ± 0.08sys

9Be 13 ± 4 1.23 ± 0.38stat ± 0.02sys

13C 7± 3 2.25 ± 0.87stat ± 0.04sys

16O 11 ± 3 1.58 ± 0.50stat ± 0.03sysTable 2Yield of the 5

ΛHe → d+ tmiss decay at rest per K−

stop.

The yield averaged over the available nuclei is (1.40± 0.24)× 10−4/K−

stop.

5 Summary

FINUDA has been able to measure several features of the two-body decaychannels 4

ΛHe → d + d, 4ΛHe → p + t an 5

ΛHe → d + t, thus conveying newresults useful to complete the meager existing database. In spite of the limitedstatistics, FINUDA has observed very clean signatures of decay events with areduced background in most of the studied channels. The particle identificationcapabilities of the magnetic spectrometer together with its high resolutionallow for the detection of a few rare events never observed before.

For 4ΛHe → d + d in different targets the decay yields are of the order of

10−4/K−

stop, while the pt decay channel is about one order of magnitude larger.The 5

ΛHe → d + t decay yield of 5ΛHe hypernuclei is (2.6 ± 1.9)× 10−5/K−

stop.The B.R. for this decay channel has been determined for the first time tobe (3.0 ± 2.3) × 10−3. This value is in rough agreement with the theoreticalprediction [2] and a factor 100 lower than the B.R. of the “standard” NMdecay [19].

References

[1] FINUDA Collaboration, M. Agnello et al., Phys. Lett. B685 (2010), 247

[2] M. Rayet, Nuovo Cim. 52B (1966), 238

[3] G. Coremans et al., Nucl. Phys. B16 (1970), 209

12

[4] M.M. Block et al., Proc. 1960 Ann. Intl. Conf. on High Energy Physics at

Rochester (New York, 1960), p. 419

[5] G. Keyes et al., Nuovo Cim. 31A (1976), 401

[6] P.G. Thurnauer, Nuovo Cim. 26 (1962), 869

[7] J. Lemonne et al., Nuovo Cim. 34 (1964), 529

[8] M.W. Holland, Nuovo Cim. 32 (1964), 48

[9] P.E. Schlein, W.E. Slater, Nuovo Cim. 21 (1961), 213

[10] J. Sacton, Nuovo Cim. 18 (1960), 266

[11] D.H. Davis et al., Nuovo Cim. 22 (1961), 275

[12] B. Dalena et al., Nucl. Instr. Methods A603 (2009), 276

[13] P. Bottan et al., Nucl. Instr. Methods A427 (1999), 423

[14] M. Agnello et al., Nucl. Instr. Methods A385 (1997), 58

[15] L. Benussi et al., Nucl. Instr. Methods A361 (1995), 180L. Benussi et al., Nucl. Instr. Methods A419 (1998), 648

[16] A. Pantaleo et al., Nucl. Instr. Methods A543 (2005), 593

[17] J.G. Fetkovich et al., Phys. Rev. D6 (1972), 3069

[18] FINUDA Collaboration, M. Agnello et al., Nucl. Phys. A775 (2006), 35

[19] FINUDA Collaboration, M. Agnello et al., Nucl. Phys. A804 (2008), 151

[20] S. Kameoka et al., Nucl. Phys A754 (2005), 173cS. Okada et al., Nucl. Phys A754 (2005), 178c

13