PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for...

87
Quiver Scaling Regimes & Black Holes PILJIN YI KOREA INSTITUTE for ADVANCED STUDY KIAS-YITP 2013, Kyoto, July 2013

Transcript of PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for...

Page 1: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Quiver Scaling Regimes & Black Holes

PILJIN YIKOREA INSTITUTE for ADVANCED STUDY

KIAS-YITP 2013, Kyoto, July 2013

Page 2: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

start with Denef’s quiver quantum mechanics

Page 3: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Type IIB on CY3 4d N=2 theories

Page 4: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

D3 wrapped on a SL cycle in CY3 4d BPS particle

charged particle-likeBPS state in 4d

Page 5: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

D3’s wrapped on SL 3-cycles in CY3 quiver quantum mechanics for particle-like BPS states in 4d

Denef 2002

Page 6: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

where the main object of interest is the equivariant index

Page 7: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

which counts BPS states with 4 supercharged preserved

marginal stability wall

Page 8: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb versus Higgs

Page 9: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

large & “positive” FI constants

Higgs : wrapped D-branes are fused into a single object

Page 10: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Higgs “phase”

= assume large values of chiral multipletsand ignore heavy vector multiplets

Page 11: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Higgs “phase” ground states ~ cohomology Euler index

Page 12: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

or the Hirzebruch characteristic

Page 13: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Higgs “phases” have branches with different vacuum geometry

marginal stability wall

Page 14: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

small & “positive” FI constants

Coulomb : wrapped D-branes are separated along real space

Page 15: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb “phase”

= assume large values of vector multipletsand integrate out heavy chiral multiplets

Page 16: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb “phase”

= assume large values of vector multipletsand integrate out heavy chiral multiplets

Page 17: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb “phase”

= assume large values of vector multipletsand integrate out heavy chiral multiplets

Page 18: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

multi-center picture of BPS states

Page 19: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

1998 Lee + P.Y.N=4 SU(n) ¼ BPS states via semiclassical multi-center dyon solitons

1999 Bak + Lee + Lee + P.Y. N=4 SU(n) ¼ BPS states via semi-classical multi-center monopole dynamics

1999-2000 Gauntlett + Kim + Park + P.Y. / Gauntlett + Kim + Lee + P.Y. / Stern + P.Y.N=2 SU(n) BPS state counting via semi-classical multi-center monopole dynamics

2000 DenefN=2 supergravity via classical multi-center black holes attractor solutions

2001 Argyres + Narayan / Ritz + Shifman + Vainshtein + VoloshinUV-incomplete string-web picture for N=2 BPS dyons

2002 Denefquiver dynamics of BPS states / primitive wall-crossing formula

Page 20: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

wall-crossing ~ supersymmetric Schroedinger problem

Page 21: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

N=4 many body quantum mechanics,to be orbifolded by the Weyl symmetry

Denef 2002Sungjay Lee+P.Y. 2011

Heeyeon Kim+Jaemo Park+Zhao-Long Wang+P.Y, 2011

Page 22: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

each charge-center feels the long-range tails due to the rest

Sungjay Lee+P.Y. 2011

Page 23: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 24: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 25: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

deform & localize N=4 3(n-1) dimensional dynamicsN=1 2(n-1) dim nonlinear sigma model with U(1) bundle

Page 26: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

an index theorem before the Weyl division

Manschot+Pioline+ Sen 2010/2011Kim+Park+Wang+P.Y. 2011

Page 27: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

division by Weyl symmetries an iterative sum overfixed submanifolds under permutation of identical particles

Page 28: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

orbifolding of the index

Page 29: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

P.Y. 1997Green + Gutperle 1997

Kim+Park+Wang+P.Y. 2011

for p identical particles & with internal degeneracy

cf) Manschot + Pioline + Sen 2010/2011

Page 30: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

e.g., for an identical pair of unit degeneracy eachP.Y. 1997

Page 31: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Manschot+Pioline+Sen 2011Kim+Park+Wang+P.Y. 2011

universal wall-crossing formula from Coulomb ‘phase’ dynamics / real space dynamics

Page 32: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Manschot+Pioline+Sen 2011Kim+Park+Wang+P.Y. 2011

an Abelianization formulavia a sum over all partitions of charges with rational invariants

Page 33: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

this computes BPS bound state indexgiven input data wall-crossing formulae

marginal stability wall

Page 34: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

reduction to

Lee+Wang+P.Y. 2012

which is easily elevated to the equivariant index of the quiver as

Page 35: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Lee+Wang+P.Y. 2012

and can be easily evaluated via localization

Page 36: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

with all charges on a single plane of charge lattice,and in the absence of a scaling regime,

the resulting wall-crossing formula has been shown to be equivalent to the Kontsevich-Soibelman proposal

(Ashoke Sen, December 2011)

Page 37: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb versus Higgs

Page 38: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Denef 2002

small FI constantslarge FI constants

why?

Page 39: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

in quantum mechanics, the word “phase” is very misleading since vacuum expectation values do not imply

superselection sectors

what one really means by this word is a truncation process depending on where the ground state wavefunctions are localized;

at large values of chiral multipletsor at large values of vector multiplets

Page 40: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

versus

Page 41: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 42: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 43: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 44: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 45: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 46: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

in quantum mechanics, the word “phase” is very misleading since vacuum expectation values do not imply

superselection sectors

what one really means by this word is a truncation process depending on where the ground state wavefunctions are localized;

at large values of chiral multipletsor at large values of vector multiplets

as long as wavefunctions do not move away to infinite,and as long as the truncation process is reliable,

the supersymmetric index seems to be preserved

Page 47: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 48: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

F. Denef 2002 + A. Sen 2011

small FI constantslarge FI constants

Page 49: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Denef + Moore 2007

however, a simple 3-body problem says otherwise

Page 50: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

small FI constantslarge FI constants

why not ?

Page 51: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

practically, however, what one also means by “phase” is certain truncation processes where we integrate out either the chiral multiplets or the vector multiplets

however, this process can sometimes fail spectacularly, if the “heavy” multiplet in question become light somewhere in the classical vacuum moduli space

precisely this happens in the Coulomb “phase” scaling regime

Page 52: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb “phase”

Page 53: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb “phase”

Page 54: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 55: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb “phase” scaling regime

Page 56: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb “phase” scaling regime

Page 57: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 58: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 59: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 60: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

Coulomb “phase” scaling regime

Page 61: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

an ad hoc, canceling Laurent polynomial of degree less than n & of the same parity as

Coulomb “phase” scaling regimeManschot+Pioline+ Sen 2011

Page 62: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

an ad hoc, canceling Laurent polynomial of degree less than n & of the same parity as

Bena + Berkooz + de Boer + El-Showk + d. Bleeken, 2012

Lee+Wang+P.Y. 2012

Page 63: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 64: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

small FI constantslarge FI constants

1) loops in the quiver = superpotentials2) geometric inequality for linking numbers3) in all branches, Higgs “vacua” exist

S.J. Lee + Z.L. Wang + P. Y., 2012Bena + Berkooz + de Boer + El-Showk + d. Bleeken, 2012

Page 65: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

back to the simple 3-body example

Page 66: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

what physical & mathematical properties characterize these intrinsically Higgs, wall-crossing-safe BPS states ?

Page 67: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

quiver invariant

Page 68: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 69: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center
Page 70: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

wall-crossing vs. wall-crossing-safe

Page 71: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

wall-crossing vs. wall-crossing-safe

S.J. Lee + Z.L. Wang + P. Y., 2012Bena + Berkooz + de Boer + El-Showk + d. Bleeken, 2012

Page 72: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

general proof & explicit counting !

S.L. Lee + Z.L. Wang + P. Y., 2012Manschot + Pioline + Sen, 2012

Page 73: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

the total equivariant index ~ Hirzebruch character

Page 74: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

which is easily computable here, via Riemann-Roch theorem

Page 75: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

and decomposed into two parts

Page 76: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

wall-crossing states vs. wall-crossing-safe states

Page 77: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

wall-crossing states vs. wall-crossing-safe states

Page 78: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

wall-crossing states vs. wall-crossing-safe states

Page 79: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

this simple dichotomy, due to the Lefschetz hyperplane theorem,is literally true only for cyclic Abelian quivers:

for general quivers, the cohomology is far more intricate

Page 80: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

intrinsic Higgs states are likely to remain angular momentum singlets

Page 81: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

wall-crossing states vs. wall-crossing-safe states

many-body bound stateswall-crossing

angular momentummultiplets

single-center stateswall-crossing-safe

angular momentumsinglets

Page 82: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

wall-crossing states vs. wall-crossing-safe states

many-body bound stateswall-crossing

polynomial degeneracy:most of familiar BPS states in

field theories belong here

single-center stateswall-crossing-safe

exponential degeneracy:single-center BH’s

belong here

Page 83: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

more examples of quiver invariants

Page 84: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

more examples of quiver invariants

Page 85: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

black hole entropy ?

Page 86: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

outstanding issues

origin & validity of the MPS Coulomb prescription for scaling cases ?

is the Coulomb-like Abelianization routine true even for Higgs “phase” with quiver invariants ?

(in-)dependence of index on superpotential choices ?

detailed string theory embeddings and microscopic counting of BH entropy ?

Page 87: PILJIN YIstring13/Yi.pdfquiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver of N=2* theories to single-center

summary

d=1 N=4 quiver quantum mechanics offers a universal framework for wall-crossing / counting of 4d BPS states

with the intuitive Coulomb “phase” for wall-crossing &the comprehensive Higgs “phase” for faithful state counting

quiver invariants must/can be computed separately as input data for wall-crossing, and appear everywhere from the BPS quiver

of N=2* theories to single-center BPS black holes

complete derivation of the index for non-Abelian quivers, in the presence of quiver invariants, is not yet available but existing

Abelianization proposals suggest the quiver invariant as a measure of single-center black hole microstates