A New Regulariation of N = 4 Super Yang-Mills Theory

Transcription

1 A New Regulariation of N = 4 Super Yang-Mills Theory Humboldt Universität zu Berlin Institut für Physik F. Alday, J. Henn, J. Plefka and T. Schuster, arxiv:

2 Outline 1 Motivation Why N = 4 Super Yang-Mills Why an Alternative Regularization 2 The New IR Regularization The General Idea Higgsing N = 4 Super Yang-Mills An Explicit Example 3 Summary and Outlook

3 Why N = 4 Super Yang-Mills Reasons to investigate N = 4 Super Yang-Mills SUSY is a candidate for solving: Coupling Constant Unification Hierarchy Problem Dark Matter Problem N = 4 SYM is the maximally supersymmetric gauge theory in 4 dimensions AdS/CFT conjecture: exact duality between N = 4 SYM and a string theory on AdS 5 S 5 posesses quantum superconformal symmetry: β g = 0 controlled by the two parameters N and g Learn about non-supersymmetric gauge theories tree-level QCD is effectively supersymmetric

4 Why N = 4 Super Yang-Mills Example: Gluon Tree Amplitudes from N = 4 SYM Formula for all tree-level super-amplitudes in N = 4 SYM [J.M. Drummond and J.M. Henn JHEP 04(2009)018] Formula for all gluon tree amplitudes! [to be published]

5 Why an Alternative Regularization Motivation For a New Regularization Form of infrared divergences and the BDS ansatz suggest to calculate the logarithm of amplitudes Inconvenience of dimensional regularization: interference between pole terms and O(ε) terms higher order terms needed

6 The General Idea An Alternative Regularization Field Theory: Consider U(N + M) symmetry with N M Higgsing U(N + M) U(N) U(1) M Renders amplitudes IR finite. Have light (m i m j ) and heavy m i fields Dual string picture: Seperating M of the D3-branes serves as regulator Extended dual conformal symmetry of amplitudes

7 Higgsing N = 4 Super Yang-Mills Higgsing N = 4 Super Yang-Mills Field content: All fields in adjoint representation of U(N + M) Gluons: A μ, μ = 0, 1, 2, 3 6 real scalars: Φ I, I = 1,..., 6 4 gluinos: Ψ A, A = 1, 2, 3, 4 The Action: uniquely fixed by supersymmetry Ŝ U(N+M) N =4 = d 4 x Tr ( 1 ˆF 2 4 μν 1 2 (D μ ˆΦ I ) 2 + g2 4 [ˆΦ I, ˆΦ ) J ] 2 + fermions, Decompose all the fields into N + M blocks ( ) ( ) (Aμ ) Â μ = ab (A μ ) aj (ΦI ) ab (Φ I ) aj, ˆΦI = m (A μ ) ia (A μ ) ij (Φ I ) ia δ i I9 g δ ij + (Φ I ) ij a, b = 1,..., N, i, j = N + 1,..., N + M, thereby switching on a vacuum expectation value: ˆΦ I = δ I9 Φ 9 + Φ I

8 Higgsing N = 4 Super Yang-Mills Higgsing N = 4 Super Yang-Mills Add R ξ gauge fixing and appropriate ghost term Qudratic terms (A M := (A μ, Φ I )) { quad Ŝ N =4 = d 4 x 1 2 Tr( μa M ) (m i m j ) 2 (A M ) ij (A M ) ji } m 2 i (A M ) ia (A M ) ai + fermions light fields O ij with mass (m i m j ) heavy fields O ib with mass m i

9 Higgsing N = 4 Super Yang-Mills Higgsing N = 4 Super Yang-Mills Higgsing leads to new bosonic interactions { O(mi Ŝ = N =4 d 4 x m i ([Φ 9, A μ ] A μ ) ii m i (A μ [Φ 9, A μ ] ) ii ) + m i ([Φ 9, Φ ] Φ ) ii m i (Φ [Φ 9, Φ ] ) ii } These vertices are neccesary in order to have extended dual conformal symmetry

10 An Explicit Example One loop test of extended dual conformal symmetry Consider the (special) colour ordered amplitude A 4 = Φ 4 (p 1 ) Φ 5 (p 2 ) Φ 4 (p 3 ) Φ 5 (p 4 ) related to the leading color contribution of four scalar amplitude by A 4 = δ jσ(1) i σ1 σ S 4 /Z 4 Relevant one-loop diagrams: δ j σ(2) i σ(2) δ j σ(3) i σ(3) δ j σ(4) i σ(4) A 4 (σ(1), σ(2), σ(3), σ(4)) + cyclic cyclic permutations + permutations

11 An Explicit Example One loop test of extended dual conformal symmetry The Triangle diagrams want to talk to each other: i 2 j 2 p 2 l 2 p1 l 3 l 1 i 3 j3 p3 p 4 = 2Ng 4 d 4 l (l 1 + p 1 ) (l 3 p 2 ) (2π) 4 (l1 2 + m2 1 )(l2 2 + m2 2 )(l2 3 + m2 3 ) j 1 i 1 i 2 j 2 i 4 j 4 i 3 j3 p 2 l 2 p 1 l 3 l 1 p3 p 4 = 2Ng 4 d 4 l (2m 1 m 2 )(2m 3 m 2 ) (2π) 4 (l1 2 + m2 1 )(l2 2 + m2 2 )(l2 3 + m2 3 ) j 1 i 1 i 4 j 4

12 An Explicit Example One loop test of extended dual conformal symmetry Both diagrams can be combined using the five dimensional momenta ˆp k = (p k, m k m k+1 ) ˆl k = (l k, m k ), i2 j2 i3 j3 p2 l2 p1 j1 i1 l3 l1 p3 p4 j4 i4 + j2 i2 p2 l2 p1 j1 i1 l3 l1 i3 j3 p3 p4 i4 j4 = 2Ng 4 d 4 l (ˆl 2 + 2ˆp 1 ) (ˆl 2 2ˆp 2 ) (2π) 4 ˆl 1ˆl 2 2ˆl Numerator algebra is the same as in the massless case: ˆp 2 k = 0 i2 j2 i3 j3 p2 l2 p1 j1 i1 l3 l1 p3 p4 j4 i4 + j2 i2 p2 l2 p1 j1 i1 l3 l1 i3 j3 p3 p4 i4 j4 + j2 i3 i2 j3 l3 p2 p3 l2 l4 p1 p4 j1 l1 i4 i1 j4 = 16Ng 4 d 4 l (ˆp 1 ˆp 2 )(ˆp 2 ˆp 3 ) (2π) 4 ˆl 1ˆl 2 2ˆl 2 3ˆl All triangle integrals have canceled!

13 An Explicit Example One loop test of extended dual conformal symmetry The four point amplitude is given by ( A 4 = ig 2 YM 1 a ) 2 I(1) (s, t, m i ) + O(a 2 ) here I (1) (s, t, m i ) is the Massive box integral I (1) d 4 l (ˆp 1 ˆp 2 )(ˆp 2 ˆp 3 ) (s, t, m i ) = c 0 (2π) 4 ˆl 1ˆl 2 2ˆl 2 3ˆl Reexpressed in 5d dual variables ˆx ij = ˆx i ˆx j : ˆp i = ˆx i i+1 I (1) (s, t, m i ) = ˆx 2 13ˆx2 24 d 5ˆx 5 δ(ˆx M=4 5 ) ˆx 2 15ˆx2 25ˆx2 35ˆx2 45 Indeed I (1) (s, t, m i ) is extended dual conformal invariant

14 An Explicit Example Extended dual conformal invariance Extended dual conformal invariance 4 ( ˆK μ I (1) (s, t, m i ) := [2x iμ i=1 x ν i x ν i (x 2 i + m 2 i ) x μ i + m i m i ) ] I (1) (s, t, m i ) = 0 Expectation: Amplitudes regulated by Higgsing should be invariant exactly under extended dual conformal symmetry! Only 4 particle invariants: m 1m 3 ˆx 2 13 and m 2m 4 ˆx 2 24 ( I (1) (x13, 2 x24, 2 m1 m 3 m i ) = f ˆx 2 13, m ) 2m 4 ˆx 24 2 Indeed one finds in the m i 0 limit: f (u, v) = 2 ln(u) ln(v) π 2 + O(m 2 )

15 An Explicit Example Higher Loops and More External Legs At 2 loops: Only one integral is allowed by extended dual conformal symmetry: Bubbles and triangles forbidden. Similarly restricts possible integrals at higher loops and more external legs. Checked exponentiation in Higgs regularization: results agree with dimensional regularization

16 Summary and Outlook New IR regularization of planar amplitudes in N = 4 SYM. Breaking of dual conformal invariance at loop level can be repaired by regularization via Higgsing with an exact extended dual conformal symmetry. Regularization via Higgsing restricts allowed integrals at higher loops. Is of pratical help for higher loop calculations. Calculation of regulated integrals is relativly easy. Proof of exact dual conformal symmetry? Proof of anomalous dual conformal Ward identity? Study exponentiation in the new regularization? Can breaking of standard conformal invariance at loop level be controlled?

17 Summary and Outlook New IR regularization of planar amplitudes in N = 4 SYM. Breaking of dual conformal invariance at loop level can be repaired by regularization via Higgsing with an exact extended dual conformal symmetry. Regularization via Higgsing restricts allowed integrals at higher loops. Is of pratical help for higher loop calculations. Calculation of regulated integrals is relativly easy. Proof of exact dual conformal symmetry? Proof of anomalous dual conformal Ward identity? Study exponentiation in the new regularization? Can breaking of standard conformal invariance at loop level be controlled?

18 Thank you for your attention!