The Ξπ
0 baryon at LHCb
Peter Griffith University of Birmingham, UK
1
UNIVERSITY OF BIRMINGHAM SCHOOL OF PHYSICS AND ASTRONOMY SEMINAR 21/10/15
0 baryon at LHCb The UNIVERSITY OF BIRMINGHAM SCHOOL OF PHYSICS - - PowerPoint PPT Presentation
0 baryon at LHCb The UNIVERSITY OF BIRMINGHAM SCHOOL OF PHYSICS AND ASTRONOMY SEMINAR 21/10/15 Peter Griffith University of Birmingham, UK 1 Contents B-physics at the LHC Heavy baryons in B-physics The LHCb detector 0
Peter Griffith University of Birmingham, UK
1
UNIVERSITY OF BIRMINGHAM SCHOOL OF PHYSICS AND ASTRONOMY SEMINAR 21/10/15
0 baryon at the LHC
0 β ππΏβπ+πβ in detail
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 2
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 3
The very successful Standard Model Dark Matter Dark Energy Gravity Hierarchy Problem Matter/antimatter asymmetry
constrain the SM
& precise measurements of SM
Lb baryon at LHCb Peter Griffith 4 University of Birmingham Seminar
reconstructed decay πΆ β π 3872 πΏ LHCb |π
π£π| measurement
π β π πX ~ 80ππ
π pairs per second
πΞπ ππ ~0.4 β plenty of Ξπ 0β²π‘ at the LHC! (20% of b hadrons)
0 has half integer spin β opens the door for unique measurements
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 5
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 6
VELO - high precision tracking and tagging RICH β large background rejection from PID Extensive muon detection system with clean muon triggering Trackers
vertices (B meson flight distance ~10ππ)
Magnet Calorimeters
π(1ππ)
switched to cancel systematic effects
rather than high instantaneous luminosity preferred.
well known luminosity
luminosities
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 7
LHCb Atlas/CMS
~5 ππΌπ¨ read out rate to disk (1 ππΌπ¨ originally planned )
High operational efficiency (~2% deadtime)
Interesting measurements and discoveries
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 8
measurements β new particle with right-handed coupling?
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 9
ππ~ β 0.2 β new particle would have ~20% coupling strength of the W boson π
π£π 2
π
ππ 2 =
πΆ Ξπ
0 β ππβ ππ
πΆ Ξπ
0 β Ξπ +πβ ππ
ππΊπΊ Where ππΊπΊ is the ratio of relevant form factors
Ξπ
0 β ππβ ππ candidates are
reconstructed using mπππ π = πβπ
2 + πβ₯ 2 + πβ₯
Visible mass Transverse momentum of βπ pair
Candidates with 100πππ/π2 uncertainty are selected
New LHCb measurement removes the need for a new particle. But why the initial disagreement?
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 10
0 β πΎ/πππΏβ
content quark cuud
Six dimensional amplitude fit. Using just Ξβ states is not adequate. Two additional states required
πΈπ (ππππ) πΈπ (ππππ) Mass (Meπ/π2) 4449.8 Β± 4.2 4380Β±37 πΎπ 5 2
+
3 2
β
Significance, π 12 9
arXiv:1507.03414
0 β J/πΞ decays
1 2 particle into spin 1 and 1 2
particles
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 11
Decay amplitudes
Transverse production polarisation: 0.06Β±0.07Β±0.02
Appears to be small (O(10%) or less)
Ξπ
0 β ΞπΏ and Ξπ 0 β ΞβπΏ decays if small βΉ
Angular analysis performed on all three angles to Transverse polarisation parameter Angular distributions
downstream long
arXiv:1302.5578
0 β πΎ/πππΏβ
decays
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 12
πΞπ
ππΌπ·π =
1.482 Β± 0.018 Β± 0.012 ps
Acceptance ratio of Ξπ
0 and πΆ0
Decay time distributions for Ξπ
0 and πΆ0
Unprecedented precision dominates world average
arXiv:1509.00292
Lb baryon at LHCb Peter Griffith 13
loops
additional diagrams from new BSM particles in loops
very sensitive to NP
π+ πβ π π‘ π+ πβ π π‘
University of Birmingham 21/10/15
An effective field theory is employed
π+ πβ π π‘ π+ πβ π π‘
Operators (local interaction terms) Wilson coefficients
All Wilson coefficients calculable β predictive NP can be seen in deviations of Wilson coefficients
Lb baryon at LHCb Peter Griffith 14
LHCb-PAPER-2013-025
University of Birmingham 21/10/15
to other π β π‘ππ measurements
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 15
arXiv:1503.07138
BR as a function of π2 Leptonic π΅πΊπΆ Hadronic π΅πΊπΆ
Branching fraction measurement
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 16
0 β Ξβ(1520)π+πβ
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 17
Non-resonant Ξπ
0 β ππΏβπ+πβ
π(ππΏβ) in Ξπ
0 β
πΎ/π β π+πβ ππΏβ
knowledge
SUSY Wilson coefficients from M. J. Aslam, Y.-M. Wang and C.-D. Lu,
Branching fraction predictions (in units of 10e6) for SCA (SM1) and MCN (SM2) models. (a and b without and with LD charmonium contributions respectively) arXiv:1108.6129
All variables blinded in mass region of the π³π
π
0 β ππΏβπ+πβ branching fraction measurement
0 β πΎ/πππΏβ
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 18
MC DATA π(ππΏβ) in Ξπ
0 β πΎ/πππΏβ
No clean separation
MC is produced with only phase- space kinematics Experimentally motivated model for decay structure would need full amplitude analysis
Dimuon mass squared Ξπ
0 β ππΏβπ+πβ MC
βTypicalβ differential decay rate (πΆ0 β πΏβ0π+πβ) as function of dimuon mass squared
0 β ππΏβπ+πβ branching fraction measurement
0 β πΎ/πππΏβ
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 19
MC DATA π(ππΏβ) in Ξπ
0 β πΎ/πππΏβ
No clean separation
MC is produced with only phase- space kinematics Experimentally motivated model for decay structure would need full amplitude analysis
Dimuon mass squared Ξπ
0 β ππΏβπ+πβ MC
βTypicalβ differential decay rate (πΆ0 β πΏβ0π+πβ) as function of dimuon mass squared Could see large signal where MC statistics are relatively low
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 20
Trigger Stripping Selection The LHCb dataflow
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 21
Trigger Stripping Selection
Offline Loose cuts to minimise kinematic biases Large mass windows accept full kinematic phase space
The LHCb dataflow
identification
Muon triggers only at L0
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 22
Trigger Stripping Selection
Offline Loose cuts to minimise kinematic biases Large mass windows accept full kinematic phase space
The LHCb dataflow
identification
Muon triggers only at L0
Imposes kinematic constraints based
final state Proton ππ > 500πππ Ξπ
0 π€π’π¦ π2/πΈππΊ < 5.0
π2 < 17.6 π»ππ2
Proton ππ background vs signal from neural net output
500 MeV
Cuts on the `probNN` variables Kaon:
Proton:
Training Samples:
0 β πΎ/πππΏβ sideband (> 6 π»πV)
0 β ππΏβπ+πβ MC
Most powerful variables:
Optimised with βpunzi figure of meritβ
π‘ πΆ+π
2
π chosen to be 5
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 23
Trigger Stripping Selection
Offline Loose cuts to minimise kinematic biases Large mass windows accept full kinematic phase space
The LHCb dataflow
identification
Muon triggers only at L0
Imposes kinematic constraints based
final state Proton ππ > 500πππ Ξπ
0 π€π’π¦ π2/πΈππΊ < 5.0
π2 < 17.6 π»ππ2
Proton ππ background vs signal from neural net output
500 MeV
Cuts on the `probNN` variables Kaon:
Proton:
Training Samples:
0 β πΎ/πππΏβ sideband (> 6 π»πV)
0 β ππΏβπ+πβ MC
Most powerful variables:
Optimised with βpunzi figure of meritβ
π‘ πΆ+π
2
π chosen to be 5
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 24
Geometric acceptance Stripping/ selection Trigger PID Neural Net
Between 10 and 400 mrad
PID behaviour difficult to replicate in MC. Highly dependent on kinematics Evaluated using βcalibration samplesβ
PID efficiency is not stable
with magnet polarity β needs to be treated separately Strong dependence on kinematics for most selections and acceptances β issue with data/MC mismatch
π = ππππ‘π‘ π
Correcting π³π
π production kinematics
0 mode?
π β π²/πππ³β has similar phase-
space coverage
π³π
π β ππ³βπ + πβ decay structure
the two masses
for all correlations* π πππ‘Ξπ, πππ‘Ξπ, Ξπ, π ππΏβ , π2
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 25
π³π
π * 7D if we do not assume negligible production polarisation
Correcting π³π
π production kinematics
π measurement
Ξπ
0 kinematic correction factor
βobtain clean sample with loose cuts
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 26
Fit πΆ0 ππ and π In data* and MC Small corrections on πΆ0 MC Fit πΆ0 ππ and π at βgenerator levelβ Fit Ξπ
0 ππ and π
at βgenerator levelβ
π₯ = πΞπ π
π
Pπ, π Γ ππππΆ0 ππ, π πππΞπ
0 ππ, π Γ
1 π₯πΆ0 ππ, π
doi:10.1007/JHEP08(2014)143
LHCb LHCb *background subtracted using fit model to π(πΆ0) (s-weighting)
Corrected MC MC Background subtracted data
Perfect agreement not expected. Corrects for production kinematics
0 decay
Finding a βgoldilocksβ model
(or an adventure in failed techniques!)
and speed
systematics (e.g Toy MCβs)
rightβ π³π
π β ππ³βπ + πβ decay structure
the two masses
for all correlations* π πππ‘Ξπ, πππ‘Ξπ, Ξπ, π ππΏβ , π2
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 27
* 7D if we do not assume negligible production polarisation
Kernel density estimation?
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 28
to estimate the pdf.
dimensional KDE on full phase-space
Good: Incredibly accurate!
Slices through each plain of the multi-dimensional phase- space, unfolded into 1 dimension, showing ratio between MC after full reconstruction/selection and generator level MC weighted with the KDE pdf.
Bad: Incredibly CPU intensive!
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 29
Good: Very fast!
Comparison
weighted generator level MC and full selection MC
Bad: Not so accurate⦠Neural Network?
Legendre polynomials
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 30
πππ€πππ’ πππ‘Ξπ, πππ‘Ξπ, Ξπβ², π ππΏβ 2β², π2β² =
π,π,π,π,π
ππππππππ cosππ π
π cosππ ππ Ξπβ² ππ π ππΏβ 2β² π π(π2β²)
ππππππ = π0ππππππ(2π + 1)(2π + 1)(2π + 1)(2π + 1)(2π + 1) ππππππ = 1 πππ€πππ’π‘
πππ€πππ’π‘
ππ cosππ π
π cosππ ππ Ξπβ² ππ π ππΏβ 2β² π π(π2β²)
π0 is chosen to be Β½ and controls
whole model is normalised to the integrated phase-space efficiency
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 31
Fast enough and accurate enough! πππ‘ππ πππ‘ππ Ξπ π2 π ππΏβ 2 5πΈ scan Efficiency correction now largely independent
Systematic uncertainty evaluated with `bootstrapping`
πΆ Ξπ
0βππΏβπ+πβ
πΆ(Ξπ
0βπΎ/πππΏβ)
π = π
π₯π ππ
β
πΆ Ξπ
0βππΏβπ+πβ
πΆ(Ξπ
0βπΎ/πππΏβ) =
π Ξπ
0βππΏβπ+πβ
π(Ξπ
0βπΎ/πππΏβ) =
π π₯π
ππ/ππ ππ
π π₯π
πΎ/π/ππ πΎ/π
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 32
Corrected yield Event efficiency correction Event weight from background subtraction
Systematics propagation:
π π =
π
π₯π ππ
2
β ππ’ππ’ π = π π 2 +
π ππππ‘ ππ‘βπππ(π) + β¦
Includes sources of systematic uncertainty on acceptance/efficiency Other sources, e.g yield extraction model uncertainties
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 33
Ξπ
0 lifetime
Ξπ
0 kinematics
correction Efficiency model + MC statistics Particle Identification Yield extraction fit
Absolute uncertainty on Ξπ
0 β ππΏβπ+πβ
Relative uncertainty near negligible thanks to LHCb and CMS lifetime measurements LHCb unofficial
Corrected MC MC Background subtracted data
PID systematic is combination
bin populations of correction table
widths ( can hide fine structure in PID efficiency space)
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 34
Ξπ
0 β πΎ/πππΏβ
LHCb unofficial
Signal: Double crystal ball πͺπ
π and πͺπ backgrounds: DCB (opposite sign tails)
Combinatorial: exponential πΆπ‘
0 β πΏ+πΏβπ+πβ yield
constrained by fitting mass reflection in upper sideband Swap proton β kaon πΆπ‘
0 pdf parameters constrained from MC.
Yield constrained from data using: π½ =
ππ ππ
π π¦ πΆπ‘ππ¦ β‘
ππ ππ
π π¦ πΆπ‘ππ¦ +
ππ ππ
π π¦ πΆπ‘ππ¦
ππ ππ ππ
Take side band β clean of Ξπ Ξπ
0 β ππΏβπ+πβ to be
unblinded in the coming weeks followed by publication
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 35
0 measurements yet
0 rare decays hint at similar pattern of SM discrepancy!
0 sector requires a lot of groundwork
0,
πΞπ ππ , polarisation, branching fraction measurements of ideal control channels , e.g
Ξπ
0 β πΎ/πππΏβ, Ξπ 0 β πΎ/πππβ
0 β πΎ/πππΏβ!
University of Birmingham 21/10/15 Lb baryon at LHCb Peter Griffith 36
Lb baryon at LHCb Peter Griffith 37 University of Birmingham 21/10/15 Fig1: penguin decay