hadronic light by light scatering from latice qcd
play

Hadronic light-by-light scatering from latice QCD Jeremy Green in - PowerPoint PPT Presentation

Hadronic light-by-light scatering from latice QCD Jeremy Green in collaboration with Nils Asmussen, Oleksii Gryniuk, Georg von Hippel, Harvey Meyer, Andreas Nyffeler, Vladimir Pascalutsa, Hartmut Witig Institut fr Kernphysik, Johannes


  1. Hadronic light-by-light scatering from latice QCD Jeremy Green in collaboration with Nils Asmussen, Oleksii Gryniuk, Georg von Hippel, Harvey Meyer, Andreas Nyffeler, Vladimir Pascalutsa, Hartmut Witig Institut für Kernphysik, Johannes Gutenberg-Universität Mainz Humboldt Universität – NIC, DESY Zeuthen Join Latice Seminar June 27, 2016

  2. Outline 1. Introduction 2. Latice four-point function 3. Light-by-light scatering amplitude 4. Strategy for g − 2 5. Summary and outlook Some of these results were published in Phys. Rev. Let. 115 , 222003 (2015) [1507.01577] PoS(Latice 2015)109 [1510.08384] Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 2 / 30

  3. Precision low-energy physics ◮ Direct way to search for new physics: try to create new particles at high energy in a collider. E.g., Higgs boson at LHC. ◮ Indirect way: measure a low-energy observable very precisely, and look for small deviations from a similarly precise Standard Model prediction. This can impose strong constraints on new physics that are complementary to direct searches. Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 3 / 30

  4. Magnetic moment of the electron In an external magnetic field, a particle will have potential energy µ · � U = − � B , where � µ is its magnetic moment. An electron’s magnetic moment is given by − e � � µ = g e S . 2 m e A classical rotating charged body has gyromagnetic ratio g = 1, whereas the Dirac equation predicts g = 2. QED and the Standard Model produce deviations from g = 2: the anomalous magnetic moment , a e ≡ g e − 2 + . . . = 2 = α 2 π + . . . Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 4 / 30

  5. Anomalous magnetic moment of the electron T. Aoyama, M. Hayakawa, T. Kinoshita, M. Nio, Phys. Rev. Let. 109 , 111807 (2012) � n � α � a ( 2 n ) + a ( had ) + a ( EW ) a SM = e e e e π n � 2 � 3 � 4 = α � α � α � α 2 π − ( 0 . 328 . . . ) + ( 1 . 181 . . . ) − ( 1 . 910 . . . ) π π π � 5 � α + 1 . 68 ( 2 ) × 10 − 12 + 0 . 0297 ( 5 ) × 10 − 12 + ( 9 . 16 . . . ) π Using α from atomic physics experiments yields = 1 159 652 181 . 78 ( 6 )( 4 )( 2 )( 77 ) × 10 − 12 [0.67 ppb] , a SM e where the dominant uncertainty comes from α . This agrees well with the experimental value, = 1 159 652 180 . 73 ( 28 ) × 10 − 12 [0.24 ppb] . a expt e D. Hanneke, S. Fogwell, G. Gabrielse, Phys. Rev. Let. 100 , 120801 (2008) This can be reversed: use the experiment as a determination of α , which results in the current most precise value of the fine structure constant. Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 5 / 30

  6. Sensitivity to heavy particles ◮ The Standard Model prediction for a e is dominated by QED diagrams containing only photons and electrons. ◮ Loops with heavier particles ( µ , τ , hadrons) contribute only ∼ 5 × 10 − 12 . ◮ Generically the contribution from a heavy particle with mass M is suppressed by ( m e M ) 2 . ◮ In the muon g − 2, contributions from heavy particles will therefore be enhanced by ( m µ m e ) 2 ≈ 40 000, relative to their contributions to a e . Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 6 / 30

  7. Muon g − 2: experiment BNL E821: G. W. Bennet et al. (Muon g − 2 Collaboration), Phys. Rev. D 73 , 072003 (2006) ◮ Polarized muons injected into a 14 m diameter storage ring. ◮ The muon’s spin precesses relative to its � � velocity: � � ω a = e 1 a µ � v × � � � B − a µ − E . γ 2 − 1 m µ ◮ “Magic” muon energy E ≈ 3 . 098 GeV used to eliminate the dependence on � E . a µ = 116 592 089 ( 63 ) × 10 − 11 ◮ The electron produced in muon decay [0.54 ppm] has direction correlated with the muon spin. BNL storage ring moved to Fermilab for new muon g − 2 experiment: E989. Goal: reduce uncertainty by a factor of four. New experiment using ultra-cold muons also planned by J-PARC E34. Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 7 / 30

  8. Muon g − 2: theory T. Blum et al. (US Physics “Snowmass” Self Study), 1311.2198 F. Jegerlehner, A. Nyffeler, Phys. Rept. 477 (2009) 1–110 QED O ( α ) 116 140 973 . 32 ( 8 ) = 0 . 5 ( α / π ) QED O ( α 2 ) 413 217 . 63 ( 1 ) = 0 . 765 857 425 ( 17 )( α / π ) 2 QED O ( α 3 ) 30 141 . 90 ( 0 ) = 24 . 050 509 96 ( 32 )( α / π ) 3 QED O ( α 4 ) 381 . 01 ( 2 ) = 130 . 8796 ( 63 )( α / π ) 4 QED O ( α 5 ) 5 . 09 ( 1 ) = 753 . 29 ( 1 . 04 )( α / π ) 5 QED combined 116 584 718 . 95 ( 8 ) T. Aoyama et al. , Phys. Rev. Let. 109 , 111808 (2012) Electroweak 154 ( 1 ) J. P. Miller et al. , Ann. Rev. Nucl. Part. Sci. 62 (2012) 237–264 HVP (LO) 6949 ( 43 ) K. Hagiwara et al. , J. Phys. G 38 (2011) 085003 HVP (HO) − 98 . 4 ( 7 ) HLbL 105 ( 26 ) “Glasgow consensus” or 116(39) × 10 − 11 Standard Model 116 591 828 ( 50 ) Jegerlehner + Nyffeler × 10 − 11 ∆ a µ ( expt − SM ) 261 ( 78 ) Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 8 / 30

  9. Muon g − 2: theory uncertainty The two dominant sources of uncertainty are hadronic effects: Hadronic vacuum polarization: a HVP,LO = 6949 ( 43 ) × 10 − 11 . µ ◮ Determined using experimental data on cross section for e + e − → hadrons. ◮ Very active field for latice QCD calculations working toward an ab initio prediction with competitive uncertainty. Hadronic light-by-light scatering: a HLbL = 105 ( 26 ) or 116 ( 39 ) × 10 − 11 . µ ◮ Determined using models that include meson exchange terms, charged meson loops, etc. ◮ Could benefit significantly with reliable input from the latice. Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 9 / 30

  10. π 0 contribution to HLbL scatering comes from π 0 exchange diagrams, About 2 / 3 of theory prediction for a HLbL µ which dominate at long distances. Large contributions also come from η , η ′ . + + Their contribution to the four-point function: Π E , π 0 µ 1 µ 2 µ 3 µ 4 ( p 4 ; p 1 , p 2 ) = − p 1 α p 2 β p 3 σ p 4 τ � F 12 ϵ µ 1 µ 2 α β F 34 ϵ µ 3 µ 4 στ + F 13 ϵ µ 1 µ 3 ασ F 24 ϵ µ 2 µ 4 βτ � ( p 1 + p 2 ) 2 + m 2 ( p 1 + p 3 ) 2 + m 2 π π � + F 14 ϵ µ 1 µ 4 ατ F 23 ϵ µ 2 µ 3 βσ , � ( p 2 + p 3 ) 2 + m 2 π j ) is the π 0 γ ∗ γ ∗ form factor. where p 3 = − ( p 1 + p 2 + p 4 ) and F ij = F ( p 2 i , p 2 Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 10 / 30

  11. Light-by-light scatering Before computing a HLbL , start by studying light-by-light scatering by itself. µ This has much more information than just a HLbL . We can: µ ◮ Compare against phenomenology. ◮ Test models used to compute a HLbL . µ Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 11 / 30

  12. Latice four-point function Directly compute four-point function of vector currents ◮ Use one local current Z V J l µ at the source point. ◮ Use three conserved currents J c µ . In position space: Π pos � Z V J l µ 3 ( 0 ) [ J c µ 1 ( x 1 ) J c µ 2 ( x 2 ) J c µ 1 µ 2 µ 3 µ 4 ( x 1 , x 2 , 0 , x 4 ) = µ 4 ( x 4 ) + δ µ 1 µ 2 δ x 1 x 2 T µ 1 ( x 1 ) J c µ 4 ( x 4 ) + δ µ 1 µ 4 δ x 1 x 4 T µ 4 ( x 4 ) J c µ 2 ( x 2 ) + δ µ 2 µ 4 δ x 2 x 4 T µ 4 ( x 4 ) J c µ 1 ( x 1 ) + δ µ 1 µ 4 δ µ 2 µ 4 δ x 1 x 4 δ x 2 x 4 J c � µ 4 ( x 4 ) ] , where T µ ( x ) is a “tadpole” contact operator. This satisfies the conserved-current relations, µ 1 Π pos µ 2 Π pos µ 4 Π pos ∆ x 1 µ 1 µ 2 µ 3 µ 4 = ∆ x 2 µ 1 µ 2 µ 3 µ 4 = ∆ x 4 µ 1 µ 2 µ 3 µ 4 = 0 . Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 12 / 30

  13. Qark contractions Compute only the fully-connected contractions, with fixed kernels summed over x 1 and x 2 : Π pos ′ � f 1 ( x 1 ) f 2 ( x 2 )Π pos µ 1 µ 2 µ 3 µ 4 ( x 4 ; f 1 , f 2 ) = µ 1 µ 2 µ 3 µ 4 ( x 1 , x 2 , 0 , x 4 ) x 1 , x 2 X X 2 X X 2 X X 2 1 1 1 Generically, need the following propagators: ◮ 1 point-source propagator from x 3 = 0 0 X 4 0 X 4 0 X 4 ◮ 8 sequential propagators through x 1 , for each µ 1 and f 1 or f ∗ 1 ◮ 8 sequential propagators through x 2 ◮ 32 double-sequential propagators through x 1 and x 2 , for each ( µ 1 , µ 2 ) and ( f 1 , f 2 ) or ( f ∗ 1 , f ∗ 2 ) Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 13 / 30

  14. Kinematical setup Obtain momentum-space Euclidean four-point function using plane waves: µ 1 µ 2 µ 3 µ 4 ( x 4 ; f 1 , f 2 ) � � e − ip 4 · x 4 Π pos ′ � � Π E µ 1 µ 2 µ 3 µ 4 ( p 4 ; p 1 , p 2 ) = � f a ( x ) = e − ipa · x . x 4 Thus, we can efficiently fix p 1 , 2 and choose arbitrary p 4 . ◮ Full 4-point tensor is very complicated: it can be decomposed into 41 scalar functions of 6 kinematic invariants. ◮ Forward case is simpler: Q 1 ≡ p 2 = − p 1 , Q 2 ≡ p 4 . Then there are 8 scalar functions that depend on 3 kinematic invariants. Jeremy Green (Mainz) Hadronic light-by-light scatering from latice QCD DESY Latice Seminar 14 / 30

Recommend


More recommend