1. High Q 2 Neutral Current Deep Inelastic Scattering at HERA 3rd year talk Imperial College 25th June 2001 José Ricardo Gonçalo Summary: Introduction HERA and the ZEUS detector Deep Inelastic Scattering Background: previous measurements My analysis and some results Future measurements at HERA

2. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 2 In practice: Measure structure functions of the proton Determine Parton Density Functions Study QCD and electroweak interactions (weak contributions to DIS) Look for contact interactions, large extra dimensions etc Introduction Central questions of particle physics: What does matter consist of? How do the fundamental constituents of matter interact? In HERA: Use electron beam to look inside the proton and “see” quarks and gluons Study the proton’s structure down to distances of the order of ~10 -18 m Study the Standard Model in a region of spacelike gauge boson exchange Look for new physics beyond SM

3. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 3 The HERA collider Electron (positron) – proton collider Bunch crossing time: 96 ns Beam energy: 27.5 GeV (e) 920 GeV (p) Centre of mass energy: 318 GeV 220 electron and proton bunches Rutherford: ~ 5 fm SLAC: ~ 0.1 fm CERN, FNAL: ~ 0.02 fm HERA: ~ 0.001 fm => equivalent to fixed target with E beam  50 TeV

4. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 4 The ZEUS detector Depleted uranium compensating sampling calorimeter:  99.7% solid angle coverage  Timing resolution 4.5 GeV)  Energy resolution (test beam) Central Tracking Detector (CTD)  9 superlayers (8 sense wires each) - 4 stereo superlayers  Angular acceptance 15 o <  <164 o   p T /p T = 0.0058.p T  0.0065  0.0014/p T (long tracks)  Vertex resolution  2 mm (Z) 1 mm (X,Y) with long tracks Magnetic field of 1.43 T achieved with thin superconducting solenoid from  hit mult.scatt. before CTD mult.scatt.

5. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 5 HERA and the ZEUS detector Three level trigger: FLT: first level, (hardware, analog pipeline), decision in 4.5  s, cuts background (veto-wall), physics related slots (large emc deposit in cal for example) SLT: second level (software on transputer network), decision in ~7 ms, based on components TLT: third level, builds event combining different components OFFLINE 40 MHz (96 ns) ~ 500 Hz ~ 100 Hz ~ 14 Hz Luminosity measurement: Bethe-Heitler: e ± p  e ± p  (well known cross section: to~0.5%) Electron and photon detectors upstream in proton direction ~2% total uncertainty Lead-scintillators sampling calorimeters

6. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 6 Deep Inelastic Scattering at HERA Kinematic variables: Q² = (k-k’) 2  four momentum transfer squared Q²  [~0.1,s] x = Q²/2P.q  Bjorken x  [0,1] y = P.q/P.k  inelasticity  [0,1] = P.q/M  E transfer in p rest frame Only 2 independent variables: Q 2 = x.y.s s = (k+P) 2  c.m. energy squared~10 5 GeV 2 e  (k) p (P) X (P’) e  (k’) ,Z 0 (Q 2 =-q 2 ) For y  1, x can go down to x min  0.1/s = 10 -6 max  50 TeV in p rest frame

7. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 7 Deep Inelastic Scattering at HERA Leading order cross section: with: F L negligible at high Q 2 xF 3 parity violating with: Pure Z 0 contribution suppressed, main e.w. contribution is from  Z 0 interference

8. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 8 Low Q 2 :  exchange  similar cross sections for e - p and e + p High Q 2 :  and Z 0 exchange plus interference term; opposite sign of xF3 for e- and e+ Effect of Z 0 visible at high Q 2 At high Q 2 NC and CC cross sections are similar: nice illustration of electroweak unification

9. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 9 Previous measurements of F2 g (p g low x) q (p,x) q (p’ low x Scaling F 2 =F 2 (x) g (p g,x) q (p’ low x q (p’’ low x Gluon density inferred from F2 or measured from D* cross section (test QCD by testing universality of xg(x,Q 2 ) ) Good agreement with fixed target

10. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 10 Previous results on xF3 xF 3 previously measured at low Q2 in fixed target experiments: CCFR (Fermilab, -Fe), CDHSW (CERN, -Fe) Recent measurements at high Q 2 made preliminary by H1 and ZEUS

11. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 11  F 2 sensitive to valence and sea quarks  All flavors contribute to F 2 Measure F 2  Gluon density can be extracted from d F 2 /dlnQ 2  xF 3 sensitive to valence quarks only Sea quarks created in pairs => (quark-antiquark) in sea =0 Measure xF 3 My Analysis: Objectives Measure NC cross sections at high Q 2 :

12. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 12 Main set of data used in previous measurements at ZEUS: 1996-1997 running:  820 GeV protons  27.5 GeV positrons  Integrated luminosity 38.6 pb -1 1999-2000 running:  920 GeV protons  27.5 GeV positrons  Integrated luminosity 66.0 pb -1 New set of data: 1998-1999 running:  920 GeV protons  27.5 GeV electrons  Integrated luminosity 16.6 pb -1 Centre of mass energy:  s = 318 GeV (98-00) / 300 GeV (92-97) The Data

13. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 13 Methods Select large sample of NC events with high purity NC Signature: Isolated electron in final state Backgrounds: Beam-gas overlayed events Photoproduction (quasi-real  ) QED-Compton Cuts: E-pz, timing, etc Electron finder: neural net Reconstruct kinematic variables: Used double angle method, little dependence on energy scale, good resolution Q2xQ2x  Q 2 /Q 2 ~5%  x/x~10% 10%<  x/x<40%

14. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 14 P T : double agle method with corrections to hadronic P T from electron variables Jacquet-Blondel: uses hadronic variables only, used in Charged Current Electron method: traditional reconstruction method

15. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 15 Corrections applied in the analysis: Noise suppression: remove calorimeter bad cells (sparks) and cells with energy deposition less than some threshold (Uranium radioactivity) Remove runs that fail data quality requirements Alignment of calorimeter and CTD For electrons: CAL energy and non-uniformity correction + dead material correction For hadronic final state: CAL energy correction + reconstruction combining tracking and CAL + backsplash correction Monte Carlo: electron energy smearing to match MC and real detector resolution

16. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 16 Low Q 2 :  exchange  similar cross sections for e - p and e + p High Q 2 :  and Z 0 exchange plus interference term Effect of Z 0 visible at high Q 2 My results so far... VERY preliminary!... Efficiency = #events generated and measured in bin #events generated in bin Purity = #events generated and measured in bin #events measured in bin

17. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 17 (Theory curve, not a fit)

18. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 18 Comparison of my reduced cross-section measurement to 96/97 preliminary results

19. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 19 Can we improve the latest F 2 measurement? Measurement from 96/97 data was systematics dominated in most of the accessible region. But there is room for improvement.

20. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 20 My very preliminary Measurement of F 2

21. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 21 My very preliminary Measurement of xF 3

22. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 22 Future Measurements at HERA Neutral current: 4 different x-sections will allow measurement of couplings a u, v u, a d, and v d (HERA sensitive to u & d; LEP sensitive to c & b) Charged Current: cross section linear on polarisation. Limits or measurement of mass of W - R, W + L Expected errors assuming

23. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 23 Summary Done: Have obtained my first results on d  /dQ 2, reduced cross section, F 2 and xF 3. xF 3 measurement can only improve previous measurement very slightly, error dominated by small e - p data sample Expect to improve the previous F 2 measurement from ZEUS for Q 2 > ~650 GeV 2 where statistical errors dominate Assume similar systematic errors Statistical errors should be 30% smaller: under investigation Plan for thesis: Introduction Theory of DIS Data and Monte Carlo sets Description of the analysis Systematic errors Conclusions

24. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 24 Most important sources of systematic errors: Energy scale: below 2% but up to 10% in 1 bin of d  /dxdQ 2 and 1 bin of d  /dQ 2 Electron track momentum: (E track > 5GeV ± 5 GeV) below 2% but up to 14% in some bins of d  /dQ 2 and 6% in bins of d  /dxdQ 2 Monte Carlo fragmentation: LEPTO (MEPS) versus ARIADNE (colour dipole) - order of 2%, can go up to 6% Photoproduction background & normalization: less than 2% but up to 4% in several bins Remaining systematics of order 0.5% to 1% in most bins To do: Measurements of single differential cross sections d  /dx and d  /dy All the systematics: but there’s a lot I can learn from other analysis using the same data

25. Ricardo Goncalo3rd year talk - Imperial College - 25/06/2001 25 Luminosity upgrade and polarisation xF 3 measurement from H1: different binning With both lepton flavors and polarizations  4 different cross sections xF 3 from H1 Loose ends Altarelli-Parisi evolution equations: