Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
First glance at B0 → η′K 0 time-dependent CPV analysis
Stefano Lacaprara
INFN Padova
23rd B2GM, Physics sessionKEK, February 1st 2016
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 1 / 38
Outline
1 Introduction and motivations
2 B0 → η′(→ ηγγπ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ ηγγπ+π−)(K 0S → π0π0)
3 B0 → η′(→ η3ππ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ η3ππ+π−)(K 0S → π0π0)
4 Summary and outlook
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 2 / 38
Outline
1 Introduction and motivations
2 B0 → η′(→ ηγγπ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ ηγγπ+π−)(K 0S → π0π0)
3 B0 → η′(→ η3ππ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ η3ππ+π−)(K 0S → π0π0)
4 Summary and outlook
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 3 / 38
Introduction and motivations
A sensitivity study for Time-Dependent CP violation analysis in theB0 → η′K 0channel, a charmless b → sqq̄ decay
CP asymmetry from time-dependent decay rateinto CP eigenstates;
Sη′K0 = sin 2φeff1 tightly related to sin 2φ1
I identical if only penguin diagram were present;I QCD factorization ∆Sη′K 0 ∈ [−0.03, 0.03]I new physics can enter in the loop, shifting ∆Sη′K 0
more than SM expectationη′ K
0 S
CP
HF
AG
Mo
rio
nd
20
14
0.5 0.6 0.7 0.8
BaBar
PRD 79 (2009) 052003
0.57 ± 0.08 ± 0.02
Belle
JHEP 1410 (2014) 165
0.68 ± 0.07 ± 0.03
Average
HFAG correlated average
0.63 ± 0.06
H F A GH F A GMoriond 2014
PRELIMINARYSimilar to B0 → φK 0, see work by A. GazI more complex final state;I η′ is a pseudo-scalar (φ a vector);I large BR: ∼ 6.6 · 10−5 (∼ 10xBR(B0 → φK 0))[CLEO(1998)]
F constructive interference between penguin diagramsI actual uncertainties σstat = 0.07, σsyst = 0.03
[Belle(2014)]
I projected for 50 ab−1 σstat = 0.008, σsyst = 0.008[Urquijo(2015)]
I no competition from LHCb (neutrals);
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 4 / 38
Status
Channel have been analyzed in B-factory[BABAR(2009), Belle(2007), Belle(2014)];
uncertainties are mostly statistical (∼ 3500 events for all final states);quasi-two body approach;
many decay channels available B0 → η′K 0SI η′ → ρ(→ π+π−)γ; BR:29% not yet: Hulya Atmacan (METU) is interested in working on thisI η′ → ηπ+π−; BR:43%ηγγ : η → γγ ; BR:40%η3π: η → π+π−π0 ; BR:23%
I K 0S → π+π−, K 0S → π0π0,I Total BR(B0 → η′(→ηγγ/η3ππ+π−)K 0S ) = 27% today
B0 → η′K 0L not yet studiedComplex final state, neutrals, large combinatorics;
Release used rel-00-05-03
Signal: MC5 BGx0 (some private production); Background MC5 BGx1
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 5 / 38
Outline
1 Introduction and motivations
2 B0 → η′(→ ηγγπ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ ηγγπ+π−)(K 0S → π0π0)
3 B0 → η′(→ η3ππ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ η3ππ+π−)(K 0S → π0π0)
4 Summary and outlook
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 6 / 38
Selections ηγγ
B0 → η′(→ ηγγπ+π−)(K 0S → π+π−)
skim at least one decay chain have been reconstructedI w/o vertex or mass fit and with loose selections (as for the background)I Using the same skim for all decay channels: run only once on background
pre-selection N ≥ 1 decay chain, with vertex/mass constraintselection cuts on reconstructed quantities
good candidate selection:I Mbc > 5.25 GeV;
I |∆E | < 0.1 GeV;I M(ηγγ) ∈ [0.45, 0.57] GeV;I M(η′) ∈ [0.93, 0.98] GeV;I M(K 0S → π+π−) ∈ [0.48, 0.52] GeV;
I PIDπ(π±) > 0.2; Should use L-ratio instead
I d0(π±) < 0.08mm;
I z0(π±) < 0.1mm;
I N hitsPXD (π±) > 1
I P-valuevtx (B0, η′,K 0S ) > 1 · 10−5
Best candidate
if Ncands > 1, select candidate with highest P-valuevtx (B0, η′, η,K 0S )
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 7 / 38
Distributions: all/good candidates
bcM5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
10000
20000
30000
40000
50000
60000
70000
Mbc
All cands
MC match
Good cands
" MC match
Mbc
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.20
10000
20000
30000
40000
50000
60000
70000
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.70
10000
20000
30000
40000
50000
60000
70000
80000
90000
'ηM0.85 0.9 0.95 1 1.05 1.1 1.150
50
100
150
200
250
3003
10×
S0
KM
0.45 0.46 0.47 0.48 0.49 0.5 0.51 0.52 0.53 0.54 0.551
10
210
310
410
510
πPID0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 11
10
210
310
410
510
)-π+π(S
0) K-π+π) γγ(η'( η→0B
combinatoricssmall
some true cand lost(mostly due to PID(π) and
N hitsPXD (π±) > 1);
almost all goodcandidate(s) aretrue;
Mη has a lowertail;
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 8 / 38
Distributions: best candidate
bcM5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
5000
10000
15000
20000
25000
30000
35000
40000
Mbc
Best cands
" MC match
" SXF
Mbc
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.20
5000
10000
15000
20000
25000
30000
35000
40000
45000
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.70
10000
20000
30000
40000
50000
'ηM0.85 0.9 0.95 1 1.05 1.1 1.150
20
40
60
80
100
120
140
160
180
200
220
310×
S0
KM
0.45 0.46 0.47 0.48 0.49 0.5 0.51 0.52 0.53 0.54 0.551
10
210
310
410
510
πPID0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 11
10
210
310
410
510
)-π+π(S
0) K-π+π) γγ(η'( η→0B
All best candidates are true: SXF is negligible
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 9 / 38
Efficiency and ∆t resolution
Cands multiplicity
N candidates0 2 4 6 8 10 12 14 16 18 20
0
100
200
300
400
500
310×
All cands multiplicity: 1.89
N good candidates0 2 4 6 8 10 12 14 16 18 20
0
100
200
300
400
500
600
310×
Good cands multiplicity: 1.06
: 0.330∈Best cands
: 0.318∈Best cands MC true
: 0.012∈Best cands MC false
)-π+π(S
0) K-π+π) γγ(η'( η→0B
t (ps)∆20− 15− 10− 5− 0 5 10 15 200
5000
10000
15000
20000
25000
30000
35000
40000
45000
/ ndf 2χ 698 / 91
Prob 0
norm 1.04e+02± 4.56e+04
CBias 0.0065± 0.0362 Cσ 0.01± 1.36
T
Bias 0.0071± 0.0432
Tσ 0.0± 3
OBias 0.035± 0.086 Oσ 0.04± 7.48
Cf 0.002± 0.523
Tf 0.003± 0.444
Fit
Core
Tail
Outlier
t>: 2.29 ps∆<
)-π+π(S
0) K-π+π) γγ(η'( η→0B
Efficiency %skim 57.1preselection 48.1good cands 33.1MC true 31.9SXF 1.2
t (ps)∆10− 8− 6− 4− 2− 0 2 4 6 8 100
10000
20000
30000
40000
50000Flavour
0B0B
t (ps)∆10− 8− 6− 4− 2− 0 2 4 6 8 10
Asy
mm
etry
1−
0
1
)-π+π(S
0) K-π+π) γγ(η'( η→0B
Flavour tagging is random,with �(1− 2w)2 = 32% dilution
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 10 / 38
Background reduction
B0 → η′(→ ηγγπ+π−)(K 0S → π+π−)Background MC sample BGx1, single skim for all channels;
Using only a fraction: L=30 fb−1;I need to process the full MC5 dataset to increase # of selected events;I Need to reduce the skim output!
NB No cut (yet) on continuum discriminating variableI still learning how to use itI Alessandro showed some issue in B0 → φK 0presentation on 2015/12/11
Sample # Ev (M) Skim (M) �skim pre-sel sel �seluū 48.15 1.023 2.13 · 10−2 8196 93 1.96 · 10−6dd̄ 13.03 0.274 2.27 · 10−2 2333 37 2.84 · 10−6ss̄ 11.49 0.238 2.07 · 10−2 2334 18 1.57 · 10−6cc̄ 38.87 1.127 2.90 · 10−2 9911 123 1.09 · 10−6total 111.54 2.662 2.39 · 10−2 22774 271 2.43 · 10−6
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 11 / 38
Background distribution at preselection level
5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
20
40
60
80
100
120
140
160
180
200
bcM
uuddsscc
0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.20
50
100
150
200
250
300
350
400
450
E∆0.4 0.45 0.5 0.55 0.6 0.65 0.70
100
200
300
400
500
600
ηM
0.85 0.9 0.95 1 1.05 1.1 1.150
200
400
600
800
1000
'ηM0.45 0.46 0.47 0.48 0.49 0.5 0.51 0.52 0.53 0.54 0.550
500
1000
1500
2000
2500
S
0K
M0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
10
210
310
410
πPID
)-π+π(S
0) K-π+π) γγ(η'( η→0B
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 12 / 38
B0 → η′(→ ηγγπ+π−)(K 0S → π0π0)
Similar selection as in K 0S → π+π−case, taking into account the moredifficult π0 reconstructionI ∆E ∈ [−0.15, 0.25] GeV; (< 0.1 for K0S → π+π−)I M(π0) ∈ [0.1, 0.15] GeV;I M(K 0S → π0π0) ∈ [0.42, 0.52] GeV;I (full list in backup)
bcM5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
10000
20000
30000
40000
50000
60000
70000
80000
90000
Mbc
All cands
MC match
Good cands
" MC match
Mbc
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.2 0.25 0.30
5000
10000
15000
20000
25000
30000
35000
40000
45000
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.70
50
100
150
200
250
310×
'ηM0.85 0.9 0.95 1 1.05 1.1 1.150
100
200
300
400
500
600
700
310×
0πM0.08 0.1 0.12 0.14 0.16 0.18 0.21
10
210
310
410
510
S0
KM
0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.71
10
210
310
410
510
)0π0π(S
0) K-π+π) γγ(η'( η→0B combinatorics larger but stillmanageable (∼ 5 cands/ev)lower tails for Mπ0,K0
S,η;
SXF still small
Efficiency %skim 40.5preselection 36.1good cands 15.6MC true 13.5SXF 2.2
�(K 00S ) ∼ 1/2 �(K+−S )Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 13 / 38
Outline
1 Introduction and motivations
2 B0 → η′(→ ηγγπ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ ηγγπ+π−)(K 0S → π0π0)
3 B0 → η′(→ η3ππ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ η3ππ+π−)(K 0S → π0π0)
4 Summary and outlook
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 14 / 38
B0 → η′(→ η3ππ+π−)(K 0S → π+π−)
Selections η3π
skim, preselection as for the ηγγchannel
good candidate selection:
Mbc > 5.25 GeV;
∆E ∈ [−0.15, 0.15] GeV;
M(η3π) ∈ [0.52, 0.57] GeV;
M(η′) ∈ [0.93, 0.98] GeV;
M(π0) ∈ [0.1, 0.15] GeV;
M(K 0S → π+π−) ∈ [0.48, 0.52] GeV;
PIDπ(π±) > 0.2;
d0(π±) < 0.08mm;
z0(π±) < 0.15mm;
N hitsPXD(π±) > 1
P-valuevtx (B0, η′, η,K 0S ) > 1 ·10−5
Best candidate
if Ncands > 1, select candidate with highest P-valuevtx (B0, η′, η,K 0S )
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 15 / 38
Distributions: all/good candidates
bcM5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.291
10
210
310
410
510
Mbc
All cands
MC match
Good cands
" MC match
Mbc
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.21
10
210
310
410
510
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.71
10
210
310
410
510
'ηM0.85 0.9 0.95 1 1.05 1.1 1.151
10
210
310
410
510
0πM0.08 0.1 0.12 0.14 0.16 0.18 0.21
10
210
310
410
510
S0
KM
0.45 0.46 0.47 0.48 0.49 0.5 0.51 0.52 0.53 0.54 0.551
10
210
310
410
510
610
)-π+π(S
0) K-π+π) 0π-π+π(η'( η→0B Combinatoricsis huge (6π±);
same problemwith π0;
η′ and η3π wellreconstructedfor goodcandidates
non negligibletails for MCtrue π0, η′ andη3π
some 20% ofgoodcandidates arefalse
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 16 / 38
Distributions: best candidate
bcM5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.291
10
210
310
Mbc
Best cands
" MC match
" SXF
Mbc
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.21
10
210
310
410
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.71
10
210
310
410
'ηM0.85 0.9 0.95 1 1.05 1.1 1.151
10
210
310
410
0πM0.08 0.1 0.12 0.14 0.16 0.18 0.21
10
210
310
410
S0
KM
0.45 0.46 0.47 0.48 0.49 0.5 0.51 0.52 0.53 0.54 0.551
10
210
310
410
)-π+π(S
0) K-π+π) 0π-π+π(η'( η→0B
Sizeable SXF ∼ 20%, is smarter best cand selection possible?
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 17 / 38
Efficiency and ∆t resolution
N candidates0 10 20 30 40 50 60 70 80 90 100
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
All cands multiplicity: 21.67
N good candidates0 2 4 6 8 10 12 14 16 18 20
0
10000
20000
30000
40000
50000
Good cands multiplicity: 1.44
)-π+π(S
0) K-π+π) 0π-π+π(η'( η→0B
t (ps)∆20− 15− 10− 5− 0 5 10 15 200
1000
2000
3000
4000
5000
6000
7000
/ ndf 2χ 115 / 91
Prob 0.0469
norm 4.8e+01± 6.9e+03 CBias 0.0166± 0.0934
C
σ 0.04± 1.16
TBias 0.01889±0.00613 −
Tσ 0.1± 2.5 OBias 0.087± 0.391
O
σ 0.13± 4.49 Cf 0.026± 0.508 Tf 0.021± 0.445
Fit
Core
Tail
Outlier
t>: 1.91 ps∆<
)-π+π(S
0) K-π+π) 0π-π+π(η'( η→0B
Better resolution (4π±)
Efficiency %skim 57.3preselection 54.8 (?)good cands 19.1MC true 15.3SXF 3.8
Lower efficiency: onlydue to π0?
pre-sel almost 100% onskim? Combinatorics?to be investigated
Got SWAP problem during
event selection on skims,
only fraction of skim
processed
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 18 / 38
Background reduction
B0 → η′(→ η3ππ+π−)(K 0S → π+π−)Background MC sample BGx1;
L=30 fb−1;
really need to process the full MC5 dataset to increment statistics ofselected events;I NB No cut (yet) on continuum discriminating variable
larger contribute from cc̄
reduction is larger (∼ 10x) than for ηγγchannel (but � as well)
Sample # Ev (M) Skim (M) �skim pre-sel sel �seluū 48.15 1.023 2.13 · 10−2 24037 14 2.91 · 10−7dd̄ 13.03 0.274 2.27 · 10−2 6552 6 1.25 · 10−7ss̄ 11.49 0.238 2.07 · 10−2 9686 5 1.04 · 10−7cc̄ 38.87 1.127 2.90 · 10−2 56653 40 8.31 · 10−7total 111.54 2.662 2.39 · 10−2 96928 271 1.35 · 10−7
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 19 / 38
Background distribution at preselection level
5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
200
400
600
800
1000
1200
1400
1600
bcM
uuddsscc
0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.20
500
1000
1500
2000
2500
3000
3500
4000
4500
E∆0.4 0.45 0.5 0.55 0.6 0.65 0.70
2000
4000
6000
8000
10000
12000
14000
ηM
0.85 0.9 0.95 1 1.05 1.1 1.150
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
'ηM0.08 0.1 0.12 0.14 0.16 0.18 0.2
0
2000
4000
6000
8000
10000
12000
14000
16000
0πM0.45 0.46 0.47 0.48 0.49 0.5 0.51 0.52 0.53 0.54 0.550
10000
20000
30000
40000
50000
60000
70000
S
0K
M
)-π+π(S
0) K-π+π) 0π-π+π(η'( η→0B
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 20 / 38
B0 → η′(→ η3ππ+π−)(K 0S → π0π0)
Channel not used in Belle
Selection ∼ as beforeI ∆E ∈ [−0.15, 0.25] GeV; (< 0.15 for K0S → π+π−)I (full list in backup)
N candidates0 10 20 30 40 50 60 70 80 90 100
0
5
10
15
20
25
30
35
All cands multiplicity: 30.20
N good candidates0 2 4 6 8 10 12 14 16 18 20
0
20
40
60
80
100
120
Good cands multiplicity: 2.45
: 0.000∈Best cands
: 0.000∈Best cands MC true
: 0.000∈Best cands MC false
)0π0π(S
0) K-π+π) 0π-π+π(η'( η→0B
bcM5.255.2555.265.2655.275.2755.285.2855.291
10
Best cands
" MC match
" SXF
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.21
10
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.71
10
210
'ηM0.85 0.9 0.95 1 1.05 1.1 1.151
10
210
0πM0.08 0.1 0.12 0.14 0.16 0.18 0.21
10
210
S0
KM0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7
1
10
210
)0π0π(S
0) K-π+π) 0π-π+π(η'( η→0B
Private production
combinatorics is even larger(∼ 30 candidates per event)Sizeable SXF ∼ 30%,smarter best cand selectionis possible
�(K 00S ) ∼ 1/2 �(K+−S )
Efficiency %preselection 35.5good cands 10.1best cand 5.96SXF 3.8
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 21 / 38
Outline
1 Introduction and motivations
2 B0 → η′(→ ηγγπ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ ηγγπ+π−)(K 0S → π0π0)
3 B0 → η′(→ η3ππ+π−)(K 0S → π+π−)BackgroundB0 → η′(→ η3ππ+π−)(K 0S → π0π0)
4 Summary and outlook
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 22 / 38
Channels summary
With some comparison to B0 → φK 0(A.Gaz) and J/ψ(→ 2µ)K 0S channels
BR Eff. SXF Bgnd.? ∆tB0 →
10−5 sel. % % ·10−6 reso31.9 K+−
S1.2 2.43
η′(→ ηγγπ+π−)K 0S 1.1 13.1 K00S 2.1 1.252.25 ps
12.5 K+−S
3.5 0.58η′(→ η3ππ+π−)K 0S 0.6 6.0 K00S 3.8 −†
1.90 ps
35.2 K+−S
∼ 20φ(→ K+K−)K 0S 0.6 13.7 K00S ∼ 4
2.11 ps
φ(→ 2π)K 0S 0.07 28.3 K+−S ∼ 700 1.42 psJ/ψ(→ 2µ)K 0S 52 - - - 0.92 ps
? NB. w/o continuum suppression cut!†: bug found in background skimming
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 23 / 38
Outlook
Study is still very preliminary
So far only looked at signal η′ → ηπ+π−, with ηγγ and η3π;I ργ final state not yet addressed;I K 0L not yet looked at, too;I selection still to be optimized, but first attempt is good enough to start
with;
first look at continuum background (from MC5 campaign);I No continuum suppression yet
need to look at peaking background;
signal extraction and fit to be implementI synergies with B0 → φK 0analysis;Still long road toward a full scale sensitivity exercise, took first steps
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 24 / 38
Additional stuff
Additional or backup slides
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 25 / 38
Technicalities
Release used rel-00-05-03
code in GIT https://github.com/lacaprara/b2pd analysis
data used:I Privately produced signal;I MC5 for signal (BGx0) (still partially);I MC5 BGx1 sample for continuum background;
F Still only a fraction of produced dataset;F Skim (with loose selection) on large dataset;F Full selection on skimmed events;F trying to produce a single skim for all channels, not finalized yetF got problem (memory and crash in mixed and charged, respectively):
to be investigated;
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 26 / 38
https://github.com/lacaprara/b2pd_analysis
Background
Background selection done in two steps:1 Skim: require a reconstructed decay chain, in any of the four channels,
w/ loose selection and w/o vertex fit and constraint (for speed);I Trying to run skim on continuum background once for all channels;
2 Selection: apply all selection cuts2.1 Pre-selection: re-reconstruct the proper decay chain (exclusive) w/ vertex
and mass constraint2.2 Final: apply all selection cuts
I NB No cut (yet) on continuum discriminating variableI still learning how to use itI Alessandro showed some issue in B0 → φK 0presentation on 2015/12/11
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 27 / 38
Selections
B0 → η′(→ ηγγπ+π−)(K 0S → π0π0)
preselection at least one decay chain have been reconstructed
good candidate selection:I Mbc > 5.25 GeV;I ∆E ∈ [−0.15, 0.25] GeV; (< 0.1 for K0S → π+π−)I M(ηγγ) ∈ [0.45, 0.57] GeV;I M(η′) ∈ [0.93, 0.98] GeV;I M(π0) ∈ [0.1, 0.15] GeV;I M(K 0S → π0π0) ∈ [0.42, 0.52] GeV;I PIDπ(π
±) > 0.2;I d0(π
±) < 0.08mm;I z0(π
±) < 0.15mm;I N hitsPXD(π
±) > 1I P-valuevtx (B0, η
′, η,K 0S ) > 1 · 10−5
if Ncands > 1, select candidate with highest P-valuevtx (B0, η′, η,K 0S )
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 28 / 38
Distributions: all/good candidates
bcM5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
10000
20000
30000
40000
50000
60000
70000
80000
90000
Mbc
All cands
MC match
Good cands
" MC match
Mbc
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.2 0.25 0.30
5000
10000
15000
20000
25000
30000
35000
40000
45000
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.70
50
100
150
200
250
310×
'ηM0.85 0.9 0.95 1 1.05 1.1 1.150
100
200
300
400
500
600
700
310×
0πM0.08 0.1 0.12 0.14 0.16 0.18 0.21
10
210
310
410
510
S0
KM
0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.71
10
210
310
410
510
)0π0π(S
0) K-π+π) γγ(η'( η→0B
combinatoricslarger but stillmanageable
lower tails forMπ0,K0
S,η;
still most of allgoodcandidate(s)are true;
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 29 / 38
Distributions: best candidate
bcM5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
2000
4000
6000
8000
10000
12000
14000
Mbc
Best cands
" MC match
" SXF
Mbc
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.2 0.25 0.30
1000
2000
3000
4000
5000
6000
7000
8000
9000
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.70
5000
10000
15000
20000
25000
30000
'ηM0.85 0.9 0.95 1 1.05 1.1 1.150
20
40
60
80
100
310×
0πM0.08 0.1 0.12 0.14 0.16 0.18 0.21
10
210
310
410
S0
KM
0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.71
10
210
310
410
)0π0π(S
0) K-π+π) γγ(η'( η→0B
Most best candidates are true: SXF is small ∼ 2%
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 30 / 38
Efficiency and ∆t resolution
N candidates0 2 4 6 8 10 12 14 16 18 20
0
20
40
60
80
100
120
310×
All cands multiplicity: 5.11
N good candidates0 2 4 6 8 10 12 14 16 18 20
0
50
100
150
200
250
310×
Good cands multiplicity: 1.22
: 0.156∈Best cands
: 0.135∈Best cands MC true
: 0.022∈Best cands MC false
)0π0π(S
0) K-π+π) γγ(η'( η→0B
t (ps)∆20− 15− 10− 5− 0 5 10 15 200
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
22000
/ ndf 2χ 358 / 91
Prob 33− 2.55e
norm 7.12e+01± 2.15e+04
CBias 0.0096± 0.0285 Cσ 0.01± 1.37
T
Bias 0.0103± 0.0468
Tσ 0.0± 3
OBias 0.051± 0.232 Oσ 0.1± 7.5
Cf 0.003± 0.522
Tf 0.004± 0.445
Fit
Core
Tail
Outlier
t>: 2.30 ps∆<
)0π0π(S
0) K-π+π) γγ(η'( η→0B
Efficiency %preselection 38.2good cands 15.3best cand 13.1
�(K 0S → π0π0) ∼ 0.5 �(K 0S → π+π−)
t (ps)∆10− 8− 6− 4− 2− 0 2 4 6 8 100
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
22000
24000 Flavour0B0B
t (ps)∆10− 8− 6− 4− 2− 0 2 4 6 8 10
Asy
mm
etry
1−
0
1
)0π0π(S
0) K-π+π) γγ(η'( η→0B
Flavour tagging is random,with �(1− 2w)2) = 32% diluition
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 31 / 38
Background distribution at preselection level
5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
bcM
uuddsscc
0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.2 0.25 0.30
1000
2000
3000
4000
5000
6000
E∆0.4 0.45 0.5 0.55 0.6 0.65 0.70
2000
4000
6000
8000
10000
12000
14000
16000
ηM
0.85 0.9 0.95 1 1.05 1.1 1.150
5000
10000
15000
20000
25000
'ηM0.08 0.1 0.12 0.14 0.16 0.18 0.2
0
10000
20000
30000
40000
50000
0πM0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.70
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
22000
S
0K
M
)0π0π(S
0) K-π+π) γγ(η'( η→0B
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 32 / 38
Selections
B0 → η′(→ η3ππ+π−)(K 0S → π0π0)
preselection at least one decay chain have been reconstructed
good candidate selection:I Mbc > 5.25 GeV;I ∆E ∈ [−0.15, 0.25] GeV;I M(η3π) ∈ [0.52, 0.57] GeV;I M(η′) ∈ [0.93, 0.98] GeV;I M(π0) ∈ [0.1, 0.15] GeV;I M(K 0S → π0π0) ∈ [0.4, 0.52] GeV;I PIDπ(π
±) > 0.2;I d0(π
±) < 0.08mm;I z0(π
±) < 0.15mm;I N hitsPXD(π
±) > 1I P-valuevtx (B0, η
′, η) > 1 · 10−5
if Ncands > 1, select candidate with highest P-valuevtx (B0, η′, η,K 0S )
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 33 / 38
Distributions: all/good candidates
bcM5.255.2555.265.2655.275.2755.285.2855.291
10
210
310
All cands
MC match
Good cands
" MC match
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.21
10
210
310
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.71
10
210
310
410
'ηM0.85 0.9 0.95 1 1.05 1.1 1.151
10
210
310
410
0πM0.08 0.1 0.12 0.14 0.16 0.18 0.21
10
210
310
410
S0
KM0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7
1
10
210
310
410
)0π0π(S
0) K-π+π) 0π-π+π(η'( η→0B
NB: channelnot used inBelle
combinatoricsis huge
η and η′
reconstructionis good
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 34 / 38
Distributions: best candidate
bcM5.255.2555.265.2655.275.2755.285.2855.291
10
Best cands
" MC match
" SXF
E∆0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.21
10
ηM0.4 0.45 0.5 0.55 0.6 0.65 0.71
10
210
'ηM0.85 0.9 0.95 1 1.05 1.1 1.151
10
210
0πM0.08 0.1 0.12 0.14 0.16 0.18 0.21
10
210
S0
KM0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7
1
10
210
)0π0π(S
0) K-π+π) 0π-π+π(η'( η→0B
Sizeable SXF ∼ 30%, smarter best cand selection is possibleStefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 35 / 38
Efficiency and ∆t resolution
N candidates0 10 20 30 40 50 60 70 80 90 100
0
5
10
15
20
25
30
35
All cands multiplicity: 30.20
N good candidates0 2 4 6 8 10 12 14 16 18 20
0
20
40
60
80
100
120
Good cands multiplicity: 2.45
: 0.000∈Best cands
: 0.000∈Best cands MC true
: 0.000∈Best cands MC false
)0π0π(S
0) K-π+π) 0π-π+π(η'( η→0B
t (ps)∆20− 15− 10− 5− 0 5 10 15 200
20
40
60
80
100
/ ndf 2χ 44.5 / 91
norm 0.9± 83.6
CBias 3.8± 2
C
σ 0.894± 0.108
T
Bias 0.0335±0.0191 −
Tσ 0.02± 1.25
OBias 0.048± 0.307
O
σ 0.03± 3.05
Cf 01− 5.05e±09 − 3.59e
Tf 0.008± 0.661
Fit
Core
Tail
Outlier
t>: 1.86 ps∆<
)0π0π(S
0) K-π+π) 0π-π+π(η'( η→0B
Private production
Efficiency %preselection 35.5good cands 10.1best cand 5.96SXF 3.8
�(K 0S → π0π0) ∼ 0.5 �(K 0S → π+π−)
t (ps)∆10− 8− 6− 4− 2− 0 2 4 6 8 100
10
20
30
40
50
60
70Flavour
0B
0B
t (ps)∆10− 8− 6− 4− 2− 0 2 4 6 8 10
Asy
mm
etry
1−
0
1
)0π0π(S
0) K-π+π) 0π-π+π(η'( η→0B
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 36 / 38
Background distribution at preselection level
5.25 5.255 5.26 5.265 5.27 5.275 5.28 5.285 5.290
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
bcM
uuddsscc
0.2− 0.15− 0.1− 0.05− 0 0.05 0.1 0.15 0.20
500
1000
1500
2000
2500
3000
3500
4000
4500
E∆0.4 0.45 0.5 0.55 0.6 0.65 0.70
2000
4000
6000
8000
10000
12000
14000
16000
ηM
0.85 0.9 0.95 1 1.05 1.1 1.150
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
22000
'ηM0.08 0.1 0.12 0.14 0.16 0.18 0.2
0
2000
4000
6000
8000
10000
12000
14000
16000
0πM0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.70
5000
10000
15000
20000
25000
S
0K
M
)0π0π(S
0) K-π+π) 0π-π+π(η'( η→0B
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 37 / 38
Bibliography I
[CLEO(1998)] CLEO. Observation of high momentum η′ production in B decays. PRL, 81:1786, 1998. doi:10.1103/PhysRevLett.81.1786. URL http://link.aps.org/doi/10.1103/PhysRevLett.81.1786.
[Urquijo(2015)] Phillip Urquijo. Comparison between belle ii and lhcb physics projections. Technical ReportBELLE2-NOTE-PH-2015-004, Apr 2015.
[BABAR(2009)] BABAR. Measurement of time dependent cp asymmetry parameters in B0 meson decays to ωK0S , η′
K0, and
π0K0S . PRD, 79:052003, 2009. doi: 10.1103/PhysRevD.79.052003. URLhttp://link.aps.org/doi/10.1103/PhysRevD.79.052003.
[Belle(2007)] Belle. Observation of time-dependent cp violation in B0 → η′
K0 decays and improved measurements of cp
asymmetries in B0 → ϕK0, K0S K0S K
0S and B
0 → j/ψK0 decays. PRL, 98:031802, 2007. doi:10.1103/PhysRevLett.98.031802. URL http://link.aps.org/doi/10.1103/PhysRevLett.98.031802.
[Belle(2014)] Belle. Measurement of time-dependent cp violation in b0 → η′k0 decays. Journal of High Energy Physics, 2014(10):165, 2014. doi: 10.1007/JHEP10(2014)165. URL http://dx.doi.org/10.1007/JHEP10%282014%29165.
Stefano Lacaprara (INFN Padova) Hbb KEK 1/2/2016 38 / 38
http://link.aps.org/doi/10.1103/PhysRevLett.81.1786http://link.aps.org/doi/10.1103/PhysRevD.79.052003http://link.aps.org/doi/10.1103/PhysRevLett.98.031802http://dx.doi.org/10.1007/JHEP10%282014%29165
Introduction and motivationsB0'(+-)(K0S+-)BackgroundB0'(+-)(K0S00)
B0'(3+-)(K0S+-)BackgroundB0'(3+-)(K0S00)
Summary and outlook