Observation of the doubly charmed baryon Xicc++ - ?· Observation of the doubly charmed baryon ++ cc…

  • Published on

  • View

  • Download

Embed Size (px)


<ul><li><p>EUROPEAN ORGANIZATION FOR NUCLEAR RESEARCH (CERN)</p><p>CERN-EP-2017-156LHCb-PAPER-2017-018</p><p>July 6, 2017</p><p>Observation of the doubly charmedbaryon ++cc</p><p>LHCb collaboration</p><p>Abstract</p><p>A highly significant structure is observed in the +c K++ mass spectrum, where</p><p>the +c baryon is reconstructed in the decay mode pK+. The structure is</p><p>consistent with originating from a weakly decaying particle, identified as the doublycharmed baryon ++cc . The mass, measured relative to that of the </p><p>+c baryon, is</p><p>found to be 3621.40 0.72 (stat) 0.27 (syst) 0.14 (+c ) MeV/c2, where the lastuncertainty is due to the limited knowledge of the +c mass. The state is observedin a sample of proton-proton collision data collected by the LHCb experiment ata center-of-mass energy of 13 TeV, corresponding to an integrated luminosity of1.7 fb1, and confirmed in an additional sample of data collected at 8 TeV.</p><p>Submitted to Phys. Rev. Lett.</p><p>c CERN on behalf of the LHCb collaboration, license CC-BY-4.0.</p><p>Authors are listed at the end of this paper.</p><p>https://creativecommons.org/licenses/by/4.0/</p></li><li><p>ii</p></li><li><p>The quark model [13] predicts the existence of multiplets of baryon and meson states.Those states composed of the lightest four quarks (u, d, s, c) form SU(4) multiplets [4].Numerous states with charm quantum number C = 0 or C = 1 have been discovered,including all of the expected qq and qqq ground states [5]. Three weakly decaying qqqstates with C = 2 are expected: one isospin doublet (++cc = ccu and </p><p>+cc = ccd) and one</p><p>isospin singlet (+cc = ccs), each with spin-parity JP = 1/2+. The properties of these</p><p>baryons have been calculated with a variety of theoretical models. In most cases, themasses of the cc states are predicted to lie in the range 3500 to 3700 MeV/c</p><p>2 [622]. Themasses of the ++cc and </p><p>+cc states are expected to differ by only a few MeV/c</p><p>2, due toapproximate isospin symmetry [2325]. Most predictions for the lifetime of the +cc baryonare in the range 50 to 250 fs, and the lifetime of the ++cc baryon is expected to be between200 and 700 fs [10, 11, 19, 2629]. While both are expected to be produced at hadroncolliders [3032], the longer lifetime of the ++cc baryon should make it significantly easierto observe than the +cc baryon in such experiments, due to the use of real-time (online)event-selection requirements designed to reject backgrounds from the primary interactionpoint.</p><p>Experimentally, there is a longstanding puzzle in the cc system. Observations of the+cc baryon at a mass of 3519 2 MeV/c2 with signal yields of 15.9 events over 6.1 0.5background in the final state +c K</p><p>+ (6.3 significance), and 5.62 events over 1.380.13background in the final state pD+K (4.8 significance) were reported by the SELEXcollaboration [33,34]. Their results included a number of unexpected features, notablya short lifetime and a large production rate relative to that of the singly charmed +cbaryon. The lifetime was stated to be shorter than 33 fs at the 90% confidence level, andSELEX concluded that 20% of all +c baryons observed by the experiment originatedfrom +cc decays, implying a relative cc production rate several orders of magnitudelarger than theoretical expectations [11]. Searches at the FOCUS [35], BaBar [36], andBelle [37] experiments did not find evidence for a state with the properties reported bySELEX, and neither did a search at LHCb with data collected in 2011 corresponding toan integrated luminosity of 0.65 fb1 [38]. However, because the production environmentsat these experiments differ from that at SELEX, which studied collisions of a hyperonbeam on fixed nuclear targets, these null results do not exclude the original observations.</p><p>This Letter presents the observation of the ++cc baryon1 via the decay mode</p><p>+c K++ (Fig. 1), which is expected to have a branching fraction of up to 10% [39].</p><p>The +c baryon is reconstructed in the final state pK+. The data consist of pp colli-</p><p>sions collected by the LHCb experiment at the Large Hadron Collider at CERN with acenter-of-mass energy of 13 TeV taken in 2016, corresponding to an integrated luminosityof 1.7 fb1.</p><p>The LHCb detector is a single-arm forward spectrometer covering the pseudorapidityrange 2 &lt; &lt; 5, designed for the study of particles containing b or c quarks, and isdescribed in detail in Refs. [40,41]. The detector elements most relevant to this analysisare a silicon-strip vertex detector surrounding the pp interaction region, a tracking systemthat provides a measurement of the momentum of charged particles, and two ring-imagingCherenkov detectors [42] that are able to discriminate between different species of chargedhadrons. The online event selection is performed by a trigger that consists of a hardwarestage, which is based on information from the calorimeter and muon systems, followed</p><p>1 Inclusion of charge-conjugate processes is implied throughout.</p><p>1</p></li><li><p>ucdduusu</p><p>d</p><p>uc</p><p>c</p><p>W+</p><p>++cc +</p><p>+c</p><p>K</p><p>+</p><p>1Figure 1: Example Feynman diagram contributing to the decay ++cc +c K++.</p><p>by a software stage, which fully reconstructs the event [43]. The online reconstructionincorporates near-real-time alignment and calibration of the detector [44], which in turnallows the reconstruction of the ++cc decay to be performed entirely in the trigger software.</p><p>The reconstruction of ++cc +c K++ decays proceeds as follows. Candidate+c pK+ decays are reconstructed from three charged particles that form a good-quality vertex and that are inconsistent with originating from any pp collision primaryvertex (PV). The associated PV of a particle is defined to be the PV with respect to whichthe particle has the smallest impact parameter 2 (2IP), which is the difference in </p><p>2 ofthe PV fit with and without the particle in question; unless otherwise specified, the PV ofa particle refers to the associated PV. The +c vertex is required to be displaced from itsPV by a distance corresponding to a decay time greater than 150 fs. The +c candidateis then combined with three additional charged particles to form a ++cc +c K++candidate. These additional particles must form a good-quality vertex with the +ccandidate, and the +c decay vertex must be downstream of the </p><p>++cc vertex. Each of the</p><p>six final-state particles is required to pass track-quality requirements, to have hadron-identification information consistent with the appropriate hypothesis (p, K, or ), andto have transverse momentum pT &gt; 500 MeV/c. To avoid duplicate tracks, the anglebetween each pair of final-state particles with the same charge is required to be largerthan 0.5 mrad. The ++cc candidate must have pT &gt; 4 GeV/c and must be consistent withoriginating from its PV.</p><p>The background level is further reduced with a multivariate selector based on themultilayer perceptron algorithm [45]. The selector is trained with simulated signal eventsand with a control sample of data to represent the background. Simulated signal events areproduced with the standard LHCb simulation software [4652] interfaced to a dedicatedgenerator, Genxicc [5355], for ++cc baryon production. In the simulation, the </p><p>++cc</p><p>mass and lifetime are assumed to be 3.6 GeV/c2 and 333 fs. The background controlsample consists of wrong-sign (WS) +c K</p><p>+ combinations. For both signal andbackground training samples, candidates are required to pass the selection described aboveand to fall within a signal search region defined as 2270 &lt; mcand(</p><p>+c ) &lt; 2306 MeV/c</p><p>2 and3300 &lt; mcand(</p><p>++cc ) &lt; 3800 MeV/c</p><p>2, where mcand(+c ) is the reconstructed mass of the</p><p>2</p></li><li><p>+c candidate, mcand(++cc ) m(+c K+)mcand(+c ) +mPDG(+c ), m(+c K+)</p><p>is the reconstructed mass of the +c K+ combination, and mPDG(</p><p>+c ) = 2286.46</p><p>0.14 MeV/c2 is the known value of the +c mass [5]. The mcand(+c ) window corresponds</p><p>to approximately 3 times the +c mass resolution.Ten input variables are used in the multivariate selector: the 2 per degree of freedom</p><p>of each of the +c vertex fit, the ++cc vertex fit, and a kinematic refit [56] of the </p><p>++cc decay</p><p>chain requiring it to originate from its PV; the smallest pT of the three decay productsof the +c ; the smallest pT of the four decay products of the </p><p>++cc ; the scalar sum of the</p><p>pT of the four decay products of the ++cc ; the angle between the </p><p>++cc momentum vector</p><p>and the direction from the PV to the ++cc decay vertex; the flight distance 2 between</p><p>the PV and the ++cc decay vertex; the 2IP of the </p><p>++cc with respect to its PV; and the</p><p>smallest 2IP of the decay products of the ++cc with respect to its PV. Here the flight</p><p>distance 2 is defined as the 2 of the hypothesis that the ++cc decay vertex coincideswith its PV. Candidates are retained for analysis only if their multivariate selector outputvalues exceed a threshold chosen by maximizing the expected value of the figure of merit/(5</p><p>2+B) [57], where is the estimated signal efficiency and B is the estimated number</p><p>of background candidates underneath the signal peak. The quantity B is computed withthe WS control sample and, purely for the purposes of this optimization, it is calculated ina window centered at a mass of 3600 MeV/c2 and of halfwidth 12.5 MeV/c2 (correspondingto approximately twice the expected resolution). Its evaluation takes into account thedifference in background rates between the WS sample and the signal mode +c K</p><p>++,as estimated from data in the sideband regions 3200 &lt; mcand(</p><p>++cc ) &lt; 3300 MeV/c</p><p>2 and3800 &lt; mcand(</p><p>++cc ) &lt; 3900 MeV/c</p><p>2.After the multivariate selection is applied, events may still contain more than one ++cc</p><p>candidate in the search region 3300 &lt; mcand(++cc ) &lt; 3800 MeV/c</p><p>2. A peaking backgroundcould arise for cases in which the same six decay products are used but two of them areinterchanged (e.g., the K particle from the ++cc decay and the K</p><p> particle from the +cdecay). In such instances, one of the candidates is chosen at random to be retained andall others are discarded.</p><p>The selection described above, developed and optimized with simulated events andcontrol samples of data, is then applied to data in the search region. Figure 2 shows the +cmass distribution with a +c purity of 72% in the signal region, and the </p><p>++cc mass spectra</p><p>after the selection. A structure is visible in the signal mode at a mass of approximately3620 MeV/c2. No significant structure is visible in the WS control sample, nor for eventsin the +c mass sidebands. To measure the properties of the structure, an unbinnedextended maximum likelihood fit is performed to the invariant mass distribution in therestricted +c K</p><p>++ mass window of 3620 150 MeV/c2 (Fig. 3). The peaking structureis empirically described by a Gaussian function plus a modified Gaussian function withpower-law tails on both sides [58]. All peak parameters are fixed to values obtainedfrom simulation apart from the mass, yield, and an overall resolution parameter. Thebackground is described by a second-order polynomial with parameters free to float inthe fit. The signal yield is measured to be 313 33, corresponding to a local statisticalsignificance in excess of 12 when evaluated with a likelihood ratio test. The fittedresolution parameter is 6.6 0.8 MeV/c2, consistent with simulation. The same structureis also observed in the +c K</p><p>++ spectrum in a pp data sample collected by LHCbats = 8 TeV (see supplemental material in Appendix A for results from the 8 TeV</p><p>cross-check sample). The local statistical significance of the peak in the 8 TeV sample is</p><p>3</p></li><li><p>]2c) [MeV/+c(candm2250 2300 2350</p><p>2 cC</p><p>andi</p><p>date</p><p>s pe</p><p>r 3 </p><p>MeV</p><p>/</p><p>0</p><p>100</p><p>200</p><p>300</p><p>400</p><p>500</p><p>600</p><p>700</p><p>800</p><p>900 LHCb 13 TeV</p><p>SignalSideband</p><p>]2c) [MeV/++cc(candm3300 3400 3500 3600 3700 3800</p><p>2 cC</p><p>andi</p><p>date</p><p>s pe</p><p>r 10</p><p> MeV</p><p>/</p><p>0</p><p>50</p><p>100</p><p>150</p><p>200</p><p>250</p><p>300</p><p>350</p><p>Data RSData WSData SB</p><p>LHCb 13 TeV</p><p>Figure 2: Mass spectra of (left) +c and (right) ++cc candidates. The full selection is applied,</p><p>except for the +c mass requirement in the case of the left plot. For the +c mass distribution the</p><p>(cross-hatched) signal and (vertical lines) sideband regions are indicated; to avoid duplication,the histogram is filled only once in events that contain more than one ++cc candidate. Inthe right plot the right-sign (RS) signal sample ++cc +c K++ is shown, along with thecontrol samples: +c sideband (SB) </p><p>+c K</p><p>++ candidates and wrong-sign (WS) +c K+</p><p>candidates, normalized to have the same area as the RS sample in the mcand(++cc ) sidebands.</p><p>above seven standard deviations, and its mass is consistent with that in the 13 TeV datasample.</p><p>Additional cross-checks are performed to test the robustness of the observation. Theseinclude fixing the resolution parameter in the invariant mass fit to the value obtainedfrom simulation, changing the threshold value for the multivariate selector, using analternative selection without a multivariate classifier, testing the presence of any fakepeaking structures in the control samples when requiring various intermediate resonancesto be present (0, K0, 0c , </p><p>++c , </p><p>+c ), studying the contributions of misidentified</p><p>D+s K+K+ and D+ K++ decays, and testing for the presence of unphysicalstructures when combining ++cc and </p><p>+c decay products. In each of the tests where a</p><p>signal is expected, the significance of the structure in the +c K++ final state remains</p><p>above 12. The significance also remains above 12 in a subsample of candidates forwhich the reconstructed decay time exceeds five times its uncertainty. This is consistentwith a weakly decaying state and inconsistent with the strong decay of a resonance.</p><p>The sources of systematic uncertainty affecting the measurement of the ++cc mass(Table 1) include the momentum-scale calibration, the event selection, the unknown ++cclifetime, the invariant mass fit model, and the uncertainty on the +c mass. The momentumscale is calibrated with samples of J/ + and B+ J/K+ decays [59,60]. Aftercalibration, an uncertainty of 0.03% is assigned, which corresponds to a systematicuncertainty of 0.22 MeV/c2 on the reconstructed ++cc mass. The selection procedure ismore efficient for vertices that are well separated from the PV, and therefore preferentiallyretains longer-lived ++cc candidates. Due to a correlation between the reconstructeddecay time and the reconstructed mass, this induces a positive bias on the mass for both++cc and </p><p>+c candidates. The effect is studied with simulation and the bias on the </p><p>++cc</p><p>mass is determined to be +0.45 0.14 MeV/c2 (assuming a lifetime of 333 fs), where theuncertainty is due to the limited size of the simulation sample. A corresponding correction</p><p>4</p></li><li><p>]2c) [MeV/++cc(candm3500 3600 3700</p><p>2 cC</p><p>andi</p><p>date</p><p>s pe</p><p>r 5 </p><p>MeV</p><p>/</p><p>0</p><p>20</p><p>40</p><p>60</p><p>80</p><p>100</p><p>120</p><p>140</p><p>160</p><p>180</p><p>DataTotalSignalBackground</p><p>LHCb 13 TeV</p><p>Figure 3: Invariant mass distribution of +c K++ candidates with fit projections overlaid.</p><p>is applied to the fitted value in data. To validate this procedure, the +c mass in aninclusive sample is measured and corrected in the same way; after the correction, the +cmass is found to agree with the known value [5]. The bias on the ++cc mass depends on theunknown ++cc lifetime, introducing a further source of uncertainty on the correction. Thisis estima...</p></li></ul>