Abstract
Using the data samples of 102 million Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) and 158 million Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) events collected by the Belle detector, we
search for a pentaquark state in the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ final state from Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays. Here, the
charge-conjugate p ¯ J / ψ ¯ 𝑝 𝐽 𝜓 \bar{p}J/\psi over¯ start_ARG italic_p end_ARG italic_J / italic_ψ is included. We observe clear p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ production in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays and measure the
branching fractions to be ℬ [ Υ ( 1 S ) → p J / ψ + a n y t h i n g ] = [ 4.27 ± 0.16 ( s t a t . ) ± 0.20 ( s y s t . ) ] × 10 − 5 {\cal B}[\Upsilon(1S)\to pJ/\psi+anything]=[4.27\pm 0.16(stat.)\pm 0.20(syst.)%
]\times 10^{-5} caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] = [ 4.27 ± 0.16 ( italic_s italic_t italic_a italic_t . ) ± 0.20 ( italic_s italic_y italic_s italic_t . ) ] × 10 start_POSTSUPERSCRIPT - 5 end_POSTSUPERSCRIPT
and ℬ [ Υ ( 2 S ) → p J / ψ + a n y t h i n g ] = [ 3.59 ± 0.14 ( s t a t . ) ± 0.16 ( s y s t . ) ] × 10 − 5 {\cal B}[\Upsilon(2S)\to pJ/\psi+anything]=[3.59\pm 0.14(stat.)\pm 0.16(syst.)%
]\times 10^{-5} caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] = [ 3.59 ± 0.14 ( italic_s italic_t italic_a italic_t . ) ± 0.16 ( italic_s italic_y italic_s italic_t . ) ] × 10 start_POSTSUPERSCRIPT - 5 end_POSTSUPERSCRIPT . We also measure the
cross section of inclusive p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ production in e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT annihilation to be σ ( e + e − → p J / ψ + a n y t h i n g ) = [ 57.5 ± 2.1 ( s t a t . ) ± 2.5 ( s y s t . ) ] fb \sigma(e^{+}e^{-}\to pJ/\psi+anything)=[57.5\pm 2.1(stat.)\pm 2.5(syst.)]~{}$%
\mathrm{fb}$ italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) = [ 57.5 ± 2.1 ( italic_s italic_t italic_a italic_t . ) ± 2.5 ( italic_s italic_y italic_s italic_t . ) ] roman_fb at s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV using an 89.5 fb − 1 89.5 superscript femtobarn 1 89.5~{}$\mathrm{fb}$^{-1} 89.5 roman_fb start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT continuum data sample. There is
no significant P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT or P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signal found in the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ final states in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays. We
determine the upper limits of ℬ [ Υ ( 1 , 2 S ) → P c + + a n y t h i n g ] ⋅ ℬ ( P c + → p J / ψ ) ⋅ ℬ delimited-[] → Υ 1 2 𝑆 superscript subscript 𝑃 𝑐 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ → superscript subscript 𝑃 𝑐 𝑝 𝐽 𝜓 {\cal B}[\Upsilon(1,2S)\to P_{c}^{+}+anything]\cdot{\cal B}(P_{c}^{+}\to pJ/\psi) caligraphic_B [ roman_Υ ( 1 , 2 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ) to be at the 10 − 6 superscript 10 6 10^{-6} 10 start_POSTSUPERSCRIPT - 6 end_POSTSUPERSCRIPT
level.
I Introduction
In the conventional quark model, a hadron is either a meson containing a quark and an anti-quark or an (anti-)baryon
containing three (anti-)quarks. However, the fundamental theory of strong interaction, Quantum Chromodynamics, does
not forbid new structures of hadrons beyond the conventional quark model, such as glueball states containing only
gluons, hybrid states containing gluons and quarks, or multi-quark states containing more than three
quarks [1 ] . Many theoretical and experimental efforts have been devoted to predicting and searching
for these exotic states [2 , 3 ] . In 2003, the Belle experiment observed the X ( 3872 ) 𝑋 3872 X(3872) italic_X ( 3872 ) in B → K + π + π − J / ψ → 𝐵 𝐾 superscript 𝜋 superscript 𝜋 𝐽 𝜓 B\to K+\pi^{+}\pi^{-}J/\psi italic_B → italic_K + italic_π start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_π start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT italic_J / italic_ψ decay [4 ] , which was the clearest evidence yet of the existence of exotic states. Five years
later, Belle observed the Z ( 4430 ) + 𝑍 superscript 4430 Z(4430)^{+} italic_Z ( 4430 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT in the decay B → K + π + ψ ( 2 S ) → 𝐵 𝐾 superscript 𝜋 𝜓 2 𝑆 B\to K+\pi^{+}\psi(2S) italic_B → italic_K + italic_π start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_ψ ( 2 italic_S ) [5 ] . The c c ¯ 𝑐 ¯ 𝑐 c\bar{c} italic_c over¯ start_ARG italic_c end_ARG component and the
non-zero net charge of the final state π + ψ ( 2 S ) superscript 𝜋 𝜓 2 𝑆 \pi^{+}\psi(2S) italic_π start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_ψ ( 2 italic_S ) indicate that the Z ( 4430 ) + 𝑍 superscript 4430 Z(4430)^{+} italic_Z ( 4430 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT is a good candidate for a tetraquark
state. Since then, many candidate multi-quark states have been observed by the Belle, LHCb, and BESIII
experiments [6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 ] . In the pentaquark
sector, the LHCb experiment discovered P c ( 4380 ) + subscript 𝑃 𝑐 superscript 4380 P_{c}(4380)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4380 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT and P c ( 4450 ) + subscript 𝑃 𝑐 superscript 4450 P_{c}(4450)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4450 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT in the decay Λ b → K + p J / ψ → subscript Λ 𝑏 𝐾 𝑝 𝐽 𝜓 \Lambda_{b}\to K+pJ/\psi roman_Λ start_POSTSUBSCRIPT italic_b end_POSTSUBSCRIPT → italic_K + italic_p italic_J / italic_ψ [14 ] , but an updated analysis using ten times the statistics divided the
structures into three states [15 ] , the P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT . The deuteron can be considered a candidate for a hexaquark
state [16 ] . The observations of deuterons in the Υ ( n S ) Υ 𝑛 𝑆 \Upsilon(nS) roman_Υ ( italic_n italic_S ) inclusive decays by the ARGUS, CLEO, and BaBar
experiments provide clues of seaching for more candidates of multi-quark states in the Υ ( n S ) Υ 𝑛 𝑆 \Upsilon(nS) roman_Υ ( italic_n italic_S ) inclusive
decays [17 , 18 , 19 ] .
The Belle experiment collected the world’s largest Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) data samples in the last years of data taking. The
Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) data sample with an integrated luminosity ℒ Υ ( 1 S ) = 5.8 fb − 1 subscript ℒ Υ 1 𝑆 5.8 superscript femtobarn 1 \mathcal{L}_{\Upsilon(1S)}=5.8~{}$\mathrm{fb}$^{-1} caligraphic_L start_POSTSUBSCRIPT roman_Υ ( 1 italic_S ) end_POSTSUBSCRIPT = 5.8 roman_fb start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT contains ( 102 ± 2 ) × 10 6 Υ ( 1 S ) plus-or-minus 102 2 superscript 10 6 Υ 1 𝑆 (102\pm 2)\times 10^{6}~{}\Upsilon(1S) ( 102 ± 2 ) × 10 start_POSTSUPERSCRIPT 6 end_POSTSUPERSCRIPT roman_Υ ( 1 italic_S ) events [20 ] , while the Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) data sample has ℒ Υ ( 2 S ) = 24.7 fb − 1 subscript ℒ Υ 2 𝑆 24.7 superscript femtobarn 1 \mathcal{L}_{\Upsilon(2S)}=24.7~{}$\mathrm{fb}$^{-1} caligraphic_L start_POSTSUBSCRIPT roman_Υ ( 2 italic_S ) end_POSTSUBSCRIPT = 24.7 roman_fb start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT and
( 158 ± 4 ) × 10 6 Υ ( 2 S ) plus-or-minus 158 4 superscript 10 6 Υ 2 𝑆 (158\pm 4)\times 10^{6}~{}\Upsilon(2S) ( 158 ± 4 ) × 10 start_POSTSUPERSCRIPT 6 end_POSTSUPERSCRIPT roman_Υ ( 2 italic_S ) events [21 ] . Using the two data samples, we search for a P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT state in
the inclusive production of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ final states via Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays. Here and hereinafter, P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT is P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ,
P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , or P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT . The charge-conjugated final state P c − → p ¯ J / ψ → superscript subscript 𝑃 𝑐 ¯ 𝑝 𝐽 𝜓 P_{c}^{-}\to\bar{p}J/\psi italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → over¯ start_ARG italic_p end_ARG italic_J / italic_ψ is included throughout this study. We
also use a Belle continuum data sample with an integrated luminosity of ℒ cont = 89.5 fb − 1 subscript ℒ cont 89.5 superscript femtobarn 1 \mathcal{L}_{\rm cont}=89.5~{}$\mathrm{fb}$^{-1} caligraphic_L start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT = 89.5 roman_fb start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT taken
at center-of-mass (c.m.) energy s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV [60 MeV 60 MeV 60~{}\hbox{MeV} 60 MeV below the peak of the Υ ( 4 S ) Υ 4 𝑆 \Upsilon(4S) roman_Υ ( 4 italic_S ) resonance] to
investigate the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ final state from continuum productions, which could be backgrounds in the Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) data
samples for studying the Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays.
II The Belle Detector and Monte Carlo simulation
The Belle detector is a large-solid-angle magnetic spectrometer [22 ] . It consists of several subdetectors,
including a silicon vertex detector, a central drift chamber with 50 layers, an array of aerogel threshold Cherenkov
counters, a barrel-like arrangement of time-of-flight scintillation counters, and an electromagnetic calorimeter
(ECL) comprised of CsI(Tl) crystals. All the above are located within a superconducting solenoid coil which generates
a magnetic field of 1.5 T. An iron flux return outside the coil is instrumented to detect K L 0 superscript subscript 𝐾 𝐿 0 K_{L}^{0} italic_K start_POSTSUBSCRIPT italic_L end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 0 end_POSTSUPERSCRIPT mesons and identify
muons. The origin of the coordinate system is defined as the position of the nominal interaction point. The z 𝑧 z italic_z axis
is aligned with the direction opposite to the e + superscript 𝑒 e^{+} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT beam and points along the magnetic field within the solenoid.
The x 𝑥 x italic_x axis points horizontally outwards of the storage ring, and the y 𝑦 y italic_y axis is vertically upwards. The angles of
the polar (θ 𝜃 \theta italic_θ ) and azimuthal (ϕ italic-ϕ \phi italic_ϕ ) are measured relative to the positive z 𝑧 z italic_z axis.
To optimize the selection criteria, we use EvtGen to simulate signal Monte Carlo (MC) samples of Υ ( 1 , 2 S ) → P c + + p ¯ + q q ¯ → Υ 1 2 𝑆 superscript subscript 𝑃 𝑐 ¯ 𝑝 𝑞 ¯ 𝑞 \Upsilon(1,2S)\to P_{c}^{+}+\bar{p}+q\bar{q} roman_Υ ( 1 , 2 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + over¯ start_ARG italic_p end_ARG + italic_q over¯ start_ARG italic_q end_ARG with P c + → p J / ψ → superscript subscript 𝑃 𝑐 𝑝 𝐽 𝜓 P_{c}^{+}\to pJ/\psi italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ according to three-body phase space [23 ] , where q q ¯ ( q = u , d , s , c ) 𝑞 ¯ 𝑞 𝑞 𝑢 𝑑 𝑠 𝑐
q\bar{q}~{}(q=u,d,s,c) italic_q over¯ start_ARG italic_q end_ARG ( italic_q = italic_u , italic_d , italic_s , italic_c ) is a
quark-antiquark pair of random flavor whose hadronization is simulated by PYTHIA6.4 [24 ] . Each P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT MC
sample has 2 × 10 4 2 superscript 10 4 2\times 10^{4} 2 × 10 start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT events, and we combine the three P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signal MC samples for the selection criteria
optimization. To study the efficiency and mass resolution of the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ invariant mass (M p J / ψ subscript 𝑀 𝑝 𝐽 𝜓 M_{pJ/\psi} italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT ), we generate efficiency MC samples of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , whose mass is fixed to different values from 4.1 GeV / c 2 4.1 GeV superscript 𝑐 2 4.1~{}\hbox{GeV}/c^{2} 4.1 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT
to 5.0 GeV / c 2 5.0 GeV superscript 𝑐 2 5.0~{}\hbox{GeV}/c^{2} 5.0 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , and the width is set to zero. To study p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ production not due to P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT decays, we generate a
no-P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT MC sample of Υ ( 1 , 2 S ) → J / ψ + p + p ¯ + q q ¯ → Υ 1 2 𝑆 𝐽 𝜓 𝑝 ¯ 𝑝 𝑞 ¯ 𝑞 \Upsilon(1,2S)\to J/\psi+p+\bar{p}+q\bar{q} roman_Υ ( 1 , 2 italic_S ) → italic_J / italic_ψ + italic_p + over¯ start_ARG italic_p end_ARG + italic_q over¯ start_ARG italic_q end_ARG according to four-body phase space [23 ] . To
simulate the hadronization of q q ¯ 𝑞 ¯ 𝑞 q\bar{q} italic_q over¯ start_ARG italic_q end_ARG , we define a state of X → q q ¯ → 𝑋 𝑞 ¯ 𝑞 X\to q\bar{q} italic_X → italic_q over¯ start_ARG italic_q end_ARG where X 𝑋 X italic_X has a mass of 2.6 GeV / c 2 2.6 GeV superscript 𝑐 2 2.6~{}\hbox{GeV}/c^{2} 2.6 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT and a
width of 2.7 GeV 2.7 GeV 2.7~{}\hbox{GeV} 2.7 GeV in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays; similarly a mass of 3.2 GeV / c 2 3.2 GeV superscript 𝑐 2 3.2~{}\hbox{GeV}/c^{2} 3.2 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT and width of 3.3 GeV 3.3 GeV 3.3~{}\hbox{GeV} 3.3 GeV in Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays.
We simulate the geometry and the response of the Belle detector using a GEANT3-based MC technique [25 ] .
III Event selection
To reconstruct the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ final state, we select events with at least three well-measured charged tracks. Two
tracks with opposite charges are chosen as candidates for J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ decaying into e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT (called the e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode) or
μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT (called the μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode). A well-measured charged track has impact parameters of d r < 0.5 cm 𝑑 𝑟 0.5 centimeter dr<0.5~{}$\mathrm{cm}$ italic_d italic_r < 0.5 roman_cm in the r − ϕ 𝑟 italic-ϕ r-\phi italic_r - italic_ϕ
plane and | d z | < 5 cm 𝑑 𝑧 5 centimeter |dz|<5~{}$\mathrm{cm}$ | italic_d italic_z | < 5 roman_cm in the r − z 𝑟 𝑧 r-z italic_r - italic_z plane with respect to the interaction point, and a transverse momentum larger
than 0.1 GeV / c 0.1 GeV 𝑐 0.1~{}\hbox{GeV}/c 0.1 GeV / italic_c . For each charged track, we combine information from subdetectors of Belle to form a likelihood
ℒ i subscript ℒ 𝑖 \mathcal{L}_{i} caligraphic_L start_POSTSUBSCRIPT italic_i end_POSTSUBSCRIPT for each putative particle species (i 𝑖 i italic_i ) [26 ] . We form the likelihood ratios ℛ e ≡ ℒ e / ( ℒ e + ℒ hadrons ) subscript ℛ 𝑒 subscript ℒ 𝑒 subscript ℒ 𝑒 subscript ℒ hadrons \mathcal{R}_{e}\equiv\mathcal{L}_{e}/(\mathcal{L}_{e}+\mathcal{L}_{\rm hadrons}) caligraphic_R start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT ≡ caligraphic_L start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT / ( caligraphic_L start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT + caligraphic_L start_POSTSUBSCRIPT roman_hadrons end_POSTSUBSCRIPT ) and ℛ μ ≡ ℒ μ / ( ℒ μ + ℒ hadrons ) subscript ℛ 𝜇 subscript ℒ 𝜇 subscript ℒ 𝜇 subscript ℒ hadrons \mathcal{R}_{\mu}\equiv\mathcal{L}_{\mu}/(\mathcal{L}_{\mu}+\mathcal{L}_{\rm
hadrons}) caligraphic_R start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT ≡ caligraphic_L start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT / ( caligraphic_L start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT + caligraphic_L start_POSTSUBSCRIPT roman_hadrons end_POSTSUBSCRIPT ) for electron and muon identifications [27 , 28 ] .
For electrons from J / ψ → e + e − → 𝐽 𝜓 superscript 𝑒 superscript 𝑒 J/\psi\to e^{+}e^{-} italic_J / italic_ψ → italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT decay, we require both tracks to have ℛ e > 0.9 subscript ℛ 𝑒 0.9 \mathcal{R}_{e}>0.9 caligraphic_R start_POSTSUBSCRIPT italic_e end_POSTSUBSCRIPT > 0.9 and include the
bremsstrahlung photons detected in the ECL within 0.05 radians of the original e + superscript 𝑒 e^{+} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT or e − superscript 𝑒 e^{-} italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT direction in
calculating the e + e − ( γ ) superscript 𝑒 superscript 𝑒 𝛾 e^{+}e^{-}(\gamma) italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ( italic_γ ) invariant mass. For muons from J / ψ → μ + μ − → 𝐽 𝜓 superscript 𝜇 superscript 𝜇 J/\psi\to\mu^{+}\mu^{-} italic_J / italic_ψ → italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT decay, we require both tracks to have
ℛ μ > 0.9 subscript ℛ 𝜇 0.9 \mathcal{R}_{\mu}>0.9 caligraphic_R start_POSTSUBSCRIPT italic_μ end_POSTSUBSCRIPT > 0.9 . The single lepton identification efficiency is ( 93.9 ± 0.2 ) % percent plus-or-minus 93.9 0.2 (93.9\pm 0.2)\% ( 93.9 ± 0.2 ) % in the e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode and
( 91.9 ± 0.2 ) % percent plus-or-minus 91.9 0.2 (91.9\pm 0.2)\% ( 91.9 ± 0.2 ) % in the μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode. We identify a track with ℛ p / K = ℒ p ℒ p + ℒ K > 0.6 subscript ℛ 𝑝 𝐾 subscript ℒ 𝑝 subscript ℒ 𝑝 subscript ℒ 𝐾 0.6 \mathcal{R}_{p/K}=\frac{\mathcal{L}_{p}}{\mathcal{L}_{p}+\mathcal{L}_{K}}>0.6 caligraphic_R start_POSTSUBSCRIPT italic_p / italic_K end_POSTSUBSCRIPT = divide start_ARG caligraphic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT end_ARG start_ARG caligraphic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT + caligraphic_L start_POSTSUBSCRIPT italic_K end_POSTSUBSCRIPT end_ARG > 0.6 and ℛ p / π = ℒ p ℒ p + ℒ π > 0.6 subscript ℛ 𝑝 𝜋 subscript ℒ 𝑝 subscript ℒ 𝑝 subscript ℒ 𝜋 0.6 \mathcal{R}_{p/\pi}=\frac{\mathcal{L}_{p}}{\mathcal{L}_{p}+\mathcal{L}_{\pi}}>%
0.6 caligraphic_R start_POSTSUBSCRIPT italic_p / italic_π end_POSTSUBSCRIPT = divide start_ARG caligraphic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT end_ARG start_ARG caligraphic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT + caligraphic_L start_POSTSUBSCRIPT italic_π end_POSTSUBSCRIPT end_ARG > 0.6 as a proton. The efficiency of proton identification is ( 97.3 ± 0.1 ) % percent plus-or-minus 97.3 0.1 (97.3\pm 0.1)\% ( 97.3 ± 0.1 ) % . To remove
backgrounds from Λ → p π → Λ 𝑝 𝜋 \Lambda\to p\pi roman_Λ → italic_p italic_π decay in the proton selection, we reconstruct all the pion candidates with
ℛ π / K = ℒ π / ( ℒ π + ℒ K ) > 0.6 subscript ℛ 𝜋 𝐾 subscript ℒ 𝜋 subscript ℒ 𝜋 subscript ℒ 𝐾 0.6 \mathcal{R}_{\pi/K}=\mathcal{L}_{\pi}/(\mathcal{L}_{\pi}+\mathcal{L}_{K})>0.6 caligraphic_R start_POSTSUBSCRIPT italic_π / italic_K end_POSTSUBSCRIPT = caligraphic_L start_POSTSUBSCRIPT italic_π end_POSTSUBSCRIPT / ( caligraphic_L start_POSTSUBSCRIPT italic_π end_POSTSUBSCRIPT + caligraphic_L start_POSTSUBSCRIPT italic_K end_POSTSUBSCRIPT ) > 0.6 and a charge opposite to that of the
proton. We remove the proton candidate if it is part of any p π 𝑝 𝜋 p\pi italic_p italic_π combination of mass 1.105 GeV / c 2 < M p π < 1.12 GeV / c 2 1.105 GeV superscript 𝑐 2 subscript 𝑀 𝑝 𝜋 1.12 GeV superscript 𝑐 2 1.105~{}\hbox{GeV}/c^{2}<M_{p\pi}<1.12~{}\hbox{GeV}/c^{2} 1.105 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT < italic_M start_POSTSUBSCRIPT italic_p italic_π end_POSTSUBSCRIPT < 1.12 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , where M p π subscript 𝑀 𝑝 𝜋 M_{p\pi} italic_M start_POSTSUBSCRIPT italic_p italic_π end_POSTSUBSCRIPT is the invariant mass of the p π 𝑝 𝜋 p\pi italic_p italic_π combination. Furthermore, to remove the proton
candidates from beam backgrounds, we require the difference of the d z 𝑑 𝑧 dz italic_d italic_z parameter for p 𝑝 p italic_p and ℓ ± superscript ℓ plus-or-minus \ell^{\pm} roman_ℓ start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT to be
| Δ d z | < 0.5 cm Δ 𝑑 𝑧 0.5 centimeter |\Delta dz|<0.5~{}$\mathrm{cm}$ | roman_Δ italic_d italic_z | < 0.5 roman_cm .
The Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) data samples, and the continuum data sample, all show clear J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signals in both the e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode
and the μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode. Figure 1 shows the invariant-mass distributions of the lepton pair (M ℓ + ℓ − subscript 𝑀 superscript ℓ superscript ℓ M_{\ell^{+}\ell^{-}} italic_M start_POSTSUBSCRIPT roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT ), which is
the sum of the e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode and the μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode, in the Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) data samples. Fitting the M ℓ + ℓ − subscript 𝑀 superscript ℓ superscript ℓ M_{\ell^{+}\ell^{-}} italic_M start_POSTSUBSCRIPT roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT distributions
using a Gaussian function for the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal and a second-order Chebychev function for the backgrounds, we get
the mass resolution of the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal to be 8.7 ± 0.6 MeV / c 2 plus-or-minus 8.7 0.6 MeV superscript 𝑐 2 8.7\pm 0.6~{}\hbox{MeV}/c^{2} 8.7 ± 0.6 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT (10.1 ± 0.5 MeV / c 2 plus-or-minus 10.1 0.5 MeV superscript 𝑐 2 10.1\pm 0.5~{}\hbox{MeV}/c^{2} 10.1 ± 0.5 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) in the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S )
[Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) ] data sample and 8.2 ± 0.1 MeV / c 2 plus-or-minus 8.2 0.1 MeV superscript 𝑐 2 8.2\pm 0.1~{}\hbox{MeV}/c^{2} 8.2 ± 0.1 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT (8.6 ± 0.1 MeV / c 2 plus-or-minus 8.6 0.1 MeV superscript 𝑐 2 8.6\pm 0.1~{}\hbox{MeV}/c^{2} 8.6 ± 0.1 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ) in the signal MC simulation of Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S )
[Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) ] decays. We define the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal region to be | M ℓ + ℓ − − m J / ψ | < 3 σ subscript 𝑀 superscript ℓ superscript ℓ subscript 𝑚 𝐽 𝜓 3 𝜎 |M_{\ell^{+}\ell^{-}}-m_{J/\psi}|<3\sigma | italic_M start_POSTSUBSCRIPT roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT - italic_m start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT | < 3 italic_σ , where m J / ψ subscript 𝑚 𝐽 𝜓 m_{J/\psi} italic_m start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT is
the nominal mass of J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ [29 ] and σ = 10 MeV / c 2 𝜎 10 MeV superscript 𝑐 2 \sigma=10~{}\hbox{MeV}/c^{2} italic_σ = 10 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT . To estimate the backgrounds to the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ , we
define the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sideband regions as | M ℓ + ℓ − − m J / ψ ± 9 σ | < 3 σ plus-or-minus subscript 𝑀 superscript ℓ superscript ℓ subscript 𝑚 𝐽 𝜓 9 𝜎 3 𝜎 |M_{\ell^{+}\ell^{-}}-m_{J/\psi}\pm 9\sigma|<3\sigma | italic_M start_POSTSUBSCRIPT roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT - italic_m start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT ± 9 italic_σ | < 3 italic_σ .
Figure 1: The invariant-mass distributions of the lepton pair from (a) the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) data sample and (b) the Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S )
data sample. The curves show the best fit results with a Gaussian function for the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal and a second-order
Chebychev function for the backgrounds. The red arrows indicate the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal region and the green ones
indicate the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sideband regions.
Figures 2 (a) and 2 (b) show the distributions of the recoil mass squared against the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ system
in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) data samples and signal MC simulations. This quantity is calculated by M recoil 2 ( p J / ψ ) ≡ ( P e + e − − P J / ψ ) 2 superscript subscript 𝑀 recoil 2 𝑝 𝐽 𝜓 superscript subscript 𝑃 superscript 𝑒 superscript 𝑒 subscript 𝑃 𝐽 𝜓 2 M_{\rm recoil}^{2}(pJ/\psi)\equiv(P_{e^{+}e^{-}}-P_{J/\psi})^{2} italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p italic_J / italic_ψ ) ≡ ( italic_P start_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT - italic_P start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , where P e + e − subscript 𝑃 superscript 𝑒 superscript 𝑒 P_{e^{+}e^{-}} italic_P start_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT is the 4-momentum of the e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT collision and P p J / ψ subscript 𝑃 𝑝 𝐽 𝜓 P_{pJ/\psi} italic_P start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT is the 4-momentum of the
p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ combination. In data, there are accumulations between − 5 GeV 2 / c 4 5 superscript GeV 2 superscript 𝑐 4 -5~{}\hbox{GeV}^{2}/c^{4} - 5 GeV start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT / italic_c start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT and 5 GeV 2 / c 4 5 superscript GeV 2 superscript 𝑐 4 5~{}\hbox{GeV}^{2}/c^{4} 5 GeV start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT / italic_c start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT for the events selected
in the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal region and these can be described well with the backgrounds estimated from the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass
sideband regions. These backgrounds appear in the e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode but are scarce in the μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode. On the other hand,
these events produce a large peak at zero and a wide distribution of the recoil mass squared against the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ
candidate, calculated by M recoil 2 ( J / ψ ) ≡ ( P e + e − − P J / ψ ) 2 superscript subscript 𝑀 recoil 2 𝐽 𝜓 superscript subscript 𝑃 superscript 𝑒 superscript 𝑒 subscript 𝑃 𝐽 𝜓 2 M_{\rm recoil}^{2}(J/\psi)\equiv(P_{e^{+}e^{-}}-P_{J/\psi})^{2} italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_J / italic_ψ ) ≡ ( italic_P start_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT - italic_P start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT ) start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , where P J / ψ subscript 𝑃 𝐽 𝜓 P_{J/\psi} italic_P start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT is the 4-momentum of
the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ candidate. They are identified as backgrounds from Bhabha events with high energy bremsstrahlung
radiation photon(s) and an additional proton from beam backgrounds. As this proton is not from an e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT collision,
this background can produce negative accumulations in the M recoil 2 ( p J / ψ ) superscript subscript 𝑀 recoil 2 𝑝 𝐽 𝜓 M_{\rm recoil}^{2}(pJ/\psi) italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p italic_J / italic_ψ ) distributions. We require M recoil 2 ( p J / ψ ) > 10 GeV 2 / c 4 superscript subscript 𝑀 recoil 2 𝑝 𝐽 𝜓 10 superscript GeV 2 superscript 𝑐 4 M_{\rm recoil}^{2}(pJ/\psi)>10~{}\hbox{GeV}^{2}/c^{4} italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p italic_J / italic_ψ ) > 10 GeV start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT / italic_c start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT to suppress these backgrounds with a selection efficiency of about 99% in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays.
Figures 2 (c) and 2 (d) show the distributions of M recoil 2 ( J / ψ ) superscript subscript 𝑀 recoil 2 𝐽 𝜓 M_{\rm recoil}^{2}(J/\psi) italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_J / italic_ψ ) after this requirement. We notice
that the data have higher distributions than signal MC simulations in the region M recoil 2 ( J / ψ ) < 30 GeV 2 / c 4 superscript subscript 𝑀 recoil 2 𝐽 𝜓 30 superscript GeV 2 superscript 𝑐 4 M_{\rm recoil}^{2}(J/\psi)<30~{}\hbox{GeV}^{2}/c^{4} italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_J / italic_ψ ) < 30 GeV start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT / italic_c start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT .
In the range M recoil 2 ( J / ψ ) > 30 GeV 2 / c 4 superscript subscript 𝑀 recoil 2 𝐽 𝜓 30 superscript GeV 2 superscript 𝑐 4 M_{\rm recoil}^{2}(J/\psi)>30~{}\hbox{GeV}^{2}/c^{4} italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_J / italic_ψ ) > 30 GeV start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT / italic_c start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT , the MC and the data are in good agreement in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) , but the MC is slightly higher than the data in Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) . It is also interesting to see an enhancement at around 22 GeV 2 / c 4 22 superscript GeV 2 superscript 𝑐 4 22~{}\hbox{GeV}^{2}/c^{4} 22 GeV start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT / italic_c start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT in the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays. However, the statistics are too limited to draw any conclusions with the presently available dataset.
Figure 2: The distributions of the recoil mass squared of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ (upper), and J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ (lower) in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) (left)
and Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) (right) decays. The dots with error bars are data, the shaded histograms are backgrounds estimated from
the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sideband regions, and the solid histograms are signal MC simulations. The arrows show the
requirement M recoil 2 ( p J / ψ ) > 10 GeV 2 / c 4 superscript subscript 𝑀 recoil 2 𝑝 𝐽 𝜓 10 superscript GeV 2 superscript 𝑐 4 M_{\rm recoil}^{2}(pJ/\psi)>10~{}{\hbox{GeV}^{2}/c^{4}} italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p italic_J / italic_ψ ) > 10 GeV start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT / italic_c start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT .
IV Invariant mass spectra of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ
All the candidates satisfying the selection criteria described above are accepted, including p 𝑝 p italic_p or p ¯ ¯ 𝑝 \bar{p} over¯ start_ARG italic_p end_ARG with the
same J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ candidate or multiple candidates sharing one lepton. We show the momentum distributions of the p / p ¯ 𝑝 ¯ 𝑝 p/\bar{p} italic_p / over¯ start_ARG italic_p end_ARG
after selection criteria in Fig. 3 .
Figure 3: The momentum distributions of p / p ¯ 𝑝 ¯ 𝑝 p/\bar{p} italic_p / over¯ start_ARG italic_p end_ARG in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays. The first row is the momenta of
p 𝑝 p italic_p and the second of p ¯ ¯ 𝑝 \bar{p} over¯ start_ARG italic_p end_ARG . The left and right panels are Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) and Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) , respectively. The dots with error
bars are data, the shaded histograms are backgrounds estimated from the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sideband regions, and the solid histograms are signal MC simulations.
According to the efficiency MC simulations, we obtain an efficiency varying from 29% (26%) to 36% (33%) in the
Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) [Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) ] decays, and the mass resolution increasing from 1.5 MeV / c 2 1.5 MeV superscript 𝑐 2 1.5~{}\hbox{MeV}/c^{2} 1.5 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT to 4.9 MeV / c 2 4.9 MeV superscript 𝑐 2 4.9~{}\hbox{MeV}/c^{2} 4.9 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT for M p J / ψ ∈ [ 4.1 , 5.0 ] GeV / c 2 subscript 𝑀 𝑝 𝐽 𝜓 4.1 5.0 GeV superscript 𝑐 2 M_{pJ/\psi}\in[4.1,~{}5.0]~{}\hbox{GeV}/c^{2} italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT ∈ [ 4.1 , 5.0 ] GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT . We notice that the width of P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT reported by LHCb is Γ P c ( 4457 ) + = 6.4 ± 2.0 − 1.9 + 5.7 MeV subscript Γ subscript 𝑃 𝑐 superscript 4457 plus-or-minus 6.4 subscript superscript 2.0 5.7 1.9 MeV \Gamma_{P_{c}(4457)^{+}}=6.4\pm 2.0^{+5.7}_{-1.9}~{}\hbox{MeV} roman_Γ start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT = 6.4 ± 2.0 start_POSTSUPERSCRIPT + 5.7 end_POSTSUPERSCRIPT start_POSTSUBSCRIPT - 1.9 end_POSTSUBSCRIPT MeV [15 ] and the mass resolution near the mass of P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT is about 3.0 MeV / c 2 3.0 MeV superscript 𝑐 2 3.0~{}\hbox{MeV}/c^{2} 3.0 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT .
Therefore, we need to consider the mass resolution in fitting the M p J / ψ subscript 𝑀 𝑝 𝐽 𝜓 M_{pJ/\psi} italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT distributions for the possible P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
signals. Here and hereinafter, the first uncertainty quoted is statistical, while the second corresponds to the total
systematic uncertainty.
We then study the M p J / ψ subscript 𝑀 𝑝 𝐽 𝜓 M_{pJ/\psi} italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT distributions from the signal MC simulations of P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT . In each
distribution, there is a clear P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT peak and a plateau of wrong combination with particle(s) from the recoil of
P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT . We perform a fit to this distribution using a Breit-Wigner function convolved with a Gaussian resolution
function to describe the signals and a first-order polynomial function to describe the plateau of the wrong
combinations. The fit range is M P c + ± 200 MeV / c 2 plus-or-minus subscript 𝑀 superscript subscript 𝑃 𝑐 200 MeV superscript 𝑐 2 M_{P_{c}^{+}}\pm 200~{}\hbox{MeV}/c^{2} italic_M start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT ± 200 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , where M P c + subscript 𝑀 superscript subscript 𝑃 𝑐 M_{P_{c}^{+}} italic_M start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT is the mass of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT . The fits yield mass
resolutions of around 3 MeV / c 2 3 MeV superscript 𝑐 2 3~{}\hbox{MeV}/c^{2} 3 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT for each P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT state. The mass resolutions obtained here agree with those
obtained from the efficiency MC simulations directly. We calculate the ratio ℛ ≡ N P c + / N p J / ψ ℛ subscript 𝑁 superscript subscript 𝑃 𝑐 subscript 𝑁 𝑝 𝐽 𝜓 \mathcal{R}\equiv N_{P_{c}^{+}}/N_{pJ/\psi} caligraphic_R ≡ italic_N start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT / italic_N start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT
to be approximately 0.6, where the N P c + subscript 𝑁 superscript subscript 𝑃 𝑐 N_{P_{c}^{+}} italic_N start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT and N p J / ψ subscript 𝑁 𝑝 𝐽 𝜓 N_{pJ/\psi} italic_N start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT are the number of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signals from the fit and the
number of all p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ combinations being selected between 4.0 GeV / c 2 4.0 GeV superscript 𝑐 2 4.0~{}\hbox{GeV}/c^{2} 4.0 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT and 5.0 GeV / c 2 5.0 GeV superscript 𝑐 2 5.0~{}\hbox{GeV}/c^{2} 5.0 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT , respectively. The
efficiencies of all combinations (ε allcmb MC subscript superscript 𝜀 MC allcmb \varepsilon^{\rm MC}_{\rm allcmb} italic_ε start_POSTSUPERSCRIPT roman_MC end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_allcmb end_POSTSUBSCRIPT ) are about 60%. We list the details of the mass
resolutions, the ratios ℛ ℛ \mathcal{R} caligraphic_R , and the efficiencies ε allcmb MC subscript superscript 𝜀 MC allcmb \varepsilon^{\rm MC}_{\rm allcmb} italic_ε start_POSTSUPERSCRIPT roman_MC end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_allcmb end_POSTSUBSCRIPT from the signal MC
simulations of P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays in Table 1 .
Table 1: The mass resolution, the ratio of the number of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signals to the number of all p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ combinations,
and the efficiency of all the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ combinations from the signal MC simulations of P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT in
Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays.
Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays
Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays
—
P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
mass resolution ( MeV / c 2 MeV superscript 𝑐 2 \hbox{MeV}/c^{2} MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT )
2.9 ± 0.1 plus-or-minus 2.9 0.1 2.9\pm 0.1 2.9 ± 0.1
3.2 ± 0.2 plus-or-minus 3.2 0.2 3.2\pm 0.2 3.2 ± 0.2
3.4 ± 0.1 plus-or-minus 3.4 0.1 3.4\pm 0.1 3.4 ± 0.1
2.8 ± 0.1 plus-or-minus 2.8 0.1 2.8\pm 0.1 2.8 ± 0.1
3.0 ± 0.2 plus-or-minus 3.0 0.2 3.0\pm 0.2 3.0 ± 0.2
3.2 ± 0.1 plus-or-minus 3.2 0.1 3.2\pm 0.1 3.2 ± 0.1
Ratio of N P c + / N p J / ψ subscript 𝑁 superscript subscript 𝑃 𝑐 subscript 𝑁 𝑝 𝐽 𝜓 N_{P_{c}^{+}}/N_{pJ/\psi} italic_N start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT / italic_N start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT
0.60
0.56
0.57
0.59
0.56
0.54
ε allcmb MC subscript superscript 𝜀 MC allcmb \varepsilon^{\rm MC}_{\rm allcmb} italic_ε start_POSTSUPERSCRIPT roman_MC end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_allcmb end_POSTSUBSCRIPT (%)
58.7 ± 0.1 plus-or-minus 58.7 0.1 58.7\pm 0.1 58.7 ± 0.1
59.2 ± 0.1 plus-or-minus 59.2 0.1 59.2\pm 0.1 59.2 ± 0.1
59.7 ± 0.1 plus-or-minus 59.7 0.1 59.7\pm 0.1 59.7 ± 0.1
58.9 ± 0.1 plus-or-minus 58.9 0.1 58.9\pm 0.1 58.9 ± 0.1
59.2 ± 0.1 plus-or-minus 59.2 0.1 59.2\pm 0.1 59.2 ± 0.1
59.5 ± 0.1 plus-or-minus 59.5 0.1 59.5\pm 0.1 59.5 ± 0.1
Figure 4: The invariant-mass distributions of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ in the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) , Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) , and continuum data samples. From
left to right, the four panels are p + J / ψ 𝑝 𝐽 𝜓 p+J/\psi italic_p + italic_J / italic_ψ in e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode, p ¯ + J / ψ ¯ 𝑝 𝐽 𝜓 \bar{p}+J/\psi over¯ start_ARG italic_p end_ARG + italic_J / italic_ψ in e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode, p + J / ψ 𝑝 𝐽 𝜓 p+J/\psi italic_p + italic_J / italic_ψ in μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode, and
p ¯ + J / ψ ¯ 𝑝 𝐽 𝜓 \bar{p}+J/\psi over¯ start_ARG italic_p end_ARG + italic_J / italic_ψ in μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT mode. From top to bottom, the three rows are the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays, the Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays, and the
continuum productions at s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV . The solid histograms are the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ signals, and the shaded
histograms are backgrounds estimated from the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sideband regions.
We study the M p J / ψ subscript 𝑀 𝑝 𝐽 𝜓 M_{pJ/\psi} italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT distributions obtained from the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) , Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) , and continuum data samples, and show
them in Figs. 4 (a-d), 4 (e-h), and 4 (i-l), respectively. There are
clear p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ signals in the three data samples. As mentioned, we use the distributions obtained from the continuum
data sample to estimate the backgrounds from e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT annihilation in the Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays. For this, we scale the
luminosities and correct for the efficiencies and the c.m. energy dependence of the Quantum Electrodynamics (QED) cross
section σ e + e − ∝ 1 / s proportional-to subscript 𝜎 superscript 𝑒 superscript 𝑒 1 𝑠 \sigma_{e^{+}e^{-}}\propto 1/s italic_σ start_POSTSUBSCRIPT italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT ∝ 1 / italic_s , resulting in scale factors f scale = ( ℒ Υ ( 1 , 2 S ) × ε Υ ( 1 , 2 S ) × s cont ) / ( ℒ cont × ε cont × s Υ ( 1 , 2 S ) ) = 0.077 subscript 𝑓 scale subscript ℒ Υ 1 2 𝑆 subscript 𝜀 Υ 1 2 𝑆 subscript 𝑠 cont subscript ℒ cont subscript 𝜀 cont subscript 𝑠 Υ 1 2 𝑆 0.077 f_{\rm scale}=(\mathcal{L}_{\Upsilon(1,2S)}\times\varepsilon_{\Upsilon(1,2S)}%
\times s_{\rm cont})/(\mathcal{L}_{\rm cont}\times\varepsilon_{\rm cont}\times
s%
_{\Upsilon(1,2S)})=0.077 italic_f start_POSTSUBSCRIPT roman_scale end_POSTSUBSCRIPT = ( caligraphic_L start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT × italic_ε start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT × italic_s start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT ) / ( caligraphic_L start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT × italic_ε start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT × italic_s start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT ) = 0.077
and 0.301 for Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) and Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) , respectively. We find no peaking component in the combined M p J / ψ subscript 𝑀 𝑝 𝐽 𝜓 M_{pJ/\psi} italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT
distribution from Figs 4 (i-l) and obtain the number of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ candidates to be
N cont p J / ψ = 383 ± 20 subscript superscript 𝑁 𝑝 𝐽 𝜓 cont plus-or-minus 383 20 N^{pJ/\psi}_{\rm cont}=383\pm 20 italic_N start_POSTSUPERSCRIPT italic_p italic_J / italic_ψ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT = 383 ± 20 after subtracting the backgrounds estimated from the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sideband
regions.
To estimate the backgrounds due to the mis-identification of proton, we replace the proton identification
requirements with ℒ p / ( ℒ p + ℒ K ) < 0.4 subscript ℒ 𝑝 subscript ℒ 𝑝 subscript ℒ 𝐾 0.4 \mathcal{L}_{p}/(\mathcal{L}_{p}+\mathcal{L}_{K})<0.4 caligraphic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT / ( caligraphic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT + caligraphic_L start_POSTSUBSCRIPT italic_K end_POSTSUBSCRIPT ) < 0.4 or
ℒ p / ( ℒ p + ℒ π ) < 0.4 subscript ℒ 𝑝 subscript ℒ 𝑝 subscript ℒ 𝜋 0.4 \mathcal{L}_{p}/(\mathcal{L}_{p}+\mathcal{L}_{\pi})<0.4 caligraphic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT / ( caligraphic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT + caligraphic_L start_POSTSUBSCRIPT italic_π end_POSTSUBSCRIPT ) < 0.4 in the signal selection. We obtain 1746 ± 42 plus-or-minus 1746 42 1746\pm 42 1746 ± 42 K ± J / ψ superscript 𝐾 plus-or-minus 𝐽 𝜓 K^{\pm}J/\psi italic_K start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT italic_J / italic_ψ
signals with kaon identification efficiency of 93.5% or 1710 ± 41 plus-or-minus 1710 41 1710\pm 41 1710 ± 41 π ± J / ψ superscript 𝜋 plus-or-minus 𝐽 𝜓 \pi^{\pm}J/\psi italic_π start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT italic_J / italic_ψ signals with pion
identification efficiency of 92.4%. Taking into account mis-identification rates, we expect the number of
backgrounds from K ± J / ψ superscript 𝐾 plus-or-minus 𝐽 𝜓 K^{\pm}J/\psi italic_K start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT italic_J / italic_ψ or π ± J / ψ superscript 𝜋 plus-or-minus 𝐽 𝜓 \pi^{\pm}J/\psi italic_π start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT italic_J / italic_ψ to be 50.3 ± 0.9 ± 1.3 plus-or-minus 50.3 0.9 1.3 50.3\pm 0.9\pm 1.3 50.3 ± 0.9 ± 1.3 , where the systematic uncertainty is
described in Sec. V .
Hence, the number of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ events after all background subtractions is found to be N cont p J / ψ = 333 ± 18 subscript superscript 𝑁 𝑝 𝐽 𝜓 cont plus-or-minus 333 18 N^{pJ/\psi}_{\rm cont}=333\pm 18 italic_N start_POSTSUPERSCRIPT italic_p italic_J / italic_ψ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT = 333 ± 18 .
With the scale factor f scale subscript 𝑓 scale f_{\rm scale} italic_f start_POSTSUBSCRIPT roman_scale end_POSTSUBSCRIPT , we expect 26 ± 2 ± 1 plus-or-minus 26 2 1 26\pm 2\pm 1 26 ± 2 ± 1 and 100 ± 5 ± 4 plus-or-minus 100 5 4 100\pm 5\pm 4 100 ± 5 ± 4 p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ signals from e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT annihilation in the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) and Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) data samples, respectively.
We use the N cont p J / ψ subscript superscript 𝑁 𝑝 𝐽 𝜓 cont N^{pJ/\psi}_{\rm cont} italic_N start_POSTSUPERSCRIPT italic_p italic_J / italic_ψ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT obtained from the continuum data sample to calculate the cross section of the
inclusive p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ production in e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT annihilation via
σ ( e + e − → p J / ψ + a n y t h i n g ) = N cont p J / ψ L cont × ε cont noP c × ℬ ( J / ψ → ℓ + ℓ − ) × ( 1 + δ ISR ) . 𝜎 → superscript 𝑒 superscript 𝑒 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 subscript superscript 𝑁 𝑝 𝐽 𝜓 cont subscript 𝐿 cont superscript subscript 𝜀 cont subscript noP c ℬ → 𝐽 𝜓 superscript ℓ superscript ℓ 1 subscript 𝛿 ISR \sigma(e^{+}e^{-}\to pJ/\psi+anything)=\frac{N^{pJ/\psi}_{\rm cont}}{L_{\rm
cont%
}\times\varepsilon_{\rm cont}^{\rm noP_{c}}\times{\cal B}(J/\psi\to\ell^{+}%
\ell^{-})\times(1+\delta_{\rm ISR})}. italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) = divide start_ARG italic_N start_POSTSUPERSCRIPT italic_p italic_J / italic_ψ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT end_ARG start_ARG italic_L start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT × italic_ε start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_noP start_POSTSUBSCRIPT roman_c end_POSTSUBSCRIPT end_POSTSUPERSCRIPT × caligraphic_B ( italic_J / italic_ψ → roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ) × ( 1 + italic_δ start_POSTSUBSCRIPT roman_ISR end_POSTSUBSCRIPT ) end_ARG .
(1)
Here, ε cont noP c = 66.1 % superscript subscript 𝜀 cont subscript noP c percent 66.1 \varepsilon_{\rm cont}^{\rm noP_{c}}=66.1\% italic_ε start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_noP start_POSTSUBSCRIPT roman_c end_POSTSUBSCRIPT end_POSTSUPERSCRIPT = 66.1 % is the efficiency obtained from no-P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT MC simulation of continuum
production, ( 1 + δ ISR ) = 0.82 1 subscript 𝛿 ISR 0.82 (1+\delta_{\rm ISR})=0.82 ( 1 + italic_δ start_POSTSUBSCRIPT roman_ISR end_POSTSUBSCRIPT ) = 0.82 is the radiative correction factor [30 , 31 ] , and ℬ ( J / ψ → ℓ + ℓ − ) = ( 11.93 ± 0.07 ) % ℬ → 𝐽 𝜓 superscript ℓ superscript ℓ percent plus-or-minus 11.93 0.07 {\cal B}(J/\psi\to\ell^{+}\ell^{-})=(11.93\pm 0.07)\% caligraphic_B ( italic_J / italic_ψ → roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ) = ( 11.93 ± 0.07 ) % is the branching fraction of J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ decaying to e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT or μ + μ − superscript 𝜇 superscript 𝜇 \mu^{+}\mu^{-} italic_μ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_μ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT [29 ] . We obtain the
cross section σ ( e + e − → p J / ψ + a n y t h i n g ) = ( 57.5 ± 2.1 ± 2.5 ) fb 𝜎 → superscript 𝑒 superscript 𝑒 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 plus-or-minus 57.5 2.1 2.5 femtobarn \sigma(e^{+}e^{-}\to pJ/\psi+anything)=(57.5\pm 2.1\pm 2.5)~{}$\mathrm{fb}$ italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) = ( 57.5 ± 2.1 ± 2.5 ) roman_fb at s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV , where the
systematic uncertainties are discussed in Sec. V .
Figure 5 shows the combined distributions of Figs. 4 (a-d) and 4 (e-h)
for Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) and Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) inclusive decays, respectively. Since we measure the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ production in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S )
inclusive decays, the background of continuum production in Fig. 5 is removed. We estimate the number
of backgrounds from K ± J / ψ superscript 𝐾 plus-or-minus 𝐽 𝜓 K^{\pm}J/\psi italic_K start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT italic_J / italic_ψ or π ± J / ψ superscript 𝜋 plus-or-minus 𝐽 𝜓 \pi^{\pm}J/\psi italic_π start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT italic_J / italic_ψ to be 17.9 ± 1.2 plus-or-minus 17.9 1.2 17.9\pm 1.2 17.9 ± 1.2 (43.9 ± 3.0 plus-or-minus 43.9 3.0 43.9\pm 3.0 43.9 ± 3.0 ) in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) [Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) ] decays.
With the backgrounds estimated from the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sidebands and those from mis-identification of proton being
subtracted, we get the final numbers of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ signal events to be N Υ ( 1 S ) p J / ψ = 363 ± 19 subscript superscript 𝑁 𝑝 𝐽 𝜓 Υ 1 𝑆 plus-or-minus 363 19 N^{pJ/\psi}_{\Upsilon(1S)}=363\pm 19 italic_N start_POSTSUPERSCRIPT italic_p italic_J / italic_ψ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_Υ ( 1 italic_S ) end_POSTSUBSCRIPT = 363 ± 19 in the
Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays and N Υ ( 2 S ) p J / ψ = 541 ± 23 subscript superscript 𝑁 𝑝 𝐽 𝜓 Υ 2 𝑆 plus-or-minus 541 23 N^{pJ/\psi}_{\Upsilon(2S)}=541\pm 23 italic_N start_POSTSUPERSCRIPT italic_p italic_J / italic_ψ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_Υ ( 2 italic_S ) end_POSTSUBSCRIPT = 541 ± 23 in the Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays. These yields are much higher than
those estimated to be due to the underlying e + e − superscript 𝑒 superscript 𝑒 e^{+}e^{-} italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT continuum production. To measure the production of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ in
Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays, we use the no-P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT MC samples to estimate the efficiencies to be
ε Υ ( 1 , 2 S ) noP c = 64.8 % superscript subscript 𝜀 Υ 1 2 𝑆 subscript noP c percent 64.8 \varepsilon_{\Upsilon(1,2S)}^{\rm noP_{c}}=64.8\% italic_ε start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_noP start_POSTSUBSCRIPT roman_c end_POSTSUBSCRIPT end_POSTSUPERSCRIPT = 64.8 % and 65.1% for the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) and Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) inclusive decays in the region of
4.0 GeV / c 2 < M p J / ψ < 5.5 GeV / c 2 4.0 GeV superscript 𝑐 2 subscript 𝑀 𝑝 𝐽 𝜓 5.5 GeV superscript 𝑐 2 4.0~{}\hbox{GeV}/c^{2}<M_{pJ/\psi}<5.5~{}\hbox{GeV}/c^{2} 4.0 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT < italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT < 5.5 GeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT . We calculate the branching fractions of Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays using
ℬ [ Υ ( 1 , 2 S ) → p J / ψ + a n y t h i n g ] = N Υ ( 1 , 2 S ) p J / ψ − f scale × N cont p J / ψ N Υ ( 1 , 2 S ) × ε Υ ( 1 , 2 S ) noPc × ℬ ( J / ψ → ℓ + ℓ − ) , ℬ delimited-[] → Υ 1 2 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 subscript superscript 𝑁 𝑝 𝐽 𝜓 Υ 1 2 𝑆 subscript 𝑓 scale subscript superscript 𝑁 𝑝 𝐽 𝜓 cont subscript 𝑁 Υ 1 2 𝑆 superscript subscript 𝜀 Υ 1 2 𝑆 noPc ℬ → 𝐽 𝜓 superscript ℓ superscript ℓ {\cal B}[\Upsilon(1,2S)\to pJ/\psi+anything]=\frac{N^{pJ/\psi}_{\Upsilon(1,2S)%
}-f_{\rm scale}\times N^{pJ/\psi}_{\rm cont}}{N_{\Upsilon(1,2S)}\times%
\varepsilon_{\Upsilon(1,2S)}^{\rm noPc}\times{\cal B}(J/\psi\to\ell^{+}\ell^{-%
})}, caligraphic_B [ roman_Υ ( 1 , 2 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] = divide start_ARG italic_N start_POSTSUPERSCRIPT italic_p italic_J / italic_ψ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT - italic_f start_POSTSUBSCRIPT roman_scale end_POSTSUBSCRIPT × italic_N start_POSTSUPERSCRIPT italic_p italic_J / italic_ψ end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_cont end_POSTSUBSCRIPT end_ARG start_ARG italic_N start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT × italic_ε start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_noPc end_POSTSUPERSCRIPT × caligraphic_B ( italic_J / italic_ψ → roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ) end_ARG ,
(2)
where N Υ ( 1 , 2 S ) subscript 𝑁 Υ 1 2 𝑆 N_{\Upsilon(1,2S)} italic_N start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT are the numbers of Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) events in the Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) data samples. We obtain that
ℬ [ Υ ( 1 S ) → p J / ψ + a n y t h i n g ] = ( 4.27 ± 0.16 ± 0.20 ) × 10 − 5 ℬ delimited-[] → Υ 1 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 plus-or-minus 4.27 0.16 0.20 superscript 10 5 {\cal B}[\Upsilon(1S)\to pJ/\psi+anything]=(4.27\pm 0.16\pm 0.20)\times 10^{-5} caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] = ( 4.27 ± 0.16 ± 0.20 ) × 10 start_POSTSUPERSCRIPT - 5 end_POSTSUPERSCRIPT and ℬ [ Υ ( 2 S ) → p J / ψ + a n y t h i n g ] = ( 3.59 ± 0.14 ± 0.16 ) × 10 − 5 ℬ delimited-[] → Υ 2 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 plus-or-minus 3.59 0.14 0.16 superscript 10 5 {\cal B}[\Upsilon(2S)\to pJ/\psi+anything]=(3.59\pm 0.14\pm 0.16)\times 10^{-5} caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] = ( 3.59 ± 0.14 ± 0.16 ) × 10 start_POSTSUPERSCRIPT - 5 end_POSTSUPERSCRIPT for the first time. Excluding the background of Υ ( 2 S ) → Υ ( 1 S ) + a n y t h i n g → Υ 2 𝑆 Υ 1 𝑆 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 \Upsilon(2S)\to\Upsilon(1S)+anything roman_Υ ( 2 italic_S ) → roman_Υ ( 1 italic_S ) + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g transitions and Υ ( 1 S ) → p J / ψ + a n y t h i n g → Υ 1 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 \Upsilon(1S)\to pJ/\psi+anything roman_Υ ( 1 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g with this measurement of ℬ [ Υ ( 1 S ) → p J / ψ + a n y t h i n g ] ℬ delimited-[] → Υ 1 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 {\cal B}[\Upsilon(1S)\to pJ/\psi+anything] caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ,
we correct the ℬ [ Υ ( 2 S ) → p J / ψ + a n y t h i n g ] ℬ delimited-[] → Υ 2 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 {\cal B}[\Upsilon(2S)\to pJ/\psi+anything] caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] and find a value of ( 2.46 ± 0.15 ± 0.17 ) × 10 − 5 plus-or-minus 2.46 0.15 0.17 superscript 10 5 (2.46\pm 0.15\pm 0.17)\times 10^{-5} ( 2.46 ± 0.15 ± 0.17 ) × 10 start_POSTSUPERSCRIPT - 5 end_POSTSUPERSCRIPT .
Systematic uncertainties are listed in Table 3 , which is described in Sec. V . The world
average values of the branching fractions of J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ production in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays are ℬ [ Υ ( 1 S ) → J / ψ + a n y t h i n g ] = ( 5.4 ± 0.4 ) × 10 − 4 ℬ delimited-[] → Υ 1 𝑆 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 plus-or-minus 5.4 0.4 superscript 10 4 {\cal B}[\Upsilon(1S)\to J/\psi+anything]=(5.4\pm 0.4)\times 10^{-4} caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] = ( 5.4 ± 0.4 ) × 10 start_POSTSUPERSCRIPT - 4 end_POSTSUPERSCRIPT and ℬ [ Υ ( 2 S ) → J / ψ + a n y t h i n g ] < 6 × 10 − 3 ℬ delimited-[] → Υ 2 𝑆 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 6 superscript 10 3 {\cal B}[\Upsilon(2S)\to J/\psi+anything]<6\times 10^{-3} caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] < 6 × 10 start_POSTSUPERSCRIPT - 3 end_POSTSUPERSCRIPT at 90%
credibility [29 ] . Thus, the ratio ℬ ( Υ → p J / ψ + a n y t h i n g ) / ℬ ( Υ → J / ψ + a n y t h i n g ) ℬ → Υ 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ → Υ 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 {\cal B}(\Upsilon\to pJ/\psi+anything)/{\cal B}(\Upsilon\to J/\psi+anything) caligraphic_B ( roman_Υ → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) / caligraphic_B ( roman_Υ → italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) is
of order 10 − 1 − 10 − 2 superscript 10 1 superscript 10 2 10^{-1}-10^{-2} 10 start_POSTSUPERSCRIPT - 1 end_POSTSUPERSCRIPT - 10 start_POSTSUPERSCRIPT - 2 end_POSTSUPERSCRIPT in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays.
Figure 5: The combined distributions of the invariant masses of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ and p ¯ J / ψ ¯ 𝑝 𝐽 𝜓 \bar{p}J/\psi over¯ start_ARG italic_p end_ARG italic_J / italic_ψ from (a) the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) inclusive
decays and (b) the Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) inclusive decays, and the fit results including P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT . The dots with
error bars are data. The shaded histograms are the backgrounds estimated from the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sidebands. The blue
histograms are the best fit results; the green histograms are the P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT components; the pink
histograms are the no-P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT components.
To estimate the production of a possible P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT state in the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) or Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) inclusive decays, we perform binned
maximum likelihood fits to the distribution of M p J / ψ subscript 𝑀 𝑝 𝐽 𝜓 M_{pJ/\psi} italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT in Fig. 5 (a) or 5 (b) with
f PDF = f P c ( 4312 ) + + f P c ( 4440 ) + + f P c ( 4457 ) + + f noP c + f bkg , subscript 𝑓 PDF subscript 𝑓 subscript 𝑃 𝑐 superscript 4312 subscript 𝑓 subscript 𝑃 𝑐 superscript 4440 subscript 𝑓 subscript 𝑃 𝑐 superscript 4457 subscript 𝑓 subscript noP c subscript 𝑓 bkg f_{\rm PDF}=f_{P_{c}(4312)^{+}}+f_{P_{c}(4440)^{+}}+f_{P_{c}(4457)^{+}}+f_{\rm
noP%
_{c}}+f_{\rm bkg}, italic_f start_POSTSUBSCRIPT roman_PDF end_POSTSUBSCRIPT = italic_f start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT + italic_f start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT + italic_f start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT + italic_f start_POSTSUBSCRIPT roman_noP start_POSTSUBSCRIPT roman_c end_POSTSUBSCRIPT end_POSTSUBSCRIPT + italic_f start_POSTSUBSCRIPT roman_bkg end_POSTSUBSCRIPT ,
(3)
where f P c ( 4312 ) + subscript 𝑓 subscript 𝑃 𝑐 superscript 4312 f_{P_{c}(4312)^{+}} italic_f start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT , f P c ( 4440 ) + subscript 𝑓 subscript 𝑃 𝑐 superscript 4440 f_{P_{c}(4440)^{+}} italic_f start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT , f P c ( 4457 ) + subscript 𝑓 subscript 𝑃 𝑐 superscript 4457 f_{P_{c}(4457)^{+}} italic_f start_POSTSUBSCRIPT italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT end_POSTSUBSCRIPT , and f noP c subscript 𝑓 subscript noP c f_{\rm noP_{c}} italic_f start_POSTSUBSCRIPT roman_noP start_POSTSUBSCRIPT roman_c end_POSTSUBSCRIPT end_POSTSUBSCRIPT are the histogram PDFs obtained from the signal MC
simulations on P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , and the no-P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT MC simulation. We use a second-order polynomial function for
the f bkg subscript 𝑓 bkg f_{\rm bkg} italic_f start_POSTSUBSCRIPT roman_bkg end_POSTSUBSCRIPT to describe the backgrounds due to J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ selection. We fit to the events from the J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal
region with f PDF subscript 𝑓 PDF f_{\rm PDF} italic_f start_POSTSUBSCRIPT roman_PDF end_POSTSUBSCRIPT and the events from J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass sidebands with f bkg subscript 𝑓 bkg f_{\rm bkg} italic_f start_POSTSUBSCRIPT roman_bkg end_POSTSUBSCRIPT simultaneously. The fit
yields the numbers of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signals [N fit A ( P c + ) superscript subscript 𝑁 fit A superscript subscript 𝑃 𝑐 N_{\rm fit}^{\rm A}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) ], as listed in Table 2 . Since none of the
P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , or P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT is significant, we integrate the likelihood versus the N fit A ( P c + ) superscript subscript 𝑁 fit A superscript subscript 𝑃 𝑐 N_{\rm fit}^{\rm A}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) and
determine the upper limits N fit A , UL ( P c + ) superscript subscript 𝑁 fit A UL
superscript subscript 𝑃 𝑐 N_{\rm fit}^{\rm A,UL}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A , roman_UL end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) at 90% credibility. We also perform a fit to the
M p J / ψ subscript 𝑀 𝑝 𝐽 𝜓 M_{pJ/\psi} italic_M start_POSTSUBSCRIPT italic_p italic_J / italic_ψ end_POSTSUBSCRIPT distribution in Fig. 5 (a) or 5 (b) with individual P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT state in the
f PDF subscript 𝑓 PDF f_{\rm PDF} italic_f start_POSTSUBSCRIPT roman_PDF end_POSTSUBSCRIPT , which yields the new number of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signal [N fit B ( P c + ) superscript subscript 𝑁 fit B superscript subscript 𝑃 𝑐 N_{\rm fit}^{\rm B}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_B end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) ]. Similarly, we determine the
related upper limits N fit B , UL ( P c + ) superscript subscript 𝑁 fit B UL
superscript subscript 𝑃 𝑐 N_{\rm fit}^{\rm B,UL}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_B , roman_UL end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) for P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT at 90% credibility. We also
estimate the upper limits by varying the masses and widths of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT states by 1 σ 1 𝜎 1\sigma 1 italic_σ in these tests. We take the largest values of the upper limits as the conservative estimations of the upper limits of the
numbers of the P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signals [N sig UL ( P c + ) subscript superscript 𝑁 UL sig superscript subscript 𝑃 𝑐 N^{\rm UL}_{\rm sig}(P_{c}^{+}) italic_N start_POSTSUPERSCRIPT roman_UL end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_sig end_POSTSUBSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) ] in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays. We then calculate the
upper limit of the branching fraction of a P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT state produced in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) [Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) ] inclusive decays at 90%
credibility with
ℬ [ Υ ( 1 , 2 S ) → P c + + a n y t h i n g ] ⋅ ℬ ( P c + → p J / ψ ) < N sig UL ( P c + ) N Υ ( 1 , 2 S ) ⋅ ε allcmb MC ⋅ ℬ ( J / ψ → ℓ + ℓ − ) ( 1 − δ sys ) , ⋅ ℬ delimited-[] → Υ 1 2 𝑆 superscript subscript 𝑃 𝑐 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ → superscript subscript 𝑃 𝑐 𝑝 𝐽 𝜓 subscript superscript 𝑁 UL sig superscript subscript 𝑃 𝑐 ⋅ subscript 𝑁 Υ 1 2 𝑆 subscript superscript 𝜀 MC allcmb ℬ → 𝐽 𝜓 superscript ℓ superscript ℓ 1 subscript 𝛿 sys {\cal B}[\Upsilon(1,2S)\to P_{c}^{+}+anything]\cdot{\cal B}(P_{c}^{+}\to pJ/%
\psi)<\frac{N^{\rm UL}_{\rm sig}(P_{c}^{+})}{N_{\Upsilon(1,2S)}\cdot%
\varepsilon^{\rm MC}_{\rm allcmb}\cdot{\cal B}(J/\psi\to\ell^{+}\ell^{-})(1-%
\delta_{\rm sys})}, caligraphic_B [ roman_Υ ( 1 , 2 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ) < divide start_ARG italic_N start_POSTSUPERSCRIPT roman_UL end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_sig end_POSTSUBSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) end_ARG start_ARG italic_N start_POSTSUBSCRIPT roman_Υ ( 1 , 2 italic_S ) end_POSTSUBSCRIPT ⋅ italic_ε start_POSTSUPERSCRIPT roman_MC end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_allcmb end_POSTSUBSCRIPT ⋅ caligraphic_B ( italic_J / italic_ψ → roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ) ( 1 - italic_δ start_POSTSUBSCRIPT roman_sys end_POSTSUBSCRIPT ) end_ARG ,
(4)
where δ sys = 5.0 % ( 4.7 % ) subscript 𝛿 sys percent 5.0 percent 4.7 \delta_{\rm sys}=5.0\%~{}(4.7\%) italic_δ start_POSTSUBSCRIPT roman_sys end_POSTSUBSCRIPT = 5.0 % ( 4.7 % ) is the systematic uncertainty of Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) [Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) ] decays, which are
described in Sec. V . We summarize the values of N fit A ( P c + ) superscript subscript 𝑁 fit A superscript subscript 𝑃 𝑐 N_{\rm fit}^{\rm A}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) , N fit A , UL ( P c + ) superscript subscript 𝑁 fit A UL
superscript subscript 𝑃 𝑐 N_{\rm fit}^{\rm A,UL}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A , roman_UL end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) ,
N fit B ( P c + ) superscript subscript 𝑁 fit 𝐵 superscript subscript 𝑃 𝑐 N_{\rm fit}^{B}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT italic_B end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) , N fit B , UL ( P c + ) superscript subscript 𝑁 fit B UL
superscript subscript 𝑃 𝑐 N_{\rm fit}^{\rm B,UL}(P_{c}^{+}) italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_B , roman_UL end_POSTSUPERSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) , N sig UL ( P c + ) subscript superscript 𝑁 UL sig superscript subscript 𝑃 𝑐 N^{\rm UL}_{\rm sig}(P_{c}^{+}) italic_N start_POSTSUPERSCRIPT roman_UL end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_sig end_POSTSUBSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) , and the upper limit of
ℬ [ Υ ( 1 , 2 S ) → P c + + a n y t h i n g ] ⋅ ℬ ( P c + → p J / ψ ) ⋅ ℬ delimited-[] → Υ 1 2 𝑆 superscript subscript 𝑃 𝑐 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ → superscript subscript 𝑃 𝑐 𝑝 𝐽 𝜓 {\cal B}[\Upsilon(1,2S)\to P_{c}^{+}+anything]\cdot{\cal B}(P_{c}^{+}\to pJ/\psi) caligraphic_B [ roman_Υ ( 1 , 2 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ) at 90% credibility in Table 2 .
Table 2: The fit results and the upper limits of P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT productions in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive
decays. N fit A superscript subscript 𝑁 fit A N_{\rm fit}^{\rm A} italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A end_POSTSUPERSCRIPT is the number of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signals in the fit with the PDF function f PDF subscript 𝑓 PDF f_{\rm PDF} italic_f start_POSTSUBSCRIPT roman_PDF end_POSTSUBSCRIPT contains
P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT states, and N fit A , UL superscript subscript 𝑁 fit A UL
N_{\rm fit}^{\rm A,UL} italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A , roman_UL end_POSTSUPERSCRIPT is the related upper limits at 90% credibility.
N fit B superscript subscript 𝑁 fit B N_{\rm fit}^{\rm B} italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_B end_POSTSUPERSCRIPT is the number of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signals in the fit with the PDF function that contains only a single P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT state, and N fit B , UL superscript subscript 𝑁 fit B UL
N_{\rm fit}^{\rm B,UL} italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_B , roman_UL end_POSTSUPERSCRIPT is the related upper limits at 90% credibility.
N sig UL subscript superscript 𝑁 UL sig N^{\rm UL}_{\rm sig} italic_N start_POSTSUPERSCRIPT roman_UL end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_sig end_POSTSUBSCRIPT is the final conservative estimation of the upper limit of the number of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signals in
Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays. ℬ UL superscript ℬ UL {\cal B}^{\rm UL} caligraphic_B start_POSTSUPERSCRIPT roman_UL end_POSTSUPERSCRIPT is the upper limit of ℬ ( Υ → P c + + a n y t h i n g ) ⋅ ℬ ( P c + → p J / ψ ) ⋅ ℬ → Υ superscript subscript 𝑃 𝑐 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ → superscript subscript 𝑃 𝑐 𝑝 𝐽 𝜓 {\cal B}(\Upsilon\to P_{c}^{+}+anything)\cdot{\cal B}(P_{c}^{+}\to pJ/\psi) caligraphic_B ( roman_Υ → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) ⋅ caligraphic_B ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ) at 90% credibility.
Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays
Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays
—
P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
N fit A superscript subscript 𝑁 fit A N_{\rm fit}^{\rm A} italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A end_POSTSUPERSCRIPT
10 ± 8 plus-or-minus 10 8 10\pm 8 10 ± 8
14 ± 12 plus-or-minus 14 12 14\pm 12 14 ± 12
− 3 ± 9 plus-or-minus 3 9 -3\pm 9 - 3 ± 9
30 ± 16 plus-or-minus 30 16 30\pm 16 30 ± 16
33 ± 15 plus-or-minus 33 15 33\pm 15 33 ± 15
0 ± 3 plus-or-minus 0 3 0\pm 3 0 ± 3
N fit A , UL superscript subscript 𝑁 fit A UL
N_{\rm fit}^{\rm A,UL} italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_A , roman_UL end_POSTSUPERSCRIPT
26
37
14
52
60
6
N fit B superscript subscript 𝑁 fit B N_{\rm fit}^{\rm B} italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_B end_POSTSUPERSCRIPT
10 ± 8 plus-or-minus 10 8 10\pm 8 10 ± 8
12 ± 11 plus-or-minus 12 11 12\pm 11 12 ± 11
3 ± 9 plus-or-minus 3 9 3\pm 9 3 ± 9
29 ± 12 plus-or-minus 29 12 29\pm 12 29 ± 12
31 ± 15 plus-or-minus 31 15 31\pm 15 31 ± 15
0 ± 3 plus-or-minus 0 3 0\pm 3 0 ± 3
N fit B , UL superscript subscript 𝑁 fit B UL
N_{\rm fit}^{\rm B,UL} italic_N start_POSTSUBSCRIPT roman_fit end_POSTSUBSCRIPT start_POSTSUPERSCRIPT roman_B , roman_UL end_POSTSUPERSCRIPT
26
33
17
50
57
7
N sig UL subscript superscript 𝑁 UL sig N^{\rm UL}_{\rm sig} italic_N start_POSTSUPERSCRIPT roman_UL end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_sig end_POSTSUBSCRIPT
31
47
34
56
77
26
ℬ UL superscript ℬ UL {\cal B}^{\rm UL} caligraphic_B start_POSTSUPERSCRIPT roman_UL end_POSTSUPERSCRIPT ( × 10 − 6 absent superscript 10 6 \times 10^{-6} × 10 start_POSTSUPERSCRIPT - 6 end_POSTSUPERSCRIPT )
4.5
6.8
4.9
5.3
7.2
2.4
V Systematic uncertainties
Table 3: The summary of the systematic uncertainties (%) in the measurements of ℬ [ Υ ( 1 , 2 S ) → p J / ψ + a n y t h i n g ] ℬ delimited-[] → Υ 1 2 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 {\cal B}[\Upsilon(1,2S)\to pJ/\psi+anything] caligraphic_B [ roman_Υ ( 1 , 2 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] and σ ( e + e − → p J / ψ + a n y t h i n g ) 𝜎 → superscript 𝑒 superscript 𝑒 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 \sigma(e^{+}e^{-}\to pJ/\psi+anything) italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) at s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV .
Source
Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decay
Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decay
σ ( e + e − → p J / ψ + a n y t h i n g ) 𝜎 → superscript 𝑒 superscript 𝑒 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 \sigma(e^{+}e^{-}\to pJ/\psi+anything) italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g )
Particle identification
2.1
2.1
2.1
Tracking
1.1
1.1
1.1
J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal region
0.6
0.5
0.4
M recoil 2 ( p J / ψ ) superscript subscript 𝑀 recoil 2 𝑝 𝐽 𝜓 M_{\rm recoil}^{2}(pJ/\psi) italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p italic_J / italic_ψ ) requirement
1.5
1.5
1.5
ℬ ( J / ψ → ℓ + ℓ − ) ℬ → 𝐽 𝜓 superscript ℓ superscript ℓ {\cal B}(J/\psi\to\ell^{+}\ell^{-}) caligraphic_B ( italic_J / italic_ψ → roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT )
0.6
0.6
0.6
1 + δ ISR 1 subscript 𝛿 ISR 1+\delta_{\rm ISR} 1 + italic_δ start_POSTSUBSCRIPT roman_ISR end_POSTSUBSCRIPT
—
—
1.0
Modeling in MC simulation
2.8
2.3
2.6
Number of Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) events
2.2
2.3
—
Integrated luminosity
—
—
1.4
Statistics of MC samples
0.5
0.5
0.5
Sum in quadrature
4.6
4.4
4.3
As listed in Table 3 , we consider the following systematic uncertainties in determining the branching
fractions ℬ [ Υ ( 1 , 2 S ) → p J / ψ + a n y t h i n g ] ℬ delimited-[] → Υ 1 2 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 {\cal B}[\Upsilon(1,2S)\to pJ/\psi+anything] caligraphic_B [ roman_Υ ( 1 , 2 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] and measuring σ ( e + e − → p J / ψ + a n y t h i n g ) 𝜎 → superscript 𝑒 superscript 𝑒 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 \sigma(e^{+}e^{-}\to pJ/\psi+anything) italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) at s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV : particle identification, tracking efficiency, J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal region, M recoil 2 ( p J / ψ ) superscript subscript 𝑀 recoil 2 𝑝 𝐽 𝜓 M_{\rm recoil}^{2}(pJ/\psi) italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p italic_J / italic_ψ ) requirement,
branching fraction of J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ decay, number of Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) events, integrated luminosity, modeling in MC simulation,
and statistics of MC samples, etc. The uncertainties due to the lepton identification are 2.0% and 0.5% for e ± superscript 𝑒 plus-or-minus e^{\pm} italic_e start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT
and μ ± superscript 𝜇 plus-or-minus \mu^{\pm} italic_μ start_POSTSUPERSCRIPT ± end_POSTSUPERSCRIPT , respectively. For the proton identification, we have applied an efficiency correction according to the
momentum and angle in the laboratory frame. Shifting the correction factor by 1 σ 1 𝜎 1\sigma 1 italic_σ , we get the related
efficiency difference of 0.43% and take 0.5% to be the systematic uncertainty of proton identification. Therefore,
the total systematic uncertainty due to the particle identification is 2.1%. In estimating the backgrounds from
K J / ψ 𝐾 𝐽 𝜓 KJ/\psi italic_K italic_J / italic_ψ or π J / ψ 𝜋 𝐽 𝜓 \pi J/\psi italic_π italic_J / italic_ψ , the mis-identification of K ( π ) 𝐾 𝜋 K(\pi) italic_K ( italic_π ) to p 𝑝 p italic_p is (1.98 ± 0.07 ) % [ ( 0.72 ± 0.02 ) % ] 1.98\pm 0.07)\%~{}[(0.72\pm 0.02)\%] 1.98 ± 0.07 ) % [ ( 0.72 ± 0.02 ) % ] . The
uncertainties of mis-identification are not listed in Table 3 but contribute 0.4, 1.1, 1.3 in the numbers
of estimated backgrounds from K J / ψ 𝐾 𝐽 𝜓 KJ/\psi italic_K italic_J / italic_ψ and π J / ψ 𝜋 𝐽 𝜓 \pi J/\psi italic_π italic_J / italic_ψ in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays, Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays, and continuum productions.
The uncertainty due to the tracking efficiency is 0.35% per track and adds linearly. Fitting the M ℓ + ℓ − subscript 𝑀 superscript ℓ superscript ℓ M_{\ell^{+}\ell^{-}} italic_M start_POSTSUBSCRIPT roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT end_POSTSUBSCRIPT
distributions from data and MC simulations with a Gaussian function for J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ signal and a second-order Chebychev
function for backgrounds, we obtain the efficiencies of J / ψ 𝐽 𝜓 J/\psi italic_J / italic_ψ mass signal window to be ε J / ψ data = ( 99.43 ± 0.58 ) % subscript superscript 𝜀 data 𝐽 𝜓 percent plus-or-minus 99.43 0.58 \varepsilon^{\rm data}_{J/\psi}=(99.43\pm 0.58)\% italic_ε start_POSTSUPERSCRIPT roman_data end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT = ( 99.43 ± 0.58 ) % , ( 99.56 ± 0.48 ) % percent plus-or-minus 99.56 0.48 (99.56\pm 0.48)\% ( 99.56 ± 0.48 ) % , and ( 99.69 ± 0.37 ) plus-or-minus 99.69 0.37 (99.69\pm 0.37) ( 99.69 ± 0.37 ) % in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays, Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays, and
continuum productions in data, and ε J / ψ MC = 99.9 % subscript superscript 𝜀 MC 𝐽 𝜓 percent 99.9 \varepsilon^{\rm MC}_{J/\psi}=99.9\% italic_ε start_POSTSUPERSCRIPT roman_MC end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT = 99.9 % in the signal MC simulations. We correct the
efficiencies by the ratios ε J / ψ data / ε J / ψ MC subscript superscript 𝜀 data 𝐽 𝜓 subscript superscript 𝜀 MC 𝐽 𝜓 \varepsilon^{\rm data}_{J/\psi}/\varepsilon^{\rm MC}_{J/\psi} italic_ε start_POSTSUPERSCRIPT roman_data end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT / italic_ε start_POSTSUPERSCRIPT roman_MC end_POSTSUPERSCRIPT start_POSTSUBSCRIPT italic_J / italic_ψ end_POSTSUBSCRIPT , and take the errors of the ratios to be
the systematic uncertainties, i.e., 0.6% in the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays, 0.5% in the Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays and 0.4% in the
continuum productions. The efficiencies of the requirement M recoil 2 ( p J / ψ ) > 10 GeV 2 / c 4 superscript subscript 𝑀 recoil 2 𝑝 𝐽 𝜓 10 superscript GeV 2 superscript 𝑐 4 M_{\rm recoil}^{2}(pJ/\psi)>10~{}\hbox{GeV}^{2}/c^{4} italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p italic_J / italic_ψ ) > 10 GeV start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT / italic_c start_POSTSUPERSCRIPT 4 end_POSTSUPERSCRIPT are 98.9%, 99.9%, and 99.9%
in signal MC simulations for the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays, the Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays, and the continuum productions. Since all the
MC simulations of the P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT decay model and no-P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT process show efficiencies higher than 98.5%, we take 1.5% as
the systematic uncertainty of the requirement on M recoil 2 ( p J / ψ ) superscript subscript 𝑀 recoil 2 𝑝 𝐽 𝜓 M_{\rm recoil}^{2}(pJ/\psi) italic_M start_POSTSUBSCRIPT roman_recoil end_POSTSUBSCRIPT start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ( italic_p italic_J / italic_ψ ) . According to the world average values [29 ] ,
ℬ ( J / ψ → ℓ + ℓ − ) ( l = e , μ ) ℬ → 𝐽 𝜓 superscript ℓ superscript ℓ 𝑙 𝑒 𝜇
{\cal B}(J/\psi\to\ell^{+}\ell^{-})(l=e,\mu) caligraphic_B ( italic_J / italic_ψ → roman_ℓ start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT roman_ℓ start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT ) ( italic_l = italic_e , italic_μ ) contributes a systematic uncertainty of 0.6%. By varying the photon energy cutoff by
50 MeV 50 MeV 50~{}\hbox{MeV} 50 MeV in the simulation of ISR, we determine the change of 1 + δ ISR 1 subscript 𝛿 ISR 1+\delta_{\rm ISR} 1 + italic_δ start_POSTSUBSCRIPT roman_ISR end_POSTSUBSCRIPT to be 0.01 and take 1.0% to be
the conservative systematic uncertainty in measuring the cross section σ ( e + e − → p J / ψ + a n y t h i n g ) 𝜎 → superscript 𝑒 superscript 𝑒 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 \sigma(e^{+}e^{-}\to pJ/\psi+anything) italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) at
s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV . There are uncertainties in modeling the final states in the MC simulations. In the
hadronization of q q ¯ 𝑞 ¯ 𝑞 q\bar{q} italic_q over¯ start_ARG italic_q end_ARG , we vary the mass and width of X 𝑋 X italic_X by 200 MeV / c 2 200 MeV superscript 𝑐 2 200~{}\hbox{MeV}/c^{2} 200 MeV / italic_c start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT and 500 MeV 500 MeV 500~{}\hbox{MeV} 500 MeV , which have differences in
efficiency that 2.6% in the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays, 1.9% in the Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays and 2.3% in the continuum production,
respectively. Considering that the proton candidate may come from Λ Λ \Lambda roman_Λ decay, we simulate the MC samples of Υ ( 1 , 2 S ) → p J / ψ + Λ ¯ + ( s q ¯ ) → Υ 1 2 𝑆 𝑝 𝐽 𝜓 ¯ Λ 𝑠 ¯ 𝑞 \Upsilon(1,2S)\to pJ/\psi+\bar{\Lambda}+(s\bar{q}) roman_Υ ( 1 , 2 italic_S ) → italic_p italic_J / italic_ψ + over¯ start_ARG roman_Λ end_ARG + ( italic_s over¯ start_ARG italic_q end_ARG ) and find the efficiency differences, from those of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signal MC samples, of 1.1% in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays, 1.2% in Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays, and 1.1% in contnuum production. We sum the two sources and obtain the systematic uncertainties
in modeling the final states in MC simulations to be to be 2.8%, 2.3%, and 2.6% in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays, Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S )
decays, and continuum productions at s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV . The uncertainties of the total numbers of Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S )
events and Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) events are 2.2% and 2.3% in the Belle data samples [20 , 21 ] . The common
uncertainty in the integrated luminosities for the Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) , Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) , and continuum data samples is 1.4%, which is canceled in calculating the scale factor f scale subscript 𝑓 scale f_{\rm scale} italic_f start_POSTSUBSCRIPT roman_scale end_POSTSUBSCRIPT . The statistical uncertainties of the signal MC samples are
0.5% in common. Assuming these uncertainties are independent and sum them in quadrature, we obtain the total
systematic uncertainties to be 4.6% in ℬ [ Υ ( 1 S ) → p J / ψ + a n y t h i n g ] ℬ delimited-[] → Υ 1 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 {\cal B}[\Upsilon(1S)\to pJ/\psi+anything] caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] , 4.4% in ℬ [ Υ ( 2 S ) → p J / ψ + a n y t h i n g ] ℬ delimited-[] → Υ 2 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 {\cal B}[\Upsilon(2S)\to pJ/\psi+anything] caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] , and 4.3% in σ ( e + e − → p J / ψ + a n y t h i n g ) 𝜎 → superscript 𝑒 superscript 𝑒 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 \sigma(e^{+}e^{-}\to pJ/\psi+anything) italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) at s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV .
In determining the upper limits of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT productions in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) decays, most of the systematic uncertainties are
the same as those listed in Table 3 , with the exception of the modeling of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ in signal MC
simulations and additional uncertainties in fits. To evaluate these, we do the similar studies, including varying the
mass and width of X → q q ¯ → 𝑋 𝑞 ¯ 𝑞 X\to q\bar{q} italic_X → italic_q over¯ start_ARG italic_q end_ARG and simulating the MC sample of Υ ( 1 , 2 S ) → P c + + Λ ¯ + ( s q ¯ ) → Υ 1 2 𝑆 superscript subscript 𝑃 𝑐 ¯ Λ 𝑠 ¯ 𝑞 \Upsilon(1,2S)\to P_{c}^{+}+\bar{\Lambda}+(s~{}\bar{q}) roman_Υ ( 1 , 2 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + over¯ start_ARG roman_Λ end_ARG + ( italic_s over¯ start_ARG italic_q end_ARG ) . We
replace the uncertainties in modeling by 3.3% in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays and 2.9% in Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays in
Table 3 . Therefore, the total systematic uncertainties of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT productions in Υ ( 1 S ) Υ 1 𝑆 \Upsilon(1S) roman_Υ ( 1 italic_S ) decays and
Υ ( 2 S ) Υ 2 𝑆 \Upsilon(2S) roman_Υ ( 2 italic_S ) decays are 5.0% and 4.7%, respectively. To estimate the systematic uncertainty of f noP c subscript 𝑓 subscript noP c f_{\rm noP_{c}} italic_f start_POSTSUBSCRIPT roman_noP start_POSTSUBSCRIPT roman_c end_POSTSUBSCRIPT end_POSTSUBSCRIPT in the
fits, we investigate the difference in the yield when using an ARGUS function to replace the histogram PDF obtained
from the no-P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT MC simulation [32 ] . We change the masses and the widths of the P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT states by
1 σ 1 𝜎 1\sigma 1 italic_σ according to LHCb measurement [15 ] . As before, we take the highest values of
N sig UL ( P c + ) subscript superscript 𝑁 UL sig superscript subscript 𝑃 𝑐 N^{\rm UL}_{\rm sig}(P_{c}^{+}) italic_N start_POSTSUPERSCRIPT roman_UL end_POSTSUPERSCRIPT start_POSTSUBSCRIPT roman_sig end_POSTSUBSCRIPT ( italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT ) to calculate the upper limit of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT production in the Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays.
VI Summary
We study the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ final states in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays and search for the P c ( 4312 ) + subscript 𝑃 𝑐 superscript 4312 P_{c}(4312)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT , P c ( 4440 ) + subscript 𝑃 𝑐 superscript 4440 P_{c}(4440)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT and P c ( 4457 ) + subscript 𝑃 𝑐 superscript 4457 P_{c}(4457)^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT
signals. To study the production of p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ in the Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) data samples, we also investigate the p J / ψ 𝑝 𝐽 𝜓 pJ/\psi italic_p italic_J / italic_ψ final
state in the Belle continuum data sample. We determine the branching fractions to be ℬ [ Υ ( 1 S ) → p J / ψ + a n y t h i n g ] = ( 4.27 ± 0.16 ± 0.20 ) × 10 − 5 ℬ delimited-[] → Υ 1 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 plus-or-minus 4.27 0.16 0.20 superscript 10 5 {\cal B}[\Upsilon(1S)\to pJ/\psi+anything]=(4.27\pm 0.16\pm 0.20)\times 10^{-5} caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] = ( 4.27 ± 0.16 ± 0.20 ) × 10 start_POSTSUPERSCRIPT - 5 end_POSTSUPERSCRIPT and ℬ [ Υ ( 2 S ) → p J / ψ + a n y t h i n g ] = ( 3.59 ± 0.14 ± 0.16 ) × 10 − 5 ℬ delimited-[] → Υ 2 𝑆 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 plus-or-minus 3.59 0.14 0.16 superscript 10 5 {\cal B}[\Upsilon(2S)\to pJ/\psi+anything]=(3.59\pm 0.14\pm 0.16)\times 10^{-5} caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] = ( 3.59 ± 0.14 ± 0.16 ) × 10 start_POSTSUPERSCRIPT - 5 end_POSTSUPERSCRIPT , and the cross section of continuum production to be σ ( e + e − → p J / ψ + a n y t h i n g ) = ( 57.5 ± 2.1 ± 2.5 ) fb 𝜎 → superscript 𝑒 superscript 𝑒 𝑝 𝐽 𝜓 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 plus-or-minus 57.5 2.1 2.5 femtobarn \sigma(e^{+}e^{-}\to pJ/\psi+anything)=(57.5\pm 2.1\pm 2.5)~{}$\mathrm{fb}$ italic_σ ( italic_e start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT italic_e start_POSTSUPERSCRIPT - end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ) = ( 57.5 ± 2.1 ± 2.5 ) roman_fb at s = 10.52 GeV 𝑠 10.52 GeV \sqrt{s}=10.52~{}\hbox{GeV} square-root start_ARG italic_s end_ARG = 10.52 GeV . No significant P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT signals exist in the Belle Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) data
samples. We determine the upper limits of P c + superscript subscript 𝑃 𝑐 P_{c}^{+} italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT productions in Υ ( 1 , 2 S ) Υ 1 2 𝑆 \Upsilon(1,2S) roman_Υ ( 1 , 2 italic_S ) inclusive decays to be
ℬ [ Υ ( 1 S ) → P c ( 4312 ) + + a n y t h i n g ] ⋅ ℬ [ P c ( 4312 ) + → p J / ψ ] ⋅ ℬ delimited-[] → Υ 1 𝑆 subscript 𝑃 𝑐 superscript 4312 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ delimited-[] → subscript 𝑃 𝑐 superscript 4312 𝑝 𝐽 𝜓 \displaystyle{\cal B}[\Upsilon(1S)\to P_{c}(4312)^{+}+anything]\cdot{\cal B}[P%
_{c}(4312)^{+}\to pJ/\psi] caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B [ italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ]
< \displaystyle< <
4.5 × 10 − 6 , 4.5 superscript 10 6 \displaystyle 4.5\times 10^{-6}, 4.5 × 10 start_POSTSUPERSCRIPT - 6 end_POSTSUPERSCRIPT ,
(5)
ℬ [ Υ ( 1 S ) → P c ( 4440 ) + + a n y t h i n g ] ⋅ ℬ [ P c ( 4440 ) + → p J / ψ ] ⋅ ℬ delimited-[] → Υ 1 𝑆 subscript 𝑃 𝑐 superscript 4440 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ delimited-[] → subscript 𝑃 𝑐 superscript 4440 𝑝 𝐽 𝜓 \displaystyle{\cal B}[\Upsilon(1S)\to P_{c}(4440)^{+}+anything]\cdot{\cal B}[P%
_{c}(4440)^{+}\to pJ/\psi] caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B [ italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ]
< \displaystyle< <
6.8 × 10 − 6 , 6.8 superscript 10 6 \displaystyle 6.8\times 10^{-6}, 6.8 × 10 start_POSTSUPERSCRIPT - 6 end_POSTSUPERSCRIPT ,
(6)
ℬ [ Υ ( 1 S ) → P c ( 4457 ) + + a n y t h i n g ] ⋅ ℬ [ P c ( 4457 ) + → p J / ψ ] ⋅ ℬ delimited-[] → Υ 1 𝑆 subscript 𝑃 𝑐 superscript 4457 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ delimited-[] → subscript 𝑃 𝑐 superscript 4457 𝑝 𝐽 𝜓 \displaystyle{\cal B}[\Upsilon(1S)\to P_{c}(4457)^{+}+anything]\cdot{\cal B}[P%
_{c}(4457)^{+}\to pJ/\psi] caligraphic_B [ roman_Υ ( 1 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B [ italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ]
< \displaystyle< <
4.9 × 10 − 6 , 4.9 superscript 10 6 \displaystyle 4.9\times 10^{-6}, 4.9 × 10 start_POSTSUPERSCRIPT - 6 end_POSTSUPERSCRIPT ,
(7)
ℬ [ Υ ( 2 S ) → P c ( 4312 ) + + a n y t h i n g ] ⋅ ℬ [ P c ( 4312 ) + → p J / ψ ] ⋅ ℬ delimited-[] → Υ 2 𝑆 subscript 𝑃 𝑐 superscript 4312 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ delimited-[] → subscript 𝑃 𝑐 superscript 4312 𝑝 𝐽 𝜓 \displaystyle{\cal B}[\Upsilon(2S)\to P_{c}(4312)^{+}+anything]\cdot{\cal B}[P%
_{c}(4312)^{+}\to pJ/\psi] caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B [ italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4312 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ]
< \displaystyle< <
5.3 × 10 − 6 , 5.3 superscript 10 6 \displaystyle 5.3\times 10^{-6}, 5.3 × 10 start_POSTSUPERSCRIPT - 6 end_POSTSUPERSCRIPT ,
(8)
ℬ [ Υ ( 2 S ) → P c ( 4440 ) + + a n y t h i n g ] ⋅ ℬ [ P c ( 4440 ) + → p J / ψ ] ⋅ ℬ delimited-[] → Υ 2 𝑆 subscript 𝑃 𝑐 superscript 4440 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ delimited-[] → subscript 𝑃 𝑐 superscript 4440 𝑝 𝐽 𝜓 \displaystyle{\cal B}[\Upsilon(2S)\to P_{c}(4440)^{+}+anything]\cdot{\cal B}[P%
_{c}(4440)^{+}\to pJ/\psi] caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B [ italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4440 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ]
< \displaystyle< <
7.2 × 10 − 6 , 7.2 superscript 10 6 \displaystyle 7.2\times 10^{-6}, 7.2 × 10 start_POSTSUPERSCRIPT - 6 end_POSTSUPERSCRIPT ,
(9)
ℬ [ Υ ( 2 S ) → P c ( 4457 ) + + a n y t h i n g ] ⋅ ℬ [ P c ( 4457 ) + → p J / ψ ] ⋅ ℬ delimited-[] → Υ 2 𝑆 subscript 𝑃 𝑐 superscript 4457 𝑎 𝑛 𝑦 𝑡 ℎ 𝑖 𝑛 𝑔 ℬ delimited-[] → subscript 𝑃 𝑐 superscript 4457 𝑝 𝐽 𝜓 \displaystyle{\cal B}[\Upsilon(2S)\to P_{c}(4457)^{+}+anything]\cdot{\cal B}[P%
_{c}(4457)^{+}\to pJ/\psi] caligraphic_B [ roman_Υ ( 2 italic_S ) → italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT + italic_a italic_n italic_y italic_t italic_h italic_i italic_n italic_g ] ⋅ caligraphic_B [ italic_P start_POSTSUBSCRIPT italic_c end_POSTSUBSCRIPT ( 4457 ) start_POSTSUPERSCRIPT + end_POSTSUPERSCRIPT → italic_p italic_J / italic_ψ ]
< \displaystyle< <
2.4 × 10 − 6 , 2.4 superscript 10 6 \displaystyle 2.4\times 10^{-6}, 2.4 × 10 start_POSTSUPERSCRIPT - 6 end_POSTSUPERSCRIPT ,
(10)
at 90% credibility.
Acknowledgements. This work, based on data collected using the Belle detector, which was
operated until June 2010, was supported by
the Ministry of Education, Culture, Sports, Science, and
Technology (MEXT) of Japan, the Japan Society for the
Promotion of Science (JSPS), and the Tau-Lepton Physics
Research Center of Nagoya University;
the Australian Research Council including grants
DP210101900, DP210102831, DE220100462, LE210100098, LE230100085; Austrian Federal Ministry of Education, Science and Research (FWF) and
FWF Austrian Science Fund No. P 31361-N36;
National Key R&D Program of China under Contract No. 2022YFA1601903,
National Natural Science Foundation of China and research grants
No. 11575017,
No. 11761141009,
No. 11705209,
No. 11975076,
No. 12135005,
No. 12150004,
No. 12161141008,
and
No. 12175041,
and Shandong Provincial Natural Science Foundation Project ZR2022JQ02;
the Czech Science Foundation Grant No. 22-18469S;
Horizon 2020 ERC Advanced Grant No. 884719 and ERC Starting Grant No. 947006 “InterLeptons” (European Union);
the Carl Zeiss Foundation, the Deutsche Forschungsgemeinschaft, the
Excellence Cluster Universe, and the VolkswagenStiftung;
the Department of Atomic Energy (Project Identification No. RTI 4002), the Department of Science and Technology of India,
and the UPES (India) SEED finding programs Nos. UPES/R&D-SEED-INFRA/17052023/01 and UPES/R&D-SOE/20062022/06;
the Istituto Nazionale di Fisica Nucleare of Italy;
National Research Foundation (NRF) of Korea Grant
Nos. 2016R1D1A1B02012900, 2018R1A2B3003643,
2018R1A6A1A06024970, RS202200197659,
2019R1I1A3A01058933, 2021R1A6A1A03043957,
2021R1F1A1060423, 2021R1F1A1064008, 2022R1A2C1003993;
Radiation Science Research Institute, Foreign Large-size Research Facility Application Supporting project, the Global Science Experimental Data Hub Center of the Korea Institute of Science and Technology Information and KREONET/GLORIAD;
the Polish Ministry of Science and Higher Education and
the National Science Center;
the Ministry of Science and Higher Education of the Russian Federation
and the HSE University Basic Research Program, Moscow; University of Tabuk research grants
S-1440-0321, S-0256-1438, and S-0280-1439 (Saudi Arabia);
the Slovenian Research Agency Grant Nos. J1-9124 and P1-0135;
Ikerbasque, Basque Foundation for Science, and the State Agency for Research
of the Spanish Ministry of Science and Innovation through Grant No. PID2022-136510NB-C33 (Spain);
the Swiss National Science Foundation;
the Ministry of Education and the National Science and Technology Council of Taiwan;
and the United States Department of Energy and the National Science Foundation.
These acknowledgements are not to be interpreted as an endorsement of any
statement made by any of our institutes, funding agencies, governments, or
their representatives.
We thank the KEKB group for the excellent operation of the
accelerator; the KEK cryogenics group for the efficient
operation of the solenoid; and the KEK computer group and the Pacific Northwest National
Laboratory (PNNL) Environmental Molecular Sciences Laboratory (EMSL)
computing group for strong computing support; and the National
Institute of Informatics, and Science Information NETwork 6 (SINET6) for
valuable network support.