1. Regina Demina, University of Rochester 04/11/2013 Asymmetries in top-antitop production

2. Outline 04/11/13Regina Demina, University of Rochester2 Introduction of asymmetry Asymmetry measurement at the Tevatron Interpretation Within Standard Model Beyond Standard Model Axigluons Experimental evidence Charge asymmetry at LHC Ttbar resonances Same sign tops Dijet resonances

3. Asymmetry in top-antitop production Reconstruction level Generator level In early 80s asymmetry observed in e + e -      at sqrt(s)=34.6 GeV<< M Z was used to verify the validity of EW theory (Phys. Rev. Lett. 48, 1701–1704 (1982) Similarly, asymmetry in production could give information about new physics Mediator with axial coupling in s-channel Abnormally enhanced t-channel production Complications: Top is not observed directly, but reconstructed through its decay products Proton and antiproton are not point-like objects, lab frame is different from rest frame   ee ee _p_p p 04/11/133Regina Demina, University of Rochester

4. Definitions Asymmetry defined for ee   In proton-antiproton collisions   y  y is invariant to boosts along z-axis Asymmetry based on  y is the same in lab and tt rest frame Asymmetry based on rapidity of lepton from top decay Lepton angles are measured with a good precision    ee ee q _q_q t _t_t 04/11/134Regina Demina, University of Rochester

5. Measured Forward-Backward Asymmetry 04/11/135Regina Demina, University of Rochester

6. Lepton Asymmetry from top decay 04/11/136Regina Demina, University of Rochester

7. A FB lep vs A FB 04/11/13Regina Demina, University of Rochester7 A FB and A FB lep provide independent information and help to distinguish between models. Left – p T dep Right– p T dep Axial – p T dep D0 SM Axigluons M=500 GeV, G=100 GeV P T (lepton) dependence provides further discrimination D0, 5.4fb -1

8. More Definitions – dilepton system Asymmetry based on the direction of both leptons (two measurements per event) Asymmetry based on  y between leptons  ee q _q_q t _t_t 04/11/138Regina Demina, University of Rochester same as in single lepton case

9. Asymmetries in dilepton system 04/11/13Regina Demina, University of Rochester9 Different quantities

10. Interpretation of the Asymmetry 04/11/13Regina Demina, University of Rochester10 Coulomb repulsion QED: e +   QCD: quark-top   ee ee q _q_q t _t_t Coulomb attraction QED: e -   QCD: antiquark-top

11. Predicted asymmetry in SM Born(  s 2 ) and box(  s 4 ) Coulomb-like repulsion of top and quark and attraction of antitop and quark in QCD Interference –  s 3 Positive asymmetry Final state with no extra partons  small transverse momentum of the tt system ISR (  s 3 ) and FSR(  s 3 ) Interference –  s 3 Negative asymmetry Final state with extra gluons  large transverse momentum of the tt system Possible extra jets + + 04/11/1311Regina Demina, University of Rochester

12. A FB vs P T (ttbar) 04/11/13Regina Demina, University of Rochester12 Gluons from initial radiation are not always reconstructed as jets. P T of the ttbar system is another way to be sensitive to gluon radiation. Need full  s 4 prediction for asymmetry. So far the total cross section is available to that order:Need full  s 4 prediction for asymmetry. So far the total cross section is available to that order:CERN-PH-TH-2013-056, TTK-13-08 Michal CzakonMichal Czakon, Paul Fiedler, Alexander Mitov. Mar 25, 2013.Paul FiedlerAlexander Mitov

13. BSM: S-channel color-octet vectors (axigluons) 04/11/13Regina Demina, University of Rochester13

14. Axigluons – a bit of history 04/11/13Regina Demina, University of Rochester14 Original paper on the subject Phys. Lett. B V190, Issues 1–2, 21 May 1987, 157–161 :Phys. Lett. BV190, Issues 1–2 Chiral color: An alternative to the standard model P. H. Frampton, S. L. Glashow aP. H. FramptonS. L. Glashow SU(3) L ×SU(3) R -8 heavy gluons with chiral structure Alternative Version of Chiral Color as Alternative to the Standard Model P. H. FramptonP. H. Frampton PhysRevD.81.095005 10/2009 SU(3)_L × U(1) – simpler model, cancellation of triangular divergences. Axigluon as Possible Explanation for ttbar Forward-Backward Asymmetry P. H. Frampton, J. Shu, K. Wang Phys. Lett. B 683, 294 (2010) Axigluons cannot explain the observed top quark forward-backward asymmetry R. S. Chivukula, E. H. Simmons, C.-P. Yuan http://arxiv.org/pdf/1007.0260v3.pdfhttp://arxiv.org/pdf/1007.0260v3.pdf Light axigluon explanation of the Tevatron ttbar asymmetry and multijet signals at the LHC C. Gross, G. M. Tavares, M. Schmaltz, C. Spethmann 10/12, http://arxiv.org/pdf/1209.6375v2.pdfG. M. TavaresM. SchmaltzC. Spethmannhttp://arxiv.org/pdf/1209.6375v2.pdf Open windows for a light axigluon explanation of the top forward-backward asymmetry M. Gresham, J. Shelton, K. M. Zurek 9 Feb 2013 http://arxiv.org/abs/arXiv:1212.1718M. GreshamJ. Sheltonhttp://arxiv.org/abs/arXiv:1212.1718 50<M(axigluon)<450 GeV

15. Axigluons: Dijet resonances 04/11/13Regina Demina, University of Rochester15 LHC limits are not applicable for low mass (<400GeV) and large width ( 

16. t-channel: Z’, W’ 04/11/13Regina Demina, University of Rochester16 Direct constraint : from like-sign tops at LHC Introduce SU(2) X that places (u t) R in the same doublet W’ carries “top number” thus suppressing like-sign top production at LHC Predicted asymmetry due to W’ ~30% More forward than SM or s-channel production As a result observed asymmetry reduced to 20% Least constrained by other experimental data, asymmetries in agreement with observed Test this hypothesis by using top polarization

17. How to compare to charge asymmetry at LHC 04/11/13Regina Demina, University of Rochester17 2 problems compared to Tevatron: Large fraction of top pairs (~90%) are produced in gluon fusion, in addition asymmetric qg provides a sizable fraction of ttbar events Anti quark direction is not known  in case of positive asymmetry tops tend to be more forward (because of the boost in the direction of valence quark) than antitops

18. Charge asymmetry at LHC 04/11/13Regina Demina, University of Rochester18 CMS, 5.0 fb -1, l+jets A C = 0.4±1.0 (stat.)±1.1(syst.)% Atlas, l+jets (1.0 fb -1) +dileptons (4.7 fb -1) A C =5.7±2.4(stat)±1.5(syst)%

19. What needs to happen to have a (more) complete story? Complete Tevatron analyses Full asymmetry, lepton based, all as a function of PT(lepton) and/or Mttbar Combine CDF and D0 LHC analyses – try to isolate the qqbar  ttbar process SM theory - Need full  s 4 prediction for the asymmetry. Need full  s 4 prediction for the asymmetry. BSM – reexamine low mass dijet resonances, maybe combine with asymmetry (e.g in bbar system) – hard! BSM for full comparison between Tevatron and LHC – agree on the framework (e.g. axigluons) 04/11/1319Regina Demina, University of Rochester

20. Instead of conclusion: Personal remarks Results are consistent between Tevatron experiment and correspond to ~20% asymmetry at production level Combination of full asymmetry with lepton-based one provides a better discrimination between different models LHC results are reaching the precision compared to the Tevatron data In agreement with the SM, but there is a class of BSM models consistent with the LHC and Tevatron data 04/11/1320Regina Demina, University of Rochester It’s a lovely mystery, Stay tuned

21. Back up 04/11/13Regina Demina, University of Rochester21

22. Systematics - Tevatron 04/11/13Regina Demina, University of Rochester22 CDF: Systematic uncertainties on the parton level AFB measurement, % Background shape 1.4 Background normalization 1.3 Hadronization 1.0 Jet energy scale 0.5 Initial- and nal-state radiation 0.5 Correction procedure 0.3 Color reconnection 0.1 Parton-distribution functions 0.1 Total systematic uncertainty 2.3 D0

23. Systematics - LHC 04/11/13Regina Demina, University of Rochester23 Atlas Systematic uncertainty JES0.003 JER0.002 Lepton ID/sel. efficiency0.006 Generator0.001 Hadronization0.001 Q2 scale0.002 PDF0.002 Pileup<0.001 W + jets0.004 Multijet0.001 Migration matrix0.002 Model dependence0.007 Total0.011 CMS

24. History of measurements and predictions D0, reconstruction level PRL 100, 142002(2008) ICHEP2010 CDF, generator level PRL 101, 202001(2008) Phys. Rev. D 83,112003 (2011) 04/11/1324Regina Demina, University of Rochester

25. Results for asymmetry, in % Reconstruction level (experiments cannot be directly compared, only to Monte Carlo after reconstruction and selection) D0 (5.4 fb -1 ) MC@NLO (D0) CDF (5.3 fb -1 ) MC@NLO (CDF) Generator level (experiments can be directly compared) D0 CDF MC@NLO 04/11/1325Regina Demina, University of Rochester

26. Dijet resonances 04/11/13Regina Demina, University of Rochester26

27. 04/11/13Regina Demina, University of Rochester27

28. 04/11/13Regina Demina, University of Rochester28

29. 04/11/13Regina Demina, University of Rochester29 Indirect D-mixing M G >200GeV EW precision (Zbb,  Z,  had ) M G >500GeV Direct – dijet resonances LHC pp  G  2 jets

30. Reconstruction of top-antitop signal Leptonic top Hadronic top 04/11/1330Regina Demina, University of Rochester

31. D0: Asymmetry at reconstruction level Using kinematic variables of l+jets events construct a discriminant and fit events with  y>0 and  y<0 for top fraction 04/11/1331Regina Demina, University of Rochester Leading b-jet p T    of kinematic fit k T min M jj Discriminant,  y<0 Discriminant,  y>0

32. Towards “true or generated” asymmetry “ Unfolding” =correcting for acceptance (A) and detector resolution (S) Method 1: 4 bin Likelihood unfolding : Problem with Method 1: migration of events near inner bin edge (  y  0) is underestimated, while for the outer edge it is overestimated Solution: fine bins closer to  y=0 Problem: statistical fluctuations in data make the fine bin unfolding unstable Solution: employ regularization Bonus: reduced statistical uncertainties Method 2: fine bin unfolding with regularization 04/11/1332Regina Demina, University of Rochester

33. Method 2: fine bin unfolding with regularization 04/11/13Regina Demina, University of Rochester33 Migration matrix

34. D0: Lepton-based asymmetry, in % Since lepton direction is defined with a very good precision, lepton based asymmetry is simpler to extract Lepton from top decay carries information about underlying asymmetry at production Can be directly compared to theoretical predictions Reconstruction level Generated level 04/11/1334Regina Demina, University of Rochester