Loop Quantum Gravity : state of the art

Karim NOUI
Laboratoire de Mathematiques et de Physique Theorique, TOURS
Federation Denis Poisson
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 1/15
Overview
Quantum Gravity : Why and How?
1. Classical framework: Ashtekar variables
Ashtekar Gravity looks like Yang-Mills
2. Quantum Geometry
Polymer representation
Kinematical States and Geometric Operators
3. Quantum Dynamics?
From Wheeler-de Witt to Spin-Foams
Successes and failures
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 2/15
Quantum Gravity : a general discussion
WHY?
Physics of a relativistic system of mass m and length

c
=
h
mc
: quantum physics
r
s
=
Gm
c
2
: general relativity

c
r
s
=
p
: quantum gravity
Where is quantum gravity : at the GR singularities !
Hawking-Penrose theorem
Origin of the universe : quantum cosmology
Black holes : microscopic explanation of entropy
But quantum physics and general relativity are not compatible
Quantum Field theory based on a xed background
General Relativity is a non renormalisable theory
Related problems : time, observables and dieomorphisms etc...
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 3/15
Quantum Gravity : a general discussion
HOW?
One thinks that General Relativty fails at
p
GR is the Fermi model of a Standard model ?
Modify classical paradigms : String theory (extra-dimensions ...)
GR appears as an eective theory with corrections at
p
One thinks that Quantum methods fail for GR
New quantisation roads : Loop Quantum Gravity (polymer states)
Problem of singularities similar of H atom : classical instability but
existence of quantum ground state
Quantisation resolves singularities : discretisation, minimal length...
Two or more roads... for one solution !
Loops and Strings are orthogonal directions
For loops : GR is fundamental = background independence
For Strings : QFT with Fock spaces and so on...
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 4/15
Classical framework : Ashtekar variables
Many classical actions for General Relativity
Lagrangian formulation : all actions lead to GR equations
Einstein-Hilbert action : functional of the metric g
S
EH
[g] =

d
4
x

[g[R
Hilbert-Palatini action : functional of g and the connection
S
HP
[g] =

d
4
x

[g[R[g, ]
Cartan formalism : g

= e
I

e
J

IJ
and F = d +
S
C
[e, ] =

e e F[]) =

d
4
x

IJKL
e
I

e
J

F
KL

Ashtekar-Barbero-Holst action : generalisation of Cartan

S
ABH
[e, ] =

e e F[]) +
1

e e F[])
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 5/15
Classical framework : Ashtekar variables
Hamiltonian analysis : GR phase space
Hamiltonian formulation : M = R (61)
ADM variables : ds
2
= N
2
dt
2
(N
a
dt + h
ab
dx
b
)(N
a
dt + h
ac
dx
c
)
ADM action : (h, ) canonical variables
S
ADM
[h, ; N, N
a
] =

dt

d
3
x(

h + N
a
H
a
[h, ] + NH[h, ])
Equations, H = 0 = H
a
, very complicated and highly non-linear
Ashtekar formulation : originally = i (86)
Partial gauge xing (time gauge) : SL(2, C) SU(2)
Variables : A : SU(2)-connection and E : electric eld
Equations H = 0 = H
a
become polynomial !
real : same structure but H not polynomial
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 6/15
Classical framework : Ashtekar variables
A summary of the classical formulation
First order Lagrangian : variables are Cartan data
A tetrad e
I

such that g

= e
I

e
J

IJ
a sl (2, C) spin-connection =
i
R
i
+
0i
B
i
; F() its curvature
Classical action depends on the free parameter ,= 0
S
P
[e, ] =

e
I
e
J
(F
IJ
()
1

F
IJ
())
Time gauge : partial gauge x SL(2, C) to SU(2)
Hamiltonian analysis : similarities with Yang-Mills
New variables : E
a
i
=
1
2

ijk

abc
e
j
b
e
k
c
and A
i
a
=
i
a
+
0i
a
A
i
a
(x), E
b
j
(y) = (8G)
b
a

i
j

3
(x, y)
The constraints are almost polynomial
H
a
= F
i
ab
E
b
i
, H = (F
ij
ab
+ (
2
+ 1)K
i
[a
K
j
b]
)E
a
i
E
b
j
One more constraint : Gauss (
i
= D
a
E
a
i
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 7/15
Quantum geometry
Quantisation of a point particle
Algebra of quantum operators
Phase space : P, Q = 1
Quantisation leads to Weyl algebra [

P,

Q] = i
Quantum states from representation theory
Schrodinger representation : L
2
(R)
(

Q)(q) = q(q) , (

P)(q) = i
(q)
q
Fock like representation : [a, a

] = 1
[0) [n) (a

)
n
[0)
Stone-Von Neumann : unique representation !
Quantum Field Theory
Representation is not unique
The Fock representation is the good one
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 8/15
Quantum geometry
The Polymer representation
Schrodinger like representation
States are functionnals of connection : (A)

A acts by multiplication,

E as a derivation
Problem : no scalar product
1
[
2
)
?
=

[TA]
1
(A)
2
(A)
The polymer representation : states are one-dimensional
Let a graph : L links, V vertices
:
i
are oriented links
:n
i
are nodes

3
n
1
n
2
Let f a function on SU(2)
L
State :
,f
(A) = f (U

1
, , U

L
) where U

= P exp(

A) SU(2)
Ashtekar-Lewandowski measure :

,f
[

,f
) =
,

i
d(U

i
)) f (U

i
)f

(U

i
)
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 9/15
Quantum geometry
Imposing the constraints
The Gauss constraint
Gauge action : A A
g
= g
1
Ag +g
1
dg = U

g(s

)
1
U

g(t

)
States are invariant under gauge action
Orthonormal basis :
i
with spins I
i
and v
i
with Clebsh-Gordan
The dieomorphisms constraint
Dieomorphisms Di () on and A
States are now labelled by knots []
Unique representation compatible with Di ()
The Hamiltonian constraint

H = 0 is Wheeler-de Witt
Very few not interesting solutions
Thiemann trick to dene

H
Spin-foams models from covariant quantisation
So far, no physical solutions...
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 10/15
Quantum geometry
Physical interpretation and discretisation of the space
Area operator /(o) acting on H
0
Classical area of a surface o : /(o) =

n
a
E
a
i
n
b
E
b
i
d
2

Quantum area operator : o =

N
n
o
n
/(o) = lim
N

E
i
(o
n
)E
i
(o
n
) with E
i
(o
n
) =

S
n
E
i
Spectrum and Quanta of area
o

/(o)[S) =
8G
c
3

PS

j
P
(j
P
+ 1)[S)
Volume operator 1() acting on H
0
Classical volume on a domain : 1() =

R
d
3
x

|
abc

ijk
E
ai
E
bj
E
ck
|
3!
It acts on the nodes of [S) : discrete spectrum
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 11/15
Quantum geometry
Picture of space at the Planck scale
From the kinematics, Space is discrete...
... It is also non-commutative in 3 dimensions
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 12/15
Quantum Dynamics ?
Hamitonian constraint
Classical Hamiltonian constraint
First part of Hamiltonian :
H(N) =

d
3
x N(x)
Tr (F
ab
E
a
E
b
)

[det(E)[
Regularization using Thiemann trick :
H(N) =
1

N(x)Tr (F(x) A(x), V(R

x
))
V(R
x
) is the volume of a region R
x
around x
Quantization of the constraint :
V is a well-dened positive self-adjoint operator on H
It creates new edges on Spin-network states
Ultra-locality : action is conned around a vertex
Ambiguities : ordering, representations etc...
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 13/15
Quantum Dynamics ?
An alternative solution : Spin-Foam models
Transition Amplitudes between states
/ = S[S

)
phys
From Topological QFT
Relation to LQG not clear
Some promissing models
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 14/15
Discussion
Successes and failures
Very interesting program for Quantum Gravity
A new quantisation scheme : polymer representation
Uniqueness theorem of the representation
Complete description of kinematical states
Discrete spectrum of area : Black Hole entropy, Cosmology
But NO physical states
No dierence with a topological theory
Role of Immirzi parameter is unclear
Value xed by S = A/4
Compact vs. non compact gauge group ?
LPSC Grenoble - february 2010 Karim NOUI Loop Quantum Gravity in a nutshell 15/15