DIFFRACTIVE DIJET PRODUCTION AT CDF. Konstantin Goulianos (for the CDF II Collaboration)

Transcription

1 DIFFRACTIVE DIJET PRODUCTION AT CDF Konstantin Goulianos (for the CDF II Collaboration) 1

2 CONTENTS Introduction / motivation Diffractive dijets Summary 2

3 STUDIES OF DIFFRACTION IN QCD Non-diffractive color-exchange gaps exponentially suppressed Diffractive Colorless vacuum exchange large-gap signature Incident hadrons acquire color and break upart rapidity gap p Incident hadrons retain their quantum numbers remaining colorless CONFINEMENT Goal: probe the QCD nature of the diffractive exchange P O M E R O N p p p p 3

4 DEFINITIONS SINGLE DIFFRACTION x,t M X p p xp p 1- x L ξ CAL ξ Σ all i 1 M 2 X s E i-tower T since no radiation no price paid for increasing diffractive gap size s e η i dσ d Δη t 0 p Rap-gap Dh=-lnx constant dσ dξ 0 ln ln s 1 ξ ln Mx 2 dσ dm dn/dh 2 M X 1 M 2 4

5 DIFFRACTION AT CDF Elastic scattering s T =Im f el (t=0) Total cross section f gap h OPTICAL THEOREM f h SD DD DPE /CD SDD Single Diffraction or Single Dissociation Double Diffraction or Double Dissociation Double Pom. Exchange or Central Dissociation Single + Double Diffraction (SDD) JJ, b, J/y, W p exclusive JJ ee mm...gg p 5

6 540 GeV 1800 GeV FACTORIZATION BREAKING IN SOFT DIFFRACTION diffractive x-section suppressed relative to Regge prediction as s increases p p x,t p M Factor of ~8 (~5) suppression at s = 1800 (540) GeV RENORMALIZATION s=22 GeV C D F see KG, PLB 358, 379 (1995) Question: does factorization breaking affect t-distributions? 6

7 Run I Hard diffraction pp ( X) gap or gap p pbar Fraction: SD/ND 1800 GeV JJ W Fraction % / / dn/dh hard b J/y / / h All fractions ~ 1% (differences due to kinematics) ~ FACTORIZATION! FACTORIZATION! 7

8 Diffractive dijets in Run I H1 CDF ~20 ~8 ~8 All hard-diffraction processes in Run I were found to be suppressed by a factor of ~8 relative to predictions based on HERA-measured PDFs. 8

9 Excusive dijets Calibrate diffractive Higgs production models Phys. Rev. D 77, (2008) 9

10 THE CDF II DETECTOR PLAN VIEW h <2 h < < h < < h <7.4 ~0.03<x< < t <4 GeV 2 10

11 The RPS 11

12 The MiniPlugs overlap bgnd (BG) is reduced by including the MPs in the x CAL evaluation 12

13 Dynamic Alignment of RPS Method: iteratively adjust the RPS X and Y offsets from the nominal beam axis until a maximum in the b-slope is t=0. ±2 mm Limiting factors 1-statistics 2-beam size 3-beam jitter ±2 mm use RPStrk data width~ 2 mm/ N N~1 K events DX,DY = ± 60 m 13

14 x CAL vs x RPS slice 14

15 Why select 0.05<x pbar <0.08? be on the plateau of the ds/dlnx distribution allow enough room to avoid edge-effects accept enough events for good statistics estimated width resulting from the Dx : Dt

16 RPS ACCEPTANCE 0.05 < x <0.08 slowly varying at high t Measure up tp t = 4 GeV2 Having acceptance beyond 4 GeV 2 minimizes edge effects 16

17 t >1 GeV 2 : asymmetric t-distributions as a tool for evaluating bgd at high t schematic view of fiber tracker 25 mm t-distributions 7.5 mm 12.5 mm x 2.5 mm Y p p 2 mm Y = 7.5 mm bgnd = 20 evts/gev 2 tracker s upper edge: t =2.3 GeV 2, estimated from t~q 2 the lower edge is at t = 6.5 GeV 2 (not shown) background level: region of Y track >Y o data for t >2.3 GeV 2 17

18 Diffractive dijet results 18

19 Measurement of F jj SD 19

20 x CAL vs x RPS As RPS tracking was not available for all analyzed data, we used x CAL and calibrated it vs x RPS from data in which RPS tracking was available. A linear relationship is observed between x CAL vs x RPS in the region of x CAL of the measurement 20

21 Average E T jet and h Jet The SD and ND E T Jet distributions are nearly identical The SD h* distribution is shifted towards the c.m.s of the Pomeron-proton collision 21

22 Azimuthal angle difference of jets Left: the SD distributions are more back-to-back Right: the SD multiplicity is peaked at zero, while the ND is peaked at 9. 22

23 x Bj Distributions vs <Q 2 > <Q 2 >=100 GeV 2 The Run I result is confirmed. The drop-off on the rhs is due to the different range of the calorimeters in Run I and Run II. The Bjorken-x distributions vary by only a factor of ~2 over a range of <Q 2 > of 2 orders of magnitude! 23

24 t-distributions and Slopes vs <Q 2 > for t<1 GeV 2 The slopes are nearly constant over a range of 4 orders of magnitude in <Q 2 >! 24

25 t-distributions for -t<4 GeV 2 The rather flat t distributions at large t are copatible with the existence of an underlying diffraction minimum around t 2.5 GeV 2. 25

26 Result of the week 26

27 EXCLUSIVE Dijet Excl. Higgs THEORY CALIBRATION _ p p } JJ PRD 77, (2008) Exclusive dijets 27

28 CDF 4 LHC Larger Energy Larger ET Multigap diffraction Diffractive Higgs production The CDF measurements are having an impact on all LHC physics the MBR (Minimum Bias Rockefeller) simulation is now in PYTHIA8 28

29 Summary We measured SD to ND ratios in dijet production vs Bjorken-x for <Q 2 > up to 10 4 GeV 2 and t > 4 GeV 2 We find: nearly identical E T jet distributions for SD and ND events small <Q 2 > dependence as a function of Bjorken-x no <Q 2 > dependence of the b-slopes at low t t distributions compatible with DL at low t at high t the distributions lie increasingly higher than DL, becoming approximately flat for t >2 GeV 2 compatible with a diffraction minimum at t >2.5 GeV 2 Our findings are compatible with models of diffraction in which the hard scattering is controlled by the PDF of the recoil antiproton, while the rapidity gap formation is governed by the color-neutral soft exchange. Thank you for your attention 29

30 BACKUP 30

31 Data samples 31

32 Ratio of SD to ND events vs E T * TABLE III Diffractive Dijet Production at CDF K. Goulianos 32

33 Slopes of t-didtributions 33

34 RIGGERS AND EVENT SAMPLES 34

35 The end! 35