1 / 30

Towards a c-theorem in Four Dimensions

Towards a c-theorem in Four Dimensions. Brian Wecht Center for Theoretical Physics, MIT. Based on work with Ken Intriligator, Eddy Barnes, and Jason Wright. Outline. I. RG flows and the c-theorem. II. a-maximization and SCFTs. III. a-maximization Along the Entire RG Flow.

Download Presentation

Towards a c-theorem in Four Dimensions

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Towards a c-theorem inFour Dimensions Brian Wecht Center for Theoretical Physics, MIT Based on work with Ken Intriligator, Eddy Barnes, and Jason Wright

  2. Outline I. RG flows and the c-theorem II. a-maximization and SCFTs III. a-maximization Along the Entire RG Flow IV. Some new SCFTs V. Conclusions UV IR

  3. I. RG Flows and the c-theorem UV In 2d, we can define a quantity c which monotonically decreases along RG flows. (Zamolodchikov’s c-theorem) IR with The endpoints of this flow are CFTs, with and at these points, Recall that the central charge appears in the trace anomaly:

  4. One can actually make an even stronger claim here: Zamolodchikov argued that 2d RG flows are gradient flows, This implies positive definite In 2d, the c-theorem confirms our intuition that d.o.f. should decrease along RG flow! What about in four dimensions?

  5. In 4d, we have more options! One way to see this: Cardy sez: It’s a! In ALL KNOWN EXAMPLES, adecreases under RG flow (but we know flows where c increases). The conjectureda-theorem then states that but there is currently no general (and generally accepted) proof! IR UV

  6. But this a weaker statement than the 2d c-theorem! Stronger Claim: One can define a quantity a along the entire flow; this quantity monotonically decreases and agrees with the conformal anomaly at the fixed points. Even Stronger Claim: As in 2d, this is a gradient flow with positive definite metric. Osborn et al investigated this latter claim and found it to be true perturbatively! Can we say anything exact?

  7. We can use SUSY to get exact results! Let’s do a quick review of 4d SCFTs: • The superconformal algebra is , saturated for chiral primaries. • The beta function for • theories is given by so this means that implies that the R-current is anomaly free!

  8. We’re not done yet… We can write other beta functions in terms of R-charges: • superpotential coupling • There is a supermultiplet with the R-current, stress tensor, • and SUSY current in it. There’s a unitarity bound, • (Freedman et al) • can compute via ‘t Hooft anomaly matching! The point: The R-symmetry is a very useful tool – provided that we can find it!

  9. II. a-maximization and SCFTs In some cases, it is easy to find the anomaly-free R symmetry: For SU(Nc) SQCD with Nf flavors, But in general the R-charges are not uniquely determined by anomaly freedom! For SQCD with an extra adjoint matter field X, means there’s a one-parameter family of possible R-charges! ? Which of these is the

  10. the appropriate values of The answer: (K. Intriligator, BW) real parameters Define all other U(1) symmetries any anomaly-free R-symmetry Finding the means finding To do this, just find the local max of the function

  11. So we now know the R-charges in ANY 4d SCFT! This means we also know the exact dimensions of lots of operators, all for the low low price of maximizing a cubic function. some restrictions apply Also, at the maximum, a is the central charge of the SCFT: (there’s a similar expression for c, too) This means that SUSY gauge theories are excellent testing grounds for the a-theorem.

  12. Quick quick example: Consider a free chiral superfield This has extrema at r=2/3 and r=4/3. min! max! And it’s basically just as easy for interacting theories! Note one other thing: Since we’re maximizing a cubic function, R-charges, chiral primary operator dimensions, and central charges must be quadratical irrationals. These cannot depend on any continuous moduli.

  13. a-maximization almost proves the a-theorem! Since relevant deformations generally break the flavor symmetries, Maximizing over a subset then implies that But this is wrong! Loopholes: 1) Accidental symmetries. 2) Only a local max. A recent proposal of Kutasov helps with the second, but let’s talk about the first right now…

  14. Accidental Symmetries In general, you never know when you’ll see an accidental symmetry in the IR. But there’s a special kind we know about! Sometimes operators like appear to violate the unitarity bound For example, this happens in SQCD: Since when we find But for this number of flavors, the theory is in a free magnetic phase!

  15. S When a gauge invariant operator appears to violate the unitarity bound, an accidental symmetry rescues it by dragging its R-charge back up to 2/3! Seiberg says: That is, the operator becomes free. Kutasov, Parnachev, and Sahakyan pointed out that this nontrivially affects a-maximization. If an operator M hits the unitarity bound, we should maximize

  16. III. a-maximization along the entire RG flow The weak form of the a-theorem, has survived many different tests! Can we extend it? Use Lagrange multipliers! Kutasov says: This Lagrange multiplier enforces anomaly freedom in the IR.

  17. Now, just extremize with respect to each R-charge! We can now interpret these as the R-charges along the flow UV IR Plugging back into a gives which interpolates between the two central charges.

  18. This proposal effectively takes care of loophole #2. 0 and until the IR fixed point, where the beta function vanishes. So a is monotonically decreasing!

  19. An Aside: These Lagrange multipliers have an interesting interpretation. (Kutasov) Compare anomalous dimensions: Expand , we find So, since But we know from gauge theory that We thus conclude that The Lagrange mutliplier acts like g2in some scheme!

  20. We can go further, too: (Barnes, Intriligator, BW, Wright; Kutasov, Schwimmer) to get Expand scheme dependent but in general, we expect This means we can extract the scheme-independent part of This matches precisely with computations by Jack, Jones, North. However, the scheme-dependent terms do not match.

  21. To match up the scheme-dependent stuff too, we must take into account the wavefunction renormalization. (Barnes, Intriligator, BW, Wright; Kutasov and Schwimmer) So it looks like the Lagrange multipliers really work, and tell us something about physics. Not just formal! Now, back to the a-theorem!

  22. We can extend the Lagrange multipliers to other cases as well: Superpotential deformations Accidental symmetries from unitarity violations Higgsing? still working on it! In these cases, we can use Lagrange multipliers to prove the a-theorem, up to the caveat of additional accidental symmetries. (Kutasov; Barnes, Intriligator, BW, Wright) Let’s do the case of superpotential deformations: As before, just write with

  23. to increase this drives and a to decrease! Notice that this automatically implies gradient flow, just as in 2d! For example, consider the superpotential case: If , the superpotential is relevant We can use similar techniques to prove the a-theorem for accidental symmetries that come from unitarity violations, and hopefully for other cases as well.

  24. There’s another computation we can compare with as well: Osborn et al computed the metric in coupling space. (e.g. Yukawa) To leading order, there are no cross terms. Computation with Lagrange multipliers: These are EXACTLY the leading order terms obtained perturbatively by Osborn et al. This is evidence for the STRONGEST version of the a-theorem, that the RG flow is gradient flow.

  25. The Problem With Higgsing It is nontrivial (and still open) to show that a decreases after Higgsing! eaten matter singlet uneaten matter R 0 2/3 charged uneaten matter but So There’s a delicate balance – still needs to be proved!

  26. IV. Some New SCFTs We can use a-maximization to explore IR phases! Consider a theory with gauge group and matter content Naively, one would suspect that each group being interacting implies that both together are.

  27. But this isn’t necessarily true! Since the 2-loop pieces are positive, it’s possible for one group to drive the other to be IR free. You don’t need a-maximization to see this, since it’s basically just asking whether or not the fixed points are perturbatively stable.

  28. What about when one of the groups is magnetically free? Then we suspect we should dualize just that group, and analyze the resulting theory. The dual theory is easy to write down, but we can show that something interesting happens! It’s possible for the dyanmics of one group to drive the other to be interacting. Intuitively, this is because the Seiberg dual superpotential battles with the gauge contributions, and can win. (whether it does or not depends on Nc , Nf , etc.)

  29. Other possibilities? The only option left is that both groups look IR free, but end up flowing to an interacting point (example: N=4 SYM). We don’t have examples of this type. Can also check that conformal window exists when both interacting. It does! This is a plot of the conformal window for Nc = Nc’. The fact that is nonempty is a CHECK of Seiberg duality.

  30. IV. Conclusions, Open Questions, and Future Work a-maximization gives us access to lots of new information. New results for previously mysterious SCFTs! The a-theorem is “almost proved,” and we’re closing in. R-charges etc. all quadratic irrational numbers! Relation between Lagrange multipliers and couplings? Which scheme? Accidental symmetries? Can we say anything general? Pick a theory – can use a-maximization to study it!

More Related