Anomalies, Conformal Manifolds, and Spheres

Similar documents
Weyl Anomalies and D-brane Charges

Sphere Partition Functions, Topology, the Zamolodchikov Metric

Weyl Anomalies and D-brane Charges. Constantin Bachas. ChrisFest. Supergravity, Strings and Dualities Imperial College London, April

Rigid SUSY in Curved Superspace

Exact Results in D=2 Supersymmetric Gauge Theories And Applications

Recent Advances in SUSY

2d N = (2, 2) supersymmetry with U(1) RV in curved space

Supercurrents. Nathan Seiberg IAS

Quantum Fields in Curved Spacetime

Some Tools for Exploring Supersymmetric RG Flows

Sphere Partition Functions and the Zamolodchikov Metric

Aspects of SUSY Breaking

A Landscape of Field Theories

Supersymmetric Gauge Theories in 3d

Renormalisation Group Flows in Four Dimensions and the a-theorem

Chern-Simons Theory and Its Applications. The 10 th Summer Institute for Theoretical Physics Ki-Myeong Lee

Semper FI? Supercurrents, R symmetries, and the Status of Fayet Iliopoulos Terms in Supergravity. Keith R. Dienes

Holography for N = 1 on S 4

Supersymmetry on Curved Spaces and Holography

D.Blanco, H.C., L.Y.Hung, R. Myers (2013)

M-theory S-Matrix from 3d SCFT

Pietro Fre' SISSA-Trieste. Paolo Soriani University degli Studi di Milano. From Calabi-Yau manifolds to topological field theories

How to resum perturbative series in supersymmetric gauge theories. Masazumi Honda ( 本多正純 )

N = 2 CHERN-SIMONS MATTER THEORIES: RG FLOWS AND IR BEHAVIOR. Silvia Penati. Perugia, 25/6/2010

THE MASTER SPACE OF N=1 GAUGE THEORIES

Topological reduction of supersymmetric gauge theories and S-duality

Symmetries Then and Now

Scale and Conformal Invariance in d = 4

Heterotic Torsional Backgrounds, from Supergravity to CFT

Seiberg-Witten Theories on Ellipsoids Kazuo Hosomichi (YITP)

Techniques for exact calculations in 4D SUSY gauge theories

One Loop Tests of Higher Spin AdS/CFT

The N = 2 Gauss-Bonnet invariant in and out of superspace

Boundaries, Interfaces and Dualities

N = 2 String Amplitudes *

Aspects of (0,2) theories

Cluster algebras from 2d gauge theories

Exact Solutions of 2d Supersymmetric gauge theories

5. a d*, Entanglement entropy and Beyond

Chern-Simons Theories and AdS/CFT

Twistor Strings, Gauge Theory and Gravity. Abou Zeid, Hull and Mason hep-th/

Local RG, Quantum RG, and Holographic RG. Yu Nakayama Special thanks to Sung-Sik Lee and Elias Kiritsis

Introduction to defects in Landau-Ginzburg models

Solution Set 8 Worldsheet perspective on CY compactification

Some applications of light-cone superspace

Generalized N = 1 orientifold compactifications

A Localization Computation in Confining Phase

Knot Homology from Refined Chern-Simons Theory

Entanglement & C-theorems

Current Algebra Constraints on Supersymmetric Quantum Field Theories

2-Group Global Symmetry

Lecture 25 Superconformal Field Theory

Numerical simulation of the N = 2 Landau Ginzburg model

The Landau-Ginzburg/Calabi-Yau Phase Transition

Holographic c-theorems and higher derivative gravity

Lecture 7: N = 2 supersymmetric gauge theory

Eric Perlmutter, DAMTP, Cambridge

15.2. Supersymmetric Gauge Theories

A Supergravity Dual for 4d SCFT s Universal Sector

S 5. On exactly marginal deformations of = 4 SYM and Type IIB Supergravity on AdS 5. Journal of High Energy Physics. Related content.

Recent Progress on Curvature Squared Supergravities in Five and Six Dimensions

Rigid Holography and 6d N=(2,0) Theories on AdS 5 xs 1

arxiv:hep-th/ v1 22 Aug 1995

Higher Spin AdS/CFT at One Loop

19 Entanglement Entropy in Quantum Field Theory

CONFORMAL INVARIANCE ON CALABI-YAU SPACES*

Quantum gravity at one-loop and AdS/CFT

Heterotic Geometry and Fluxes

First Year Seminar. Dario Rosa Milano, Thursday, September 27th, 2012

Yet Another Alternative to Compactification by Heterotic Five-branes

MIFPA PiTP Lectures. Katrin Becker 1. Department of Physics, Texas A&M University, College Station, TX 77843, USA. 1

Central charges and RG flow of strongly-coupled N = 2 theory

Three-Charge Black Holes and ¼ BPS States in Little String Theory

Notes on N=2 σ-models

2 Type IIA String Theory with Background Fluxes in d=2

SUPERSPACE ANALYSIS OF_LQCAJ _ LORENTZ AND GAUGE ANOMALIES IN THE HETEROTIC STRING THEORY * ABSTRACT

Off-shell conformal supergravity in 3D

1 4-dimensional Weyl spinor

Aspects of (susy) RG flows, (s)cfts

A Higher Derivative Extension of the Salam-Sezgin Model from Superconformal Methods

Singlet Couplings and (0,2) Models arxiv:hep-th/ v1 14 Jun 1994

A Brief Introduction to AdS/CFT Correspondence

8.821 String Theory Fall 2008

Katrin Becker, Texas A&M University. Strings 2016, YMSC,Tsinghua University

ON ULTRAVIOLET STRUCTURE OF 6D SUPERSYMMETRIC GAUGE THEORIES. Ft. Lauderdale, December 18, 2015 PLAN

Scale without conformal invariance

Think Globally, Act Locally

Non-Geometric Calabi- Yau Backgrounds

Gravity, Strings and Branes

Looking Beyond Complete Intersection Calabi-Yau Manifolds. Work in progress with Hans Jockers, Joshua M. Lapan, Maurico Romo and David R.

Calabi-Yau Fourfolds with non-trivial Three-Form Cohomology

Twistors, amplitudes and gravity

Symmetries of curved superspace

Flux Compactification of Type IIB Supergravity

S 2 partition functions: Coulomb vs Higgs localization and vortices

Exact results in AdS/CFT from localization. Part I

Sphere partition functions and the 3d superconformal R-charge

Comments on Chern-Simons Contact Terms in Three Dimensions

Lifshitz Geometries in String and M-Theory

Transcription:

Anomalies, Conformal Manifolds, and Spheres Nathan Seiberg Institute for Advanced Study Jaume Gomis, Po-Shen Hsin, Zohar Komargodski, Adam Schwimmer, NS, Stefan Theisen, arxiv:1509.08511

CFT Sphere partition function log Z Power divergent terms are not universal. Can be removed by counterterms like Λ d γ + Λ d 2 γ R + In addition for odd d log Z = F is universal (ambiguity in quantized imaginary part due to Chern-Simons terms depends on framing) for even d log Z = C log (rλ) F. C is universal (in 2d it is c/3 and in 4d it is a) F is not universal; it can be absorbed in a local counterterm, γ E d F with E d the Euler density Used in c-theorem and its generalizations, entanglement entropy, 2

Conformal manifolds S= S 0 + λ i O i (x) is an exactly marginal deformation Family of CFTs labeled by coordinates λ i Metric on the conformal manifold the Zamolodchikov metric O i x O j 0 = g ij(λ) x 2d Focus on d = 2 (later d = 4) 3

Conformal manifolds of 2d (2,2) SCFT Typical example: sigma models with Calabi-Yau target space. In string theory the coordinates on the conformal manifold are the moduli massless fields in 4d 2d chiral fields λ 2d twisted chiral fields λ The conformal manifold is Kahler with K = K c λ, λ + K tc λ, λ 4

2d (2,2) curved superspace For rigid SUSY in curved spacetime use supergravity [Festuccia, NS] Simplification in 2d (locally) pick conformal gauge γ μν = e 2σ δ μν for SUSY (locally) pick superconformal gauge use flat space expressions with explicit σ s Two kinds of (2,2) supergravities. In the superconformal gauge they depend on A chiral Σ = σ + i a + with A μ = ε μν ν a an axial R- gauge field (Lorentz gauge). The curvature is in a chiral superfield R = D 2 Σ. We will focus on this. A twisted chiral Σ = σ + i a + with A μ = ε μν ν a a vector R-gauge field (Lorentz gauge). 5

2d (2,2) curved superspace In this language the curvature is in a chiral superfield R = D 2 Σ. It contains 2 σ. To preserve rigid SUSY in a non-conformal theory we need to add terms to the flat superspace Lagrangian. Various backgrounds in the literature are easily described, e.g. Topological twist is Σ = 0, Σ = 2σ 0 Omega background Σ = 0 Σ = 2σ + 2 i εz z 2 σ + log z θ 2 6

2d (2,2) curved superspace Supersymmetry on S 2 [Benini, Cremonesi; Doroud, Gomis, Le Floch, Lee] is achieved with (suppress r dependence) Σ = log 1 + z 2 + θ 2 i 1+ z 2 Σ = log 1 + z 2 + θ 2 i 1+ z 2 Σ is not necessarily the complex conjugate of Σ. 7

(2,2) sphere partition functions [Benini, Cremonesi; Doroud, Gomis, Le Floch, Lee] placed a nonconformal gauged linear sigma model on a sphere with Σ = log 1 + z 2 + θ 2 i 1+ z 2 Σ = log 1 + z 2 + θ 2 i 1+ z 2 This theory flows in the IR to a nonlinear sigma model with Calabi-Yau target space. They computed Z of the IR theory using localization in the UV theory. 8

(2,2) sphere partition functions Amazing conjecture [Jockers, Kumar, Lapan, Morrison, Romo]: this S 2 partition function is Z = r c/3 e K c(λ,λ) (restoring the radius r, whose power reflects the ordinary conformal anomaly). Similarly, using Σ it is Z = r c/3 e K tc(λ,λ). Proofs based on localization, squashed sphere, tt*, twisting, counterterms and properties of the background [Gomis, Lee; Gerchkovitz, Gomis, Komargodski;...]. 9

Questions/confusions Given that the one point function of a marginal operator vanishes, how can the sphere partition function depend on λ? Why is it meaningful? Can add a local counterterm γ R f λ, λ, making the answer non-universal In an SCFT on the sphere there is no need to add terms to the Lagrangian to preserve SUSY Why does it depend on the background Σ? If it does not, what determines whether we used Σ or Σ to find e K c or e K tc? Where is the freedom in Kahler transformations? What s the conceptual reason for it? Is it UV or IR? 10

Conformal manifolds and anomalies (w/o SUSY) Zamolodchikov metric O i x O j 0 = g ij(λ) In momentum space e ipx O i x O j 0 g ij λ p 2 log(μ 2 /p 2 ) x 4 Dependence on the scale μ leads to a conformal anomaly: with position dependent λ (suppressed coefficients) [Osborn; Friedan, Konechny] T μ μ = c R + g ij λ μ λ i μ λ j + Ordinary conformal anomaly A more subtle anomaly (actually, less subtle) 11

More about anomalies The partition function Z is a nonlocal functional of the background fields (the metric γ μν, exactly marginal couplings λ i, background gauge fields, etc.). Its variation under changing the conformal factor δ σ log Z is an integral of a local functional of the background fields and δσ. δ σ log Z Must be coordinate invariant in spacetime (assume that the regularization preserves it, i.e. T μν is conserved also at coincident points) Must be coordinate invariant in the conformal manifold Must obey the Wess-Zumino consistency conditions In SUSY theories it must be supersymmetric 12

More about anomalies A term in δ σ log Z that is a Weyl variation of a local term is considered trivial. An anomaly is a cohomologically nontrivial term. It cannot be removed by changing a local counterterm. It cannot change by changing the renormalization scheme. Therefore, even though it arises due to a short distance regulator, it is universal it does not depend on the choice of regulator. 13

Returning to conformal manifolds and anomalies (w/o SUSY) O i x O j 0 = g ij(λ) x 4 e ipx O i x O j 0 g ij λ p 2 log(μ 2 /p 2 ) leads to [Osborn; Friedan, Konechny] δ σ log Z γ δσ c R + g ij λ μ λ i μ λ j + Ordinary conformal anomaly A more subtle anomaly 14

Conformal manifolds and anomalies (w/o SUSY) Another possible anomaly in 2d δ σ log Z γ δσε μν B ij λ μ λ i ν λ j can be ruled out as follows. O i1 (p 1 )O i2 (p 2 ) where are finite. = log μ 2 δ p i A i1 i 2 + A i1 i 2 is a second order polynomial in the momenta. Set e.g. p 3 = p 4 = = 0 and find A i1 i 2 = p 1 2 i3 i4 g i1 i 2. Therefore, all these anomalies are generated from the metric and there is no B-field anomaly. 15

Warmup: 2d N = 1 The conformal manifold is parameterized by real superfields λ i. The anomaly should be expressed in superspace need to use curved superspace. Use the superconformal gauge the conformal factor σ is in a real superfield Σ and the curvature is in R = D 2 Σ. The anomaly δ σ log Z γ δσ c R + g ij λ μ λ i μ λ j + is supersymmetrized in the superconformal gauge as δ Σ log Z d 2 θ δσ c D 2 Σ + g ij λ D + λ i D λ j + Ordinary conformal anomaly A more subtle anomaly Ambiguity due to a local counterterm d 2 θ R f λ prevents us from making any statement about Z. 16

(2,2) SCFT The conformal manifold is parameterized by chiral superfields λ i and twisted chiral superfields λ a. The anomaly should be expressed in superspace need to use curved superspace. Focus on the supergravity, where the conformal factor σ is in a chiral superfield Σ and the curvature is in a chiral superfield R = D 2 Σ. 17

(2,2) SCFT The supersymmetrization of the anomaly δ σ log Z γ δσ c R + g ii μ λ i μ λ i + g aa μ λ a μ λ a after gauge fixing is δ Σ log Z d 4 θ δσ + δσ c Σ + Σ + K c λ, λ K tc λ, λ The first term can also be written as c d 2 θ δσ R + c. c. Lack of B-anomaly proves that K is such a sum of two terms. Invariance under Kahler transformations of K tc λ, λ by f λ and of K c λ, λ by f λ up to a variation of a local counter term d 4 θδσ f λ = d 2 θ δr f λ. Alternatively, Kahler invariance, involves a shift Σ Σ + 1 c f λ. 18

(2,2) SCFT With an axial R-symmetry the supersymmetry current multiplet J ±± is a real superfield satisfying (slightly simplified) D J ±± = D ± W with chiral W. In an SCFT W = 0 at separated points, but the anomaly sets a contact term W = D 2 c Σ + K c λ, λ K tc λ, λ = = c R + D 2 K c λ, λ K tc λ, λ It is invariant under Kahler transformations provided we also shift Σ Σ + 1 c f λ. 19

Ambiguities In the superconformal gauge the local terms are expressed in terms of the chiral curvature superfield R = D 2 Σ. Terms that depend on Σ not through R are non-local. Therefore, the anomaly is not a variation of a local term. Freedom in the local term d 2 θ R f λ + c. c. = d 4 θ Σf λ + c. c. with holomorphic f λ leads to freedom in Kahler transformations of K c λ, λ. (It can be absorbed in a shift of Σ.) Other than that, the anomaly is unambiguous. 20

The anomaly in components For a purely conformal variation δσ = δσ the anomaly is δ Σ log Z d 4 θ δσ + δσ c Σ + Σ + K c λ, λ K tc λ, λ = δσ c σ + g ii μ λ i μ i λ + g aa μ λ a μ λ a δσ K c λ, λ The last term leads to and hence on S 2 log Z γ R K c λ, λ + Q.E.D. Z = r c/3 e K c 21

The anomaly in components For a purely conformal variation δσ = δσ the anomaly is δ Σ log Z d 4 θ δσ + δσ c Σ + Σ + K c λ, λ K tc λ, λ = δσ c σ + g ii μ λ i μ i λ + g aa μ λ a μ λ a δσ K c λ, λ The last term leads to log Z γ R K c λ, λ + Without SUSY this term is not universal. It is the variation of the local counterterm. With SUSY it is related to the universal term and hence it is meaningful. 22

Kahler invariance Z = r c/3 e K c Kahler transformations of K c can be absorbed in a local counter term. Alternatively, full Kahler invariance when Σ is also shifted. With this interpretation Z is invariant under the combined transformation K c K c + f λ + f (λ ) r r e 3 c f λ +f λ 23

Extensions Trivial to repeat with Σ and to find K tc. Here we choose the contact terms to preserve the vector R-symmetry use the other 2,2 supergravity. Equivalently, we use a different regulator that preserves the supergravity of Σ. 2d N = (0,2) 4d N = 1 N = 2 24

2d N = (0,2) The conformal manifold is parameterized by chiral superfields λ i. The conformal factor σ is in a chiral superfield Σ and the curvature is in a chiral superfield R = D + Σ. Several possible anomalies including: The ordinary anomaly i dθ + c δσ R + c. c. = i d 2 θ + c δσ Σ + c. c. The anomaly associated with the metric i d 2 θ + δσ + δσ)( K i λ i K i λi 25

2d N = (0,2) The anomaly associated with the metric i d 2 θ + δσ + δσ)( K i λ i K i λi Using the lack of B anomaly i d 2 θ + δσ + δσ)( K i λ i K i λi = i d 2 θ + δσ + δσ i K λ i i K λi with real K. Hence, the metric on the conformal manifold must be Kahler. The last term includes in its component expansion δσk and could lead to an interesting Z. But 26

Ambiguities in 2d N = (0,2) As in all our examples, a local counterterm i d θ + R f λ + c. c. = i d 2 θ + Σ f λ + c. c. with holomorphic f λ Kahler transformations of K. Can redefine the 2d metric by a function of the moduli. Locally a chiral superfield is the same as a real superfield (not in (2,2)). Hence, in the conformal gauge we can shift (more carefully, use SUGRA) Σ Σ + i D + D + H λ, λ. ++ This shifts the ordinary anomaly term i d 2 θ + δσ Σ i d 2 θ + δσ Σ + i d 2 θ + δσ H, which includes in components δσh. Hence, Z is ambiguous (note, there is no local counterterm). 27

4d Without SUSY Anomaly γ δσ(a E 4 c W 2 + g ij λ i λ j + ) Ambiguous counterterms γ R 2 F 1 λ + R 2 μν F 2 λ + R2 μνρσ F 3 λ + With SUSY need to Supersymmetrize in N = 1, 2 σ and λ in chiral superfields. Covariantize in spacetime Covariantize in the conformal manifold 28

4d N = 1 Anomaly γ δσ(a E 4 c W 2 + g i i λ i λ i + ) Ambiguous counterterms γ R 2 F 1 λ, λ + R 2 μν F 2 λ, λ + R2 μνρσ F 3 λ, λ + Can be supersymmetrized and then the local counterterm makes the sphere partition function ambiguous [Gerchkovitz, Gomis, Komargodski]. 29

4d N = 2 Example: theories of class S with moduli λ After a lot of algebra (using relevant formulas in the literature) the expressions without SUSY are supersymmetrized and covariantized to γ δσ a E 4 + g i i λ i λ i + + γ K 2 δσ + Universal Without SUSY this is not universal, but SUSY relates it to this and hence it is universal. 30

4d N = 2 Key fact, In N=1 the counterterm that is proportional to the curvature square is an arbitrary function of λ and λ. In N=2 the counterterm that is proportional to the curvature square must be holomorphic in λ. Therefore, here the ambiguity is only in a holomorphic function of λ. Only Kahler transformations of K. 31

4d N = 2 Collecting all the terms log Z γ E 4 K + and Z = r a ek(λ,λ)/12 Ordinary conformal anomaly The more subtle anomaly This reproduces a result of [Gerchkovitz, Gomis, Komargodski; Gomis, Ishtiaque] 32

Conclusions Anomaly under conformal transformations when the coupling constants depend on position Unrelated to supersymmetry This is a UV phenomenon Visible on flat R d Independent of the background Supersymmetry restricts the form of the anomaly the ambiguity due to local counterterms 33

Conclusions In 2d N = (2,2) the S 2 partition function depends on the anomaly. New derivation of Z = r c/3 e K c The usual conformal anomaly The more subtle conformal anomaly Addresses the questions/confusions we raised 34

Conclusions Addresses the questions/confusions we raised One point function nonzero because of a term in the operator proportional to the curvature (like in the dilaton) Z can be unambiguous when the counterterm proportional to the curvature depends holomorphically on the couplings. Dependence on Σ due to an anomaly. Different choices lead to K c or K tc. The anomaly is set in the UV and is detected by the sphere (IR). 35

Conclusions Three step process the anomaly the ambiguity (freedom in counterterms) the sphere Other cases 2d N = 1 ambiguous 2d N = 0,2 ambiguous 4d N = 1 ambiguous 4d N = 2 Z = r a e K/12 36