13
Analysis of the anomalous- dimension matrix of n-jet operators to four loops in SCET Leonardo Vernazza In collaboration with Valentin Ahrens and Matthias Neubert EPS HEP-2011 Grenoble, 21.07.2011 1

Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

Analysis of the anomalous-dimension matrix of n-jet

operators to four loops in SCET

Leonardo Vernazza

In collaboration with

Valentin Ahrens and Matthias Neubert

EPS HEP-2011 Grenoble, 21.07.2011

1

Page 2: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

INTRODUCTION

� Understanding the structure of infrared singularities in gauge theory amplitude has been a long standing issue.

� Recently, it has been shown that they can be mapped onto UV � Recently, it has been shown that they can be mapped onto UV divergences of n-jet operators in SCET.

� This means they can be described by means of an anomalous dimension, whose structure is constrained by:

� soft-collinear factorization,

� color conservation,

� non-abelian exponentiation,

(Becher,Neubert, 2009)

(Becher,Neubert, 2009;

Gardi, Magnea 2009)

� A conjecture has been formulated, which has an extremely simple formand it should hold to all order in perturbation theory.

L. Vernazza, EPS HEP-2011

� non-abelian exponentiation,

� collinear limit.

2

Page 3: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

MOTIVATION

� Important phenomenological applications in higher order log resummation for n-jet processes.

� Interesting for the understanding of the deeper structure of QCD: the anomalous dimension predicts only pairwise interactions among different partons.

� It implies Casimir scaling of the cusp anomalous dimension, in contrastwith results obtained using the AdS/CFT correspondence in the strong-coupling behavior.

This does not tell if and at which order a violation of the Casimir scaling � This does not tell if and at which order a violation of the Casimir scaling could arise in perturbation theory. A diagrammatic analysis excluded it up to 3 loop, and at 4 loops in terms with higher Casimir invariants.

� Our aim is to complete the diagrammatic analysis at four loop.

L. Vernazza, EPS HEP-2011

3

(Becher, Neubert 2009; Gardi, Magnea 2009)

Page 4: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

INFRARED DIVERGENCES OF GAUGE

THEORY AMPLITUDES

� Given a UV renormalized, on-shell n-parton scattering amplitude with IRdivergences regularized in dimensions, one obtains the finite remainder free from IR divergences from

µ µ−⟩ = ⟩

4 2d = − ε

� The multiplicative renormalization factor Z derives from an anomalous dimension :

� The anomalous dimension is conjectured to be very simple:

1

0| ({ }, ) lim ( ,{ }, )| ( ,{ }) .n np p pµ µ−

→⟩ = ⟩Z

ε

M ε M ε

( ,{ }, ) exp ({ }, ) .p pµ

µ µ∞ = ∫Z P Γε

Γ

� Semiclassical origin of IR singularities: completely determined by color charges and momenta of external partons; only color dipole correlations.

L. Vernazza, EPS HEP-2011

2

cusp

( , )

·({ }, ) ( )ln ( ).

2

i j i

s s

i j iij

ps

µµ γ α γ α= +

−∑ ∑

T TΓ

4

(Becher, Neubert 2009; Gardi, Magnea 2009)

Page 5: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

CONSTRAINT ON ΓΓΓΓ: SOFT-COLLINEAR

FACTORIZATION

� The conjecture is driven by the identification of on-shell amplitudes with Wilson coefficients of n-jet operators in SCET:

| ({ }, ) | ({ }, )n np pµ µ⟩ = ⟩M C

(Becher, Neubert 2009)

[on-shell spinors and polarization vectors]×

� Amplitudes of n-jet operators in SCET factorizes into

� A hard function which depends on large momentum transfer

� n-jets depending on the collinear

| nH = ⟩C

2( )ij i j

s p p= ±

L. Vernazza, EPS HEP-2011

� n-jets depending on the collinear momenta of each collinear sector

� A soft function depending on the soft scales

2 with | |i i ijp sp �

S

2 2

i j

ij

ij

p p

sΛ = 5

Page 6: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

CONSTRAINT ON ΓΓΓΓ: SOFT-COLLINEAR

FACTORIZATION

� The identification allows to use properties of the soft-collinear factorization to constrain . First

� Then, invariance under the renormalization group assure that

|| n n⟩ = ⟩M C

Γ= hΓ Γ

\� Then, invariance under the renormalization group assure that

� Soft-collinear factorization gives then

� Given that with , one obtains

+=h c sΓ Γ

2

2

dependence must cancel

( ) ( )) (

i

i

ij ij c i

p

ps+ = +Λ Γ∑c s s

i

Γ Γ 1

���������

cusp) ( ) (( )i i i

c i s i c sL Lα γ αΓ = −Γ +2

2ln

i

ip

=−

·T T({ })∂Γ L ?

L. Vernazza, EPS HEP-2011

cusp

( , )

·({ }, ) ( ) ( ).

2

i j i

s s ij s s

i j i

β µ γ α β γ α= − +∑ ∑T T

Γcusp

({ })( )is

s

iLα

∂= Γ

Γ L ?

6

(Becher, Neubert 2009; Gardi,Magnea 2009)

2 2

2 2ln ln

( )( )

ij

ij i j

i j ij

Ls

Lp p s

βµ µ−

+ −− − −

= =

hard logcollinear log

soft log

Page 7: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

CONSTRAINT ON ΓΓΓΓ: NON-ABELIAN

EXPONENTIATION

� The soft function is a matrix element of Wilson lines:

� The exponent receives contributions only from Feynman diagrams

�1(0) (0)({ }, ) 0 | | 0 exp( ({ }, ))

nn nµ µ= ⟨ … ⟩ =S SS S

�S� The exponent receives contributions only from Feynman diagrams whose color weights are color-connected (“maximally non-abelian”)

� Color structures can be simplified using the Lie commutation relation:

� Use this to decompose color structures into a sum over products of

�S

a bT Tb aT T

c cabif T=−

− =

(Gatheral 1983; Frenkel and Taylor 1984)

L. Vernazza, EPS HEP-2011

� Use this to decompose color structures into a sum over products of connected webs

� Only single connected webs contribute to the exponent .�S 7

= +− −

Page 8: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

CONSTRAINT ON ΓΓΓΓ: CONSISTENCY

WITH THE COLLINEAR LIMIT

� When two partons become collinear, an n-point amplitudes reduces to a (n-1)-parton amplitude times a splitting function:

(Berends, Giele 1989; Mangano, Parke 1991; Kosower1999; Catani,De Florian, Rodrigo 2003)1999; Catani,De Florian, Rodrigo 2003)

1 2 3 1 2 1 3({ , , , , }) ({ , }) ({ , , , })| |n n n n

p p p p p p P p p−… ⟩ = … ⟩ +…SpM M

1

2

1

2

1 2+1|| 2

L. Vernazza, EPS HEP-2011

� must be independent of momenta and colors of partons 3,…n.

Sp 1 2 1 3({ , }, ) ({ , , }, ) ({ , , , }, ) |n np p p p P p pµ µ µ → += … − …P 1 2T T TΓ Γ Γ

(Becher, Neubert 2009)

8SpΓ

Page 9: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

DIAGRAMMATIC ANALYSIS:

ONE, TWO AND THREE LOOPS

� One loop

� one leg:

� two legs:

2

i iC=T

·i jT T

� Two loops

� Three loops

� Three new structures compatible with soft-

collinear factorization:

� and are not compatible with collinear

1

( , , , )

2

( , , )

( )( )( )ln ( , )

4 ( )(

(

)

( ) )

ij kladx bcx a b c si j k l ijkl iklj iljk

i j k l ik jl

adx bcx a b c d

s i i j k

i j k

ds sf

f f Fs s

f f f

αβ β β

α +

− −+

=

− − −

3∆Γ T T T T

T T T T

f f

ijkl ij kl ik jlβ β β β β= + − −

conformalcross ratio

L. Vernazza, EPS HEP-2011

� Two loops

� one leg:

� two legs:

� three legs:

� Incompatible with soft-

collinear factorization.

� and are not compatible with collinear

limit: the splitting function depends on

colors and momenta of additional partons.

� An exception is , if it

vanishes in all collinear limits. It is possible

that such a function exists.

9

2

abc a b c A ii i i

C Cif− =T T T

·2

abc a b c Ai i i jj

Cif− =T T T T T

abc a b c

i j kif− T T T

1f 2f

( , )ijkl iklj iljkF β β β−

(Becher, Neubert 2009; Dixon,Gardi,Magnea,2009)

Page 10: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

DIAGRAMMATIC ANALYSIS:

FOUR LOOPS I

� At four loops structures involving

higher Casimir invariants appears:

� There are possible new structures

compatible with soft-collinear

factorization:

( ) 1 1, tr[( ) ]n na a aaabcd a b c d

ijkl F i j k RRl Rd d…

++= …=T T T T T TD

, 1 2 3

( , ) ( , , )

( ) ( ) ( ),ij iijj s iiij s ij ijkk s

i j i j k

D g D g D gβ α α β α ∝ + + ∑ ∑s4 1∆Γ

L. Vernazza, EPS HEP-2011

� Again, they are not compatible with

the collinear limit, except

if it vanishes in all collinear limits. 10

( , ) ( , , )

, 4 5 1

( , ) ( , , , )

( ) ( ) ( , ).

i j i j k

iijj s iiii s ijkl ijkl iklj iljk

i j i j k l

D g D g D Gα α β β β = + + − ∑ ∑s4 2∆Γ

1( , ),ijkl iklj iljkG β β β−

Page 11: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

DIAGRAMMATIC ANALYSIS:

FOUR LOOPS II

� The two webs have color structure

( ) .adx bcy exy a b c d e

ijklm i j k l mf f f +≡ T T T T TT

� There are two structures compatible with soft-collinear factorization:

� The first function is incompatible with the collinear limit, the second function cannot be excluded, if it vanishes in all collinear limits.

2

( , , ) ( , , , ,

,

)

4 ( ) ( , , , ),iijkk s ij ijklm ijkm ikmj imjk ijml imlj iljm

i j k i j k l m

s g Gα β β β β β β β= −+ −∑ ∑∆Γ T T

L. Vernazza, EPS HEP-2011

� Applied to the two-jet case, it means that the Casimir scaling of the cusp anomalous dimension is still preserved:

11

cusp cusp

cusp

((

()

) )q

s s

F

g

s

AC C

α αγ α

Γ Γ= =

Page 12: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

CONCLUSION

� Infrared singularities in gauge theory amplitude can be mapped onto UV divergences of n-jet operators in SCET.

� They can be described by means of an anomalous � They can be described by means of an anomalous dimension, whose structure is constrained by soft-collinear factorization, non-abelian exponentiation, and two-partoncollinear limit.

� The anomalous dimension is expected to have a very simple structure. It should hold to all order in perturbation theory.

� We have completed a diagrammatic analysis up to four loop, � We have completed a diagrammatic analysis up to four loop, showing that only new structures proportional to functions vanishing in all collinear limits can appear.

� No violation of Casimir scaling of the cusp anomalous dimension arise.

L. Vernazza, HEP-EPS 2011

12

Page 13: Analysis of the anomalous- dimension matrix of n-jet ... · The conjecture is driven by the identification of on-shell amplitudes with ... abc a b c A i i i i C C ... Microsoft PowerPoint

BACKUP

� The formal solution for Z up to four loops in perturbation theory reads

2

0 0 0 0 1 0 0 1' '' 3 4ln s sα α β βΓ Γ Γ −

= + + − + + Γ Γ Γ

Z 0 0 0 0 1 0 0 1

2 3 2

3 2 2

0 0 0 1 1 0 0 0 2 0 1 1 0 2

4 3 2

4 3 2

0 0 0 1 1 0 0

5

' ' '

ln4 4 2 4 16 16 4

11 5 8 12 6 6

4 72 72 36 6

25 1

'

' ' '3 40

4 192

s s

s

s

π π

α β β β β β β

π

α β β β β

π

= + + − + +

Γ Γ + Γ − Γ − − + − + +

Γ Γ + Γ − + − +

Z

Γ Γ Γ Γ

ε ε ε ε ε

ε ε ε ε

ε

3

0 0

4

2

0 2 1 2 0 0 1 0 1 0

3

1

24

192

7 9 15 24 48' ' '

192

β

β β β β β βΓ + Γ + Γ − −−

Γ

Γ Γ

ε

ε

L. Vernazza, EPS HEP-2011

3

53 0 2 1 1 2 0 3

2

192

8 8 8( )

64 8

'sO

β β βα

Γ − − − + + +

Γ Γ Γ Γ

ε

ε ε

13