On the possible mathematical connections between the Hartle-Hawking no boundary

proposal concerning the Randall-Sundrum cosmological scenario, Hartle-Hawking wave-

function in the mini-superspace sector of physical superstring theory, p-adic Hartle-Hawking
wave function and some sectors of Number Theory.

Michele Nardelli 1, 2

1
Dipartimento di Scienze della Terra
Università degli Studi di Napoli Federico II, Largo S. Marcellino, 10
80138 Napoli, Italy

2
Dipartimento di Matematica ed Applicazioni “R. Caccioppoli”
Università degli Studi di Napoli “Federico II” – Polo delle Scienze e delle Tecnologie
Monte S. Angelo, Via Cintia (Fuorigrotta), 80126 Napoli, Italy

Abstract

In this paper we have described the Hartle-Hawking no boundary proposal concerning the Randall-
Sundrum cosmological scenario, nonlocal braneworld action in the two-brane Randall-Sundrum
model, Hartle-Hawking wave-function in the mini-superspace sector of physical superstring theory,
p-adic models in the Hartle-Hawking proposal and p-adic and adelic wave functions of the universe.
Furthermore, we have showed some possible mathematical connections between some equations of
these arguments and, in conclusion, we have also described some mathematical connections
between some equations of arguments above mentioned and some equations concerning the
Riemann zeta function, the Ramanujan’s modular equations and the Palumbo-Nardelli model.
In the section 1, we have described the Hartle-Hawking “no boundary” proposal applied to Randall-
Sundrum cosmological scenario. In the section 2, we have described nonlocal braneworld action in
the two-brane Randall-Sundrum model. In the section 3, we have described the compactifications
of type IIB strings on a Calabi-Yau three-fold and Hartle-Hawking wave-function in the mini-
superspace sector of physical superstring theory. In the section 4, we have described the p-Adic
models in the Hartle-Hawking proposal. In the section 5, we have described the p-Adic and Adelic
wave functions of the Universe. In the section 6, we have described some equations concerning the
Riemann zeta function, specifically, the Goldston-Montgomery Theorem, the study of the behaviour
of the argument of the Riemann function ζ (s ) with the condition that s lies on the critical line
1
s = + it , where t is real, the P-N Model (Palumbo-Nardelli model) and the Ramanujan identities.
2
In conclusion, in the section 7, we have described some possible mathematical connections between
some equations of arguments above discussed and some equations concerning the Riemann zeta-
function, the Ramanujan’s modular equations and the Palumbo-Nardelli model.

1
1. Hartle-Hawking “No Boundary” proposal applied to Randall-Sundrum cosmological
scenario. Randall-Sundrum from AdS/CFT, CFT in the Domain Wall. [1]

The AdS/CFT correspondence relates IIB supergravity theory in AdS 5 × S 5 to a Ν = 4U ( N )

superconformal field theory. If gYM is the coupling constant of this theory then the‘t Hooft
2
parameter is defined to be λ = g YM N . The CFT parameters are related to the supergravity
parameters by
l 3 2N 2
l = λ1/ 4ls , (1.1) = , (1.2)
G π

where ls is the string length, l the AdS radius and G the five dimensional Newton constant. Note
that λ and N must be large in order for stringy effects to be small. The CFT lives on the conformal
boundary of AdS 5 . The correspondence takes the following form:

Z [h] ≡ ∫ d [g]exp(− S grav [g]) = ∫ d [φ ]exp(− S CFT [φ ; h]) ≡ exp(− WCFT [h]) , (1.3)

here Z [h] denotes the supergravity partition function in AdS 5 . This is given by a path integral over
all metrics in AdS 5 which induce a given conformal equivalence class of metrics h on the
conformal boundary of AdS 5 . A problem with equation (1.3) as it stands is that the usual
gravitational action in AdS is divergent, rendering the path integral ill-defined. A procedure for
solving this problem is the following: first one brings the boundary into a finite radius, next one
adds a finite number of counterterms to the action in order to render it finite as the boundary is
moved back off to infinity. These counterterms can be expressed solely in terms of the geometry of
the boundary. The total gravitational action for AdS d +1 becomes

S grav = S EH + S GH + S1 + S 2 + .... (1.4)

The first term is the usual Einstein-Hilbert action with a negative cosmological constant:

1  d (d − 1) 
S EH = −
16πG ∫ d d +1 x g  R +
 l 2 
(1.5)

the overall minus sign arises because we are considering a Euclidean theory. The second term in the
action is the Gibbons-Hawking boundary term, which is necessary for a well-defined variational
problem:
1
8πG ∫
S GH = − d d x h K , (1.6)

where K is the trace of the extrinsic curvature of the boundary and h the determinant of the
induced metric. The first two counterterms are given by the following expressions:

d −1 d l
S1 =
8πGl ∫
d x h, (1.7) S2 =
16πG (d − 2 ) ∫ d d x hR , (1.8)

where R now refers to the Ricci scalar of the boundary metric. The third counterterm is

2
 l3  d 
S3 =
16πG (d − 2 ) (d − 4 )
2 ∫ d d x h  Rij R ij −

R2  ,
4(d − 1) 
(1.9)

where Rij is the Ricci tensor of the boundary metric and boundary indices i, j are raised and
lowered with the boundary metric hij . This expression is ill-defined for d = 4, which is the case of
most interest to us. We can now use the AdS/CFT correspondence to explain the behaviour
discovered by Randall and Sundrum. The (Euclidean) RS model has the following action:

S RS = S EH + SGH + 2S1 + S m . (1.10)

Here 2S1 is the action of a domain wall with tension (d − 1) / (4πGl ) . The final term is the action for
any matter present on the domain wall. We want to study quantum fluctuations of the metric on the
domain wall. Let g0 denote the five dimensional background metric we have just described and h0
the metric it induces on the wall. Let h denote a metric perturbation on the wall. If we wish to
calculate correlates of h on the domain wall then we are interested in a path integral of the form

hij ( x )hi ' j ' ( x') = ∫ d [h]Z [h]hij ( x )hi ' j ' (x') , (1.11)

where
Z [h] = ∫ d [δg]d [φ ]exp(− S RS [g0 + δg]) = exp(− 2 S1 [h0 + h]) ×
B1 ∪ B2

×∫ d [δg]d [φ ]exp(− S EH [g0 + δg] − S GH [g0 + δg] − S m [φ ; h0 + h]) , (1.12)

B1 ∪ B2

δg denotes a metric perturbation in the bulk that approaches h on the boundary and φ denotes the
matter fields on the domain wall. The integrals in the two balls are independent so we can replace
the path integral by
(
Z [h] = exp(− 2 S1 [h0 + h]) ∫ d [δg]exp(− S EH [g0 + δg] − S GH [g0 + δg]) ×
B
)
2

× ∫ d [φ ]exp(− S m [φ ; h0 + h]) , (1.13)

where B denotes either ball. We now take d = 4 and use the AdS/CFT correspondence (1.3) to
replace the path integral over δg by the generating functional for a conformal field theory:

∫ d [δg]exp(− S [g
B
EH 0 + δg] − S GH [g0 + δg]) =
exp(− WRS [h0 + h] + S1 [h0 + h] + S 2 [h0 + h] + S 3 [h0 + h]) . (1.14)

This is the RS (Randall-Sundrum) CFT since it arises as the dual of the RS geometry.
Now we will consider the RS analogue of Starobinsky’s model by putting a CFT on the domain
wall. Our five dimensional (Euclidean) action is the following:

S = S EH + S GH + 2 S1 + WCFT . (1.15)

We seek a solution in which two balls of AdS 5 are separated by a spherical domain wall. Inside
each ball, the metric can be written

3
 (
ds 2 = l 2 dy 2 + sinh 2 ydΩ 2d , ) (1.16)

with 0 ≤ y ≤ y0 . The domain wall is at y = y0 and has radius

R = l sinh y0 . (1.17)

The effective tension of the domain wall is given by the Israel equations as

3
σ eff = coth y0 . (1.18)
4πGl

The actual tension of the domain wall is

3
σ= . (1.19)
4πGl

We therefore need a contribution to the effective tension from the CFT. This is provided by the
conformal anomaly, which takes the value

3N 2
T =− . (1.20)
8π 2 R 4

This contributes an effective tension − T / 4 . We can now obtain an equation for the radius of the
domain wall:
R3 R2 N 2G R 4
+ 1 = + . (1.21)
l3 l2 8πl 3 l 4

It is easy to see that this has a unique positive solution for R .

The AdS/CFT correspondence can be used to give the generating functional of the CFT on the
perturbed sphere:
WCFT = S EH + S GH + S1 + S 2 + .... (1.22)

We shall give the terms on the right hand side for d = 4. The Einstein-Hilbert action with
cosmological constant is
1  12 
S EH = −
16πG ∫ d 5 x g  R + 2  , (1.23)
 l 

and perturbing this gives

1  8 1 1 
S bulk = −
16πG ∫ d 5 x g  − 2 + h µν ∇ 2 hµν + 2 h µν hµν 
 l 4 2l 
1  1 3 
−
16πG ∫ d 4 x γ  − n µ hνρ ∇ν hµρ + hνρ n µ ∇ µ hνρ  ,
 2 4 
(1.24)

where Greek indices are five dimensional and we are raising and lowering with the unperturbed five
dimensional metric. n = ldy is the unit normal to the boundary and ∇ is the covariant derivative
defined with the unperturbed bulk metric. γ ij = R 2γˆij is the unperturbed boundary metric.
Evaluating on shell gives

4
 l3 y0 l3  3 coth 
∫ ∫ ∫
4 4
S EH = d x γˆ dy sinh y − d 4 x γˆ  4 h ij ∂ y hij − 4 h ij hij  , (1.25)
2πG 0 16πG  4l l 

where we are now raising and lowering with γˆij . The Gibbons-Hawking term is

l3  1 
S GH =−
2πG ∫ d 4 x γˆ  sinh 3 y0 cosh y0 − 4 h ij ∂ y hij  .
 8l 
(1.26)

The first counter term is

3 3l 3  1 
∫ ∫
4
S1 = d x γ = d 4 x γˆ  sinh 4 y0 − 4 h ij hij  . (1.27)
ˆ
8πGl 8π G  4l 

The second counter term is

l l3  2 1 ˆ 2 h  .
∫ ∫ d 4 x γˆ 12 sinh 2 y0 − 4
4
S2 = d x γ R = 2
h ij hij + 4 2
h ij ∇ ij 
32πG 32πG  l sinh y0 4l sinh y0 
(1.28)
Thus with only two counter terms we would have

3N 2Ω 4 R l3  1 1 ij  3 l 2  1 ij 1 
− 
l 16πG ∫
4 ij ij ˆ 2
WCFT = log − d x γˆ h ∂ h + h h − 1 + + h h − h ∇ h
8π 2  4l 4 y ij
l4
ij
2 R 2 
l 2 2
R
ij
8l 2 2
R
ij

   
(1.29)

l 3 2N 2
Ω 4 is the area of a unit four-sphere and we have used the following equation: = . The
G π
expansion of ∂ y hij at y = y0 is obtained from

f p' ( y0 )
∂ y hij = ∑ H ij( p ) ( x )∫ d 4 x' γˆ h kl ( x')H kl( p ) ( x') (1.30)
p f p ( y0 )

and
f p' ( y0 ) l2 l4 l4
f p ( y0 )
=2+
2R 2
( p + 1)( p + 2 ) + p ( p + 1)( p + 2 )( p + 3)
4R 4
log (l / R ) +
8R 4
p4 + 2 p3 [
 l6 
− 5 p 2 − 10 p − 2 − p ( p − 1)( p + 2 )( p + 3)(ψ (1) + ψ (2) − ψ ( p / 2 + 2 ) − ψ ( p / 2 + 5 / 2 ))] + Ο 6 log(l / R )
R 
(1.31)

The psi function is defined by ψ ( z ) = Γ' (z ) / Γ( z ) . Substituting into the action we find that the
divergences as l → 0 cancel at order R 4 / l 4 and R 2 / l 2 . The term of order l 4 / R 4 in the above
expansion makes a contribution to the finite part of the action:

5
 3N 2Ω 4 N2
∑ (∫ d x' γˆ h (x')H ( ) (x'))
R 4 kl p
2
WCFT = log + kl
8π 2 l 256π 2 R 4 p

(2 p( p + 1)( p + 2)( p + 3) log(l / R ) + Ψ( p )) , (1.32)

where
Ψ ( p ) = p ( p + 1)( p + 2 )( p + 3)[ψ ( p / 2 + 5 / 2 ) + ψ ( p / 2 + 2 ) − ψ (2 ) − ψ (1)] + p 4 + 2 p 3 − 5 p 2 − 10 p − 6

(1.33)

To cancel the logarithmic divergences as l → 0 , we have to introduce a length scale ρ defined by

l = ερ and add a counter term proportional to log ε to cancel the divergence as ε tends to zero.
The counter term is
l3  1 
S3 = − log ε ∫ d 4 x γ  γ ik γ jl Rij Rkl − R 2 
64πG  3 
l 3
 1  3 ˆ2 1 ˆ 4 
=− log ε ∫ d 4 x γˆ  − 12 + 4 2h ij hij − h ij ∇ hij + h ij ∇ hij   . (1.34)
64πG  R  2 4 

This term does indeed cancel the logarithmic divergence, leaving us with

3N 2 Ω 4 N2
∑ (∫ d )
R 2
WCFT = log + 4
x' γˆ h kl ( x') (2 p ( p + 1)( p + 2 )( p + 3) log(ρ / R ) + Ψ ( p )) .
8π 2 ρ 256π 2 R 4 p

(1.35)

Now, recall that our five dimensional action is

S = S EH + S GH + 2 S1 + WCFT . (1.36)

In order to calculated correlators of the metric, we need to evaluate the path integral

Z [h] = ∫ d [δg]exp(− S ) =
B1 ∪ B2

(
= exp(− 2 S1 [h0 + h] − WCFT [h0 + h]) ∫ d [δg]exp(− S EH [g0 + δg] − S GH [g0 + δg]) .
B
)2
(1.37)

Here g0 and h0 refer to the unperturbed background metrics in the bulk and on the wall
respectively and h denotes the metric perturbation on the wall. Replacing l and G with l and G ,
from equation (1.27) we obtain
3l 3  1 
S1 [h0 + h] = ∫ d 4 x gˆ  sinh 4 y0 − 4  , (1.38)
8πG  4l 

where y0 is defined by R = l sinh y0 . The path integral over δg is performed by splitting it into a
classical and quantum part:
δg = h + h' , (1.39)

where the boundary perturbation h is extended into the bulk using the linearized Einstein equations
and the requirement of finite Euclidean action. h' denotes a quantum fluctuation that vanishes at

6
the domain wall. The gravitational action splits into separate contributions from the classical and
quantum parts:
S EH + S GH = S 0 [h] + S ' [h'] , (1.40)

where S 0 can be read off from the equations (1.25) and (1.26) as

3l 3Ω 4 y0 l3  1 coth y0 ij 
∫ ∫
2 2
S0 = − dy sinh y0 cosh y0 + d 4 x γˆ  4 h ij ∂ y hij + h hij  . (1.41)
2πG 0 16πG  4l l4 

We shall not need the explicit form for S ' since the path integral over h' just contributes a factor of
some determinant Z 0 to Z [h]. We obtain

Z [h] = Z 0 exp(− 2S 0 [h0 + h] − 2 S1 [h0 + h] − WCFT [h0 + h]) . (1.42)

The exponent is given by

3l 3Ω 4 y0 3Ω 4 R 4 3 N 2 Ω 4 R
2 S 0 + 2 S1 + WCFT = −
πG ∫ 0
dy sinh 2 y cosh 2 y +
4πGl
+
8π 2
log
ρ
 l 3  f p' ( y0 ) 
1
+ 4
l
( 4 kl ( p)
)
∑p ∫ d x' γˆ h (x')H kl (x')  32πG  f ( y ) + 4 coth y0 − 6 
2

  p 0 
N 2

+ 2 4
(2 p( p + 1)( p + 2)( p + 3) log(ρ / R ) + Ψ ( p )) . (1.43)
256π sinh y0 

The (Euclidean) graviton correlator can be read off from the action as

128π 2 R 4 ∞
hij ( x )hi ' j ' ( x') =
N2
∑ W ( ) (x, x')F ( p, y )
p =2
p
iji ' j ' 0
−1
(1.44)

where we have eliminated l 3 / G using the equation (1.21). The function F ( p, y0 ) is given by

 f p' ( y0 ) 
F ( p, y0 ) = e y0 sinh y0  + 4 coth y0 − 6  + Ψ ( p ) , (1.45)
 f (y ) 
 p 0 

and the bitensor Wiji( p' j)' (x, x') is defined as

Wiji( p' j)' (x, x') = ∑ H ( ) (x )H ( ) (x') ,
k ,l ,m ,...
ij
p p
i' j' (1.46)

with the sum running over all the suppressed labels k , l , m,... of the tensor harmonics.
Now, we consider the radius R of the domain wall given by equation (1.21). It is convenient to
write this in terms of the rank N RS of the RS CFT (given by l 3 / G = 2 N RS 2
/π )

R3 R2 N2 R4
+ 1 = + . (1.47)
l3 l2 2
16 N RS l4

7
If we assume N >> N RS >> 1 then the solution is

 N RS
) .
2
R N
=
l 2 2 N RS
1 +
N 2
4
+ Ο N RS (
/ N4 (1.48)
 

Note that this implies R >> l , i.e., the domain wall is large compared with the anti-de Sitter length
scale. Now let’s turn to a four dimensional description in which we are considering a four sphere
with no interior. The only matter present is the CFT. The metric is simply

ds 2 = R42γˆij dx i dx j , (1.49)

where R4 remains to be determined. The action is the four dimensional Einstein-Hilbert action
(without cosmological constant) together with WCFT . There is no Gibbons-Hawking term because
there is no boundary. Without a metric perturbation, the action is simply

1 3Ω 4 R42 3 N 2 Ω 4 R
∫
4
S =− d x γ R + WCFT = − + 2
log 4 . (1.50)
16πG4 4πG4 8π ρ

where G4 is the four dimensional Newton constant. Varying R4 gives

N 2 G4
R42 = , (1.51)
4π

and N is large hence R4 is much greater than the four dimensional Planck length.
Now we can to include the metric perturbation. The perturbed four dimensional Einstein-Hilbert
action is
1  2 1 ij ˆ 2 
∫
(4 )
S EH =− d 4 x γˆ 12 R42 − 2 h ij hij + h ∇ hij  . (1.52)
16πG4  R4 4 R42 

Adding the perturbed CFT gives

3N 2Ω 4 3N 2Ω 4 2
S=−
16π 2
+
8π 2
R
ρ
(
log 4 + ∑ ∫ d 4 x' γˆ h kl ( x')H kl( p ) (x') 
64πG
1
R 2
)
p2 + 3 p + 6 ( )
p  4 4

N2 
+ 2 4
(2 p( p + 1)( p + 2)( p + 3) log(ρ / R4 ) + Ψ ( p )) . (1.53)
256π R4 

Setting ρ = R4 we find that the graviton correlator for a four dimensional universe containing the
CFT is
∞
[ ]
hij ( x )hi ' j ' ( x') = 8 N 2G42 ∑ Wiji( p' j)' ( x, x') p 2 + 3 p + 6 + Ψ ( p ) .
−1
(1.54)
p =2

Hence, we have computed the graviton correlator using the Hartle-Hawking “No Boundary”
proposal.

8
We shall show how the Euclidean correlator calculated above is analytically continued to give a
correlator for Lorentzian signature.
We begin by continuing the graviton correlator (equation 1.44) obtained via the five dimensional
calculation. The analytic continuation of the correlator for four dimensional gravity (equation 1.54)
is completely analogous. In terms of the new label p ' , the Euclidean correlator 1.44 between two
points on the wall is given by
128π 2 R 4 i∞ ( p ' )
hij (Ω )hi ' j ' (Ω') =
N2
∑ Wiji ' j ' (µ )G ( p ' , y0 )
p '= 7 i / 2
−1
(1.55)

where
 g 'p ' ( y0 ) 
G ( p ' , y0 ) = F (− ip '−3 / 2, y0 ) = e sinh y0  y0
+ 4 coth y0 − 6
 g p ' ( y0 ) 
+ ( p ' −4ip ' + p ' / 2 − 5ip '−63 / 16 + ( p ' +1 / 4 )( p ' +9 / 4 )[ψ (− ip ' / 2 + 5 / 4 ) + ψ (− ip ' / 2 + 7 / 4 )
4 3 2 2 2

− ψ (1) − ψ (2 )]) ,
(1.56)

with g p ' ( y ) = Q−2ip '−1/ 2 (coth y ) . The function G ( p ' , y0 ) is real and positive for all values of p' in the
sum and for arbitrary y0 ≥ 0 . We have the Euclidean correlator defined as an infinite sum.
Now we write the sum in equation (1.55) as an integral along a contour C1 encircling the points
p ' = 7i / 2,9i / 2,...ni / 2 , where n tends to infinity. This yields

− i 64π 2 R 4
hij (Ω )hi ' j ' (Ω') = ∫ dp ' tanh p ' πWiji( p' j' ') (µ )G ( p ' , y0 ) .
−1
(1.57)
N2 C1

When the contribution from the closing of the contour in the upper half p' -plane vanishes, the final
result for the Euclidean correlator reads

− i 64π 2 R 4  +∞
hij (Ω )hi ' j ' (Ω') = ∫ dp ' tanh p ' πWiji( p' j' ') (µ )G ( p ' , y0 )
−1

N 2 
 − ∞
2
(
+ 2π ∑ tan Λ k πWiji(i'Λj 'k ) (µ )Res G ( p' , y0 ) ; iΛ k .
−1
)] (1.58)
k =1

The analytic continuation from a four sphere into Lorentzian closed de Sitter space is given by
setting the polar angle Ω = π / 2 − it . We may take µ = Ω , and µ then continues to π / 2 − it .
In conclusion, we note that the Lorentzian tensor Feynman (time-ordered) correlator is

128π 2 R 4  + ∞
N 2  ∫0
hij ( x )hi ' j ' ( x') = dp ' tanh p ' πWijiL'(jp' ' ) (µ )R G ( p ' , y0 ) (−1
)
2
128π 2 R 4  +∞
(
+ π ∑ tan Λ k πWijiL'(jiΛ' k ) (µ )Res G ( p' , y0 ) ; iΛ k + i
−1

N 2
)]
 ∫ 0
dp 'WijiL'(jp' ' ) (µ )R G ( p ' , y0 )
−1
( )
k =1
2
(
− π ∑ WijiL'(ji'Λ k ) (µ )Res G ( p ' , y0 ) ; iΛ k .
−1
)] (1.59)
k =1

9
In this integral the bitensor WijiL'(jp' ' ) (µ ( x, x')) may be written as the sum of the degenerate rank-two
tensor harmonics on closed de Sitter space with eigenvalue λ p ' = p '2 +17 / 4 of the Laplacian. ( )
We can understand the effect of the strongly coupled CFT on the microwave fluctuation spectrum
by comparing the result (1.59) with the transverse traceless part of the graviton propagator in four-
dimensional de Sitter spacetime. On the four-sphere, this is easily obtained by varying the Einstein-
Hilbert action with a cosmological constant. In terms of the bitensor, this yields

i∞ Wiji( p' j'') (µ (Ω, Ω'))

hij (Ω )hi ' j ' (Ω') = 32πG4 R 2 ∑
p '=7 i / 2 λ p' − 2
, (1.60)

which continues to
+∞ dp '
hij ( x )hi ' j ' ( x') = 32πG4 R 2 ∫ WijiL'(jp' ' ) (µ (x, x')) . (1.61)
0 λ p' − 2

We note that this can be compared with equation (1.59).

2. Nonlocal braneworld action in the two-brane Randall-Sundrum model. [2]

In the following definition

exp(iS eff [φ ]) = ∫ DΦ exp(iS [Φ ]) Φ (Σ )=φ , (2.1)
the effective action by construction depends on the four-dimensional fields associated with
brane(s)1. The number of these fields equals the number of branes, geometrically each field being
carried by one of the branes in the system. In the generalized Randall-Sundrum setup, the
braneworld effective action is generated by the path integral of the type (2.1),

∫ DG exp(iS [G, g , φ ]) 4
G ( Σ )= g
= exp(iS eff [g , φ ]) , (2.1a)

where the integration over bulk metrics runs subject to fixed induced metrics on the branes – the
arguments of S eff [g , φ ] . Here S [G , g , φ ] is the action of the five-dimensional gravitational field with
the metric G = G AB ( x, y ) , A = (µ ,5) , µ = 0,1,2,3, propagating in the bulk spacetime
(x = (x, y ), x = x , x = y ), and matter fields φ are confined to the branes Σ I - four-dimensional
A µ 5

timelike surfaces embedded in the bulk,

 1 
S [G , g , φ ] = S 5 [G ] + ∑ ∫ d 4 x Lm (φ , ∂φ , g ) − g 1/ 2σ I + [K ] , (2.2)
I
ΣI
 8πG5 
1
S 5 [G ] = ∫
16πG5 M 5
(
d 5 xG1/ 2 5 R(G ) − 2Λ 5 . (2.3) )

The branes are enumerated by the index I and carry induced metrics g = g µν ( x ) and matter field
Lagrangians Lm (φ , ∂φ , g ) . The bulk part of the action contains the five-dimensional gravitational

1
The scope of this formula is very large, because it arises in very different contexts. In particolar, its Euclidean version
(iS → − S Euclid ) underlies the construction of the no-boundary wavefunction in quantum cosmology. (Hartle-
Hawking Phys. Rev. D 28, 2960 (1983))

10
and cosmological constants, G5 and Λ 5 , while the brane parts have four-dimensional cosmological
constants σ I . The bulk cosmological constant Λ 5 is negative and, therefore, is capable of
generating the AdS geometry, while the brane cosmological constants play the role of brane
tensions σ I and, depending on the model, can be of either sign. The Einstein-Hilbert bulk action
(2.3) is accompanied by the brane “Gibbons-Hawking” terms containing the jump of the extrinsic
curvature trace [K ] associated with both sides of each brane. In the tree-level approximation the
path integral (2.1a) is dominated by the stationary point of the action (2.2). Its variation is given as a
sum of five- and four-dimensional integrals,

1  1 
δS [G , g , φ ] = − ∫ d 4 xdyG1/ 2  5 R AB − 5 RG AB + Λ 5G AB δG AB ( x, y )
16πG5  2 

+ ∑ ∫ d 4 xg 1/ 2  −
1 1
[
K µν − g µν K + T µν − g µν σ ] ( )δg (x ) ,
µν (2.4)
I
ΣI
 16πG5 2 

[ ]
where K µν − g µν K denotes the jump of the extrinsic curvature terms across the brane, and T µν ( x )
is the corresponding four-dimensional stress-energy tensor of matter fields on the branes,

2 δS m [g , φ ]
T µν ( x ) = , (2.5) S m [g , φ ] = ∑ ∫ d 4 xLm (φ , ∂φ , g ) , (2.6)
g 1/ 2 δg µν (x ) I
ΣI

hence
2 1
T µν ( x ) = δ ∑ ∫ d 4 xLm (φ , ∂φ , g ) . (2.6a)
g 1/ 2
δg µν ( x ) I
ΣI

The action is stationary when the integrands of both integrals in (2.4) vanish, which gives rise to
Einstein equations in the bulk,

δS [G, g , φ ] 1  1 
≡− G1/ 2  5 R AB − G AB 5 R + Λ 5G AB  = 0 , (2.7)
δG AB ( x, y ) 16πG5  2 

which are subject to (generalized) Neumann type boundary conditions – the well-known Israel
junction conditions –

δS [G , g , φ ] 1 1
≡− g 1/ 2 [K µν − g µν K ] + g 1/ 2 (T µν − g µν σ ) = 0 , (2.8)
δg µν ( x ) 16πG5 2

or to Dirichlet type boundary conditions corresponding to fixed (induced) metrics on the branes,
with δg µν = 0 in the variation (2.4),
4
Gµν Σ = g µν (x ) . (2.9)

The solution of the latter, Dirichlet, problem is obviously a functional of brane metrics,
G AB = G AB [g µν ( x )], and it enters the tree-level approximation for the path integral (2.1a). S eff [g , φ ]
in this approximation reduces to the original action (2.2)-(2.3) calculated on this solution
G AB [g µν ( x )], S eff [g , φ ] = S [G[g ], g , φ ] + O(h ) . With this definition, the matter part of effective action
coincides with the original action Eq. (2.6)

11
 S eff [g , φ ] = S 4 [g ] + S m [g , φ ] , (2.10)

while all non-trivial dependence on g arising from the functional integration is contained in S 4 [g ] .
Given the action (2.10) as a result of solving the Dirichlet problem (2.7), (2.9), one can further
apply the variational procedure, with respect to the induced metric g µν , to get the effective
equations
δS eff [g , φ ] δS 4 [g ] 1 1/ 2 µν
= + g T ( x ) = 0 , (2.11)
δg µν ( x ) δg µν ( x ) 2

which are equivalent to the Israel junction conditions – a part of the full system of the bulk-brane
equations of motion (2.7), (2.8).
The action of the two-brane Randall-Sundrum model is given by Eq. (2.2) in which the index I = ±
enumerates two branes with tensions σ ± . The fifth dimension has the topology of a circle labelled
by the coordinate y , − d < y ≤ d , with an orbifold Z 2 -identification of points y and − y . The
branes are located at antipodal fixed points of the orbifold, y = y± , y+ = 0 , y− = d . When they are
empty, Lm (φ , ∂φ , g µν ) = 0 , and their tensions are opposite in sign and fine-tuned to the values of Λ 5
and G5 ,
6 3
Λ5 = − 2
, σ + = −σ − = , (2.12)
l 4πG5l

this model admits a solution with an AdS metric in the bulk (l is its curvature radius),

−2 y / l
ds 2 = dy 2 + e η µν dx µ dxν , (2.13)

0 = y+ ≤ y ≤ y− = d , and with a flat induced metric η µν on both branes. The metric on the negative
tension brane is rescaled by the warp factor exp(− 2d / l ) providing a possible solution for the
hierarchy problem. With the fine tuning (2.12) this solution exists for arbitrary brane separation d -
two flat branes stay in equilibrium. Their flatness is the result of compensation between the bulk
cosmological constant and brane tensions.
Now consider the Randall-Sundrum model with small matter sources for metric perturbations
hAB ( x, y ) on the background of this solution,

ds 2 = dy 2 + e η µν dx µ dxν + hAB ( x, y )dx A dx B , (2.14)

−2 y / l

such that this five-dimensional metric induces on the branes two four-dimensional metrics of the
form
±
g µν (x ) = a±2η µν + hµν± (x ) . (2.15)

Here the scale factors a± = a ( y ± ) can be expressed in terms of the interbrane distance

a+ = 1 , a− = e −2 d / l ≡ a , (2.16)

12
 ±
and hµν (x ) are the perturbations by which the brane metrics g µν± (x ) differ from the (conformally)
flat metrics of the Randall-Sundrum solution (2.13). The main result is the braneworld effective
action (2.10) calculated for the boundary conditions (2.9) of this perturbed form (2.15). The
braneworld effective action is invariant under the four-dimensional diffeomorphisms acting on the
branes. In the linearized approximation they reduce to the transformations of metric perturbations,

± ±
hµν → hµν + f µ±,ν + fν±,µ (2.17)

with two independent local vector field parameters f µ± = f µ± ( x ) . Therefore, rather than in terms of
metric perturbations themselves, the action in question is expressible in terms of the tensor
invariants of these transformations – linearized Ricci tensors of hµν = hµν ±
(x ) ,

1
Rµν =
2
(
− hµν + hνλ,λµ + hµλ,λν − h,µν , ) (2.18)

−
on flat four-dimensional backgrounds of both branes. (Strictly speaking, Rµν is the linearized Ricci
−
tensor of the artificial metric η µν + hµν . It differs from the linearized Ricci curvature of the second
( )
brane, Rµν a 2η + h − = Rµν
−
/ a 2 , by a factor of a 2 ). Commas denote partial derivatives, raising and
lowering of braneworld indices here and everywhere is performed with the aid of the flat four-
dimensional metric η µν , hνλ ≡ η λσ hσν , h ≡ η µν hµν , R = η µν Rµν , and denotes the flat spacetime
d’Alambertian
= η µν ∂ µ ∂ν . (2.19)

Hence, we have to describe the variables which determine the embedding of branes into the bulk.
Due to metric perturbations the branes no longer stay at fixed values of the fifth coordinate. Up to
four-dimensional diffeomorphism (2.17), their embedding variables consist of two four-dimensional
scalar fields – the radions ψ ± ( x ) – and the braneworld action can depend on these scalars. Their
geometrically invariant meaning is revealed in a special coordinate system where the bulk metric
perturbations hAB ( x, y ) of Eq. (2.14) satsfy the so called Randall-Sundrum gauge conditions,
,ν
hA5 = 0 , hµν = hµµ = 0 . In this coordinate system the brane embeddings are defined by the equations

l ±
Σ ± : y = y± + ψ (x ) , y+ = 0 , y− = d . (2.20)
a±2

In the approximation linear in perturbation fields and vector gauge parameters, these radion fields
are invariant under the action of diffeomorphism (2.17). The answer for the braneworld effective
±
action, is given in terms of the invariant fields, Rµν ( )
(x ),ψ ± (x ) , by the following spacetime integral
of a 2 × 2 quadratic form,

1  2F( ) 1 K( ) − 6F( )
[±
S 4 g µν ,ψ ± =] ∫ d 4 x RTµν 2 2 R µν + RT
l2 2
R
16πG4  l 6
1  K( ) 
T
 1 
− 3 Ψ + R  2 2  Ψ + R  . (2.21)
 6  l  6 

13
Here G4 is an effective four-dimensional gravitational coupling constant,

G5
G4 = , (2.22)
l

(R µν
( )
, Ψ ) and RTµν , Ψ T are the two-dimensional columns

 R + ( x ) ψ + ( x )
R µν =  µν , Ψ= −  (2.23)
 Rµν ( x ) ψ (x )
−

and rows
[
+
RTµν = Rµν (x )Rµν− (x ) , ] [ ]
Ψ T = ψ + (x )ψ − (x ) , (2.24)

of two sets of curvature perturbations and radion fields, associated with two branes ( T denotes the
matrix and vector transposition).
The quadratic approximation for the action and its nonlocal formfactors obviously determines the
spectrum of excitations in the theory. Now we show that in the graviton sector this spectrum
corresponds to the tower of Kaluza-Klein modes well-known from a conventional Kaluza-Klein
setup. The graviton sector arises when one decomposes metric perturbations on both branes into
irreducible components – transverse-traceless tensor, vector and scalar parts,

,ν
±
hµν ±
= γ µν + ϕ ±η µν + f µ ,ν + fν ,µ , γ µν = η µν γ µν = 0 . (2.25)

On substituting this decomposition in the linearized curvatures of (2.21) one finds that the vector
parts do not contribute to the action, and the latter reduces to the sum of the graviton and scalar
sectors,
±
S 4 g µν [ ]
,ψ ± = S graviton γ µν
±
[ ]
+ S scalar ϕ ± ,ψ ± . (2.26) [ ]
The graviton part in entirely determined by the operator F( ) and reads

− F( ) γ + 
µν
1 4 1
[ ]
S graviton γ µν±
=
16πG4 ∫
+
[
d x γ µν γ µν 2  µν  ,
2 l γ − 
] (2.27)

while the scalar sector consists of the radion fields of Eq. (2.23) and the doublets of the trace (or
conformal) parts of the metric perturbations ϕ ± ,

ϕ + ( x )
Φ =  − , [ ]
ΦT = ϕ + ( x )ϕ − ( x ) . (2.28)
ϕ ( x )

Their action diagonalizes in terms of the conformal modes and the (redefined) radion modes
2Ψ − Φ ,

3  1  3 T K( ) (2Ψ − Φ ) .
[ ]
S scalar ϕ ± ,ψ ± = ∫ d 4 x − ϕ + ϕ + + 2 ϕ − ϕ −  − ∫ d 4 x(2Ψ − Φ )
l2
32πG4  a  16πG4
(2.29)

14
Now we consider in the low-energy approximation on the positive-tension brane the case of large
brane separation, when a << 1 and
l
l << 1 , >> 1 . (2.30)
a

This range of coordinate distances 1 / corresponds to the long-distance approximation on the

Σ + -brane. Now one should use the asymptotic expressions of large arguments of the Bessel
functions (Jν− , Yν− ) , ν = 1,2,

2a l π πν  2a l π πν 
Jν− ≅ cos − − , Yν− ≅ sin  − − , (2.31)
πl 1/ 2
 a 4 2  πl 1/ 2
 a 4 2 

and the small-argument expansions for (Jν+ , Yν+ ). Then, in the leading order the operator F( ) reads

 l2 l2 
 2 2 J 2− 
F( ) ≅  2 . (2.32)
l J− l 
 2J − − 1− 
 2 J2 a 

In contrast to the case of small brane separation, the short-distance corrections to this matrix
operator contain a nonlocal 2 ln -term. We present it for the F+ + ( ) -element,

F+ + (
2
)= l + l
2
( ) k ( ) + O[(l ) ],
2
2 3
(2.33) k2 ( ) =
1 4 Y− 
 ln 2 − 2C + π 2−  . (2.34)
2
2 2 4 l J2 

This is a manifestation of the well-known phenomenon of AdS/CFT-correspondence when typical

quantum field theoretical effects in four-dimensional theory can be generated from the classical
theory in the bulk.
If we take the usual viewpoint of the braneworld framework, that our visible world is one of the
branes embedded in a higher-dimensional bulk, then the fields living on other branes are not
directly observable. In this case the effective dynamics should be formulated in terms of fields on
the visible brane. In the two-brane Randall-Sundrum model this is equivalent to constructing the
reduced action – an action with on-shell reduction for the invisible fields in terms of the visible
ones. We perform the reduction of the action to the Σ + -fields separately in the graviton and scalar
−
sectors. In the graviton sector (2.27) the on-shell reduction – the exclusion of γ µν − perturbations in

+
terms of γ µν = γ µν – corresponds to the replacement of the original action by the new one,

1 + Fred ( ) µν
[ ] ±
S graviton γ µν red
[ ]
⇒ S graviton g µν =
16πG4 ∫ d 4 x g γ µν
2l 2
γ+ , (2.35)

with the original kernel F( ) going over to the new one-component kernel Fred ( ) according to the
following simple prescription

15
 1
F( ) ⇒ Fred ( ) = F+ + ( ) − F+ − ( ) F−+ ( ) . (2.36)
F−− ( )
It is useful to rewrite (2.35) back to the covariant form in terms of (linearized) Ricci curvatures on a
single visible brane,

1  F ( ) µν 1 Fred ( ) 
red
S graviton [ ]
g µν = ∫ d 4 x g  Rµν red
l2 2
R − R 2 2 R  . (2.37)
8πG4  3 l 

A similar reduction in the scalar sector implies omitting in the first integral of (2.29) the negative-
tension term and replacing the 2 × 2 quadratic form in the second integral by the quadratic form in
ψ + with the reduced operator

1 det K( )
K red ( ) = K + + ( ) − K + − ( ) K −+ ( ) = . (2.38)
K −− ( ) K −− ( )

Thence, we express the conformal mode in terms of the (linearized) Ricci scalar ϕ + = −(1 / 3 )R ,
and denote the radion by ψ + = ψ . Then the combination of the reduced scalar sector together with
the graviton part (2.37) yields the reduced action in its covariant form

1  2F  1  1  1 2F 
[
S red g µν ,ψ =] ∫ d 4 x g  Rµν 2 red2  R µν − g µν R  − R − 2 red2  R
16πG4  l  2  6  l 
 R  2K  R 
− 6l 2  ψ +  2 red 2 
ψ +  . (2.39)
 6 l  6 

l
In the regime of small or finite brane separation l << 1 , << 1 , the calculation of the reduced
a
operator (2.36) gives the following result,

) ( )
2

Fred (
2

2
2
) = l 1 − a 2 + l κ1 (a ) + O l 2
( 2
[( ) ], 3
(2.40) κ1 (a ) =
1 1

4 a
( 1
2
) ( 2


)
ln 2 − 1 − a 2 − 1 − a 2  .

(2.41)
Similarly, in view of det K( ) =
4 ln a 2
1 − a4
(l )2
[ ) ],
+ O (l 2
3
the reduced operator in the radion sector
(2.38) is at least quadratic in ,

( )
K red ( ) = κ 2 (a ) l 2
2
+ O l2 [( ) ], 3
(2.42)
1 1
κ 2 (a ) = ln 2 , (2.43)
4 a

so that the low-energy radion turns out to be a dipole ghost.

The reduced braneworld action in the low-energy regime of finite interbrane distance is:

1  a2 1 
2
R  l2 
[ ]
S red g µν ,ψ = ∫ d 4
x g  (
1 − a 2
)R − R R − 6l 2
κ (a ) ψ +  + κ1 (a )C µναβ
2
. (2.44)
16πG4  6  6 2 

16
Furthermore, the following reparametrization from ψ to the new field ϕ ,

3   1  κ 2 (a )  1 
ϕ= a1 − R  − l  ψ + R  (2.45)
4πG4   6  −  6 

converts the action (2.44) to the local form

 1 1  1 l2 
[ ]
S red g µν , ϕ = ∫ d 4 x g  − ϕ 2  R + ϕ ϕ + κ1 (a )C µναβ
2
. (2.46)
 16πG4 12  2 32πG4 

The field ϕ introduced here by the formal transformation (2.45) directly arose as a local
redefinition of the radion field relating the Randall-Sundrum coordinates to the Gaussian normal
coordinates associated with the positive tension brane.

Large interbrane distance and Hartle boundary conditions

In the limit a → 0 the nonlocal and correspondingly non-minimal terms of (2.44) and (2.46) vanish
and the low-energy model seems to reproduce the Einstein theory. However, this limit corresponds
to another energy regime (2.30) in which one should use the expressions (2.32) – (2.34) in order to
obtain the reduced operator (2.36). Then the latter, up to quadratic in terms inclusive, reads as

Fred (
2
)= l + l
2
( ) 2
 4  l
 ln 2 − 2C  +
2
( )
 π Y2−
 +
2
a 
. (2.47)
2 8  l  2  4 J 2− 2l − − 
J1 J 2 

As in (2.33) it involves the logarithmic nonlocality (2.34) in 2 -terms. Moreover, the last term here
simplifies to the ratio of the first order Bessel functions Y1− / J1− , so that Fred ( ) takes a form very
similar to that of the large interbrane separation (2.33),

Fred ( ) =
l2
+
( )
l2
2
 4
ln − 2C + π
Y1− 
+ O l2 [( ) ].
3
(2.48)
 2 −
2 8  l J 1 

The calculation of the radion operator (2.38) with K( ) following from (2.32) for l 2 / a 2 >> 1
results in
( ) k ( ),
K red ( ) = l 2
2
2 (2.49)

where k 2 ( ) is defined by (2.34). Thus, Fred ( ) and K red ( ) are given by the following two
nonlocal operators,
1 4 Y−
kν ( ) = ln 2 − 2C + π ν−  , ν = 1,2, (2.50)
4 l Jν 

and the reduced (one-brane) action finally reads

17
 1  l2  R  R 
[ ]
S red g µν ,ψ = ∫ d 4
x g  R + Cµναβ k1 ( )C µναβ − 6l 2  ψ + k 2 ( ) ψ +  . (2.51)
16πG4  2  6  6 

Here terms quadratic in curvature represent short distance corrections with form factors whose
logarithmic parts have an interpretation in terms of the AdS/CFT-correspondence. With the usual
Wick rotation prescription → + iε these ratios tend to

Yν− l π πν 
≅ tan + iε − − →i, a → 0, (2.52)
Jν− a 4 2 

and both form factors (2.50) for < 0 (Euclidean or spacelike momenta) become real and can be
expressed in terms of one Euclidean form factor as

1 4  1 1 4
kν ( + iε ) a→0 = k ( ) ≡  ln 2 − C  , (2.53) and hence k ( ) + C ≡ ln 2 . (2.54)
4  l (− )  4 4 l (− )
This Wick rotation after moving the second brane to the AdS horizon impose a special choice of
vacuum or special boundary conditions at the AdS horizon. The Hartle boundary conditions
corresponding to this type of analytic continuation imply that the basis function u − ( z ) is given by
the Hankel function, u − ( z ) = H 2(1) z = J2 z ( )
+ iY2 z ( ) ( )
, and thus corresponds to ingoing
waves at the horizon. This is equivalent to the replacement Y1 , Y2− → 1, J 1− , J 2− → −i , in (2.50) and,
−

thus, justifies the Wick rotation of the above type. Hartle boundary conditions and the Euclidean
form factor (2.53) naturally arise when the Lorentzian AdS spacetime is viewed as the analytic
continuation from the Euclidean AdS (EAdS) via Wick rotation in the complex plane of time.

3. Compactifications of Type IIB Strings on a Calabi-Yau three-fold and Hartle-Hawking

wave-function in the mini-superspace sector of physical superstring theory.

A. Compactifications of Type IIB Strings on a Calabi-Yau three-fold. [3]

Although the self-duality of the five-form field strength in type IIB string theory implies that the
latter cannot be described by a supersymmetric 10-dimensional action, the bosonic fields can be
described by a non-self-dual action in which the equation of motion for the five-form field strength
is replaced by its Bianchi identity. This is consistent with self-duality, but does not imply it. When
self-duality is imposed as compactification condition, the non-self-dual action yields the correct
compactified theory. In the Einstein frame, the action is,

1 1
S = ∫ d 10 xˆ − gˆ  Rˆ − Tr ∂ µˆ Μ ( ˆ ∂ µˆ Μ ) ˆ −1 + 3 Η
ˆ Tˆ ˆ ˆ Μ
µνρ
ˆΗˆ µˆνˆρˆ + 5 Fˆ ˆ2ˆ ˆ ˆ ˆ +
µνρστ
2 8 8 12
1 ij µˆνˆρˆσˆτˆλˆαˆβˆγˆδˆ 
+ ε εˆ Dµˆνˆρˆσˆ Hˆ τ(ˆλiˆ)αˆ Hˆ β(ˆγjˆδ)ˆ  (3.1)
192 

The field definitions are

18
 1  λˆ 2 − Re λˆ 
Μˆ =  ; λˆ = lˆ + ie −ϕˆ , (3.2a)
ˆ 
Im λ − Re λˆ 
 1 
 ˆ (1) 
Ηˆ µˆνˆρˆ =  H µˆνˆρˆ  ; Hˆ µ(ˆiν)ˆρˆ = ∂ [µˆ Bˆν(ˆρiˆ)] , (3.2b)
 H (ˆ2ˆ)ˆ 
 µνρ 
3
Fˆµˆνˆρˆσˆτˆ = ∂ [µˆ Dˆνˆρˆσˆτˆ ] + ε ij Bˆ [(µiˆ)νˆ ∂ ρˆ Bˆσ(ˆτjˆ)] . (3.2c)
4

Also, εˆ0....9 = − g . It is in the final term of equation (3.1) that D = 4 vectors (from D̂ ) interact
with the D = 4 scalars from the three-form field strengths. It is convenient to rewrite this (up to an
overall constant) as
ε ij ∫ Fˆ ∧ Hˆ (i ) ∧ Bˆ ( j ) . (3.3)

To compactify to four dimensions use

Fˆ = F Λ ∧ α Λ − GΛ ∧ β Λ + ... , (3.4)

where (α Λ , β Λ ), Λ = {0,1,..., h21} are some choice of symplectic basis for H 3 (CY ) , FµνΛ are the 4-
dimensional vector field strengths and GΛµν are the magnetic field strengths and the dual
relationship between FµνΛ and GΛµν is due to self-duality of F̂ ; the terms that have been left out of
equation (3.4) are those which will not contribute to the integral in equation (3.3). The 3-form field
strengths are given Calabi-Yau expectation values via

< Hˆ (1) >= ν e(1Λ) β Λ − ν m(1)Λα Λ , (3.5a) < Hˆ (2 ) >= ν e(Λ2 ) β Λ − ν m(2 )Λα Λ , (3.5b)

where the ν m (e ) are constants that have been prematurely identified as values of the magnetic
(electric) charges. Using equations (3.5), integration of equation (3.3) over the Calabi-Yau gives

ε ij ∫ (ν e(Λi ) F Λ ∧ B ( j ) − ν m(i )Λ GΛ ∧ B ( j ) ) . (3.6)

~
Writing F Λ and GΛ in terms of electric and magnetic vector potentials AµΛ and AΛµ , gives, after an
integration by parts, (again up to a constant)

( )
ε ij ∫ d 4 x − g ν e(Λi ) AµΛ H ( j )µ − ν m(i )Λ Aˆ Λµ H ( j )µ , (3.7)

where
1
H (i )µ = ε νστµ ∂ν Bστ(i ) . (3.8)
6

The result is that to lowest order in the coupling constant, with, for simplicity, the fields
corresponding to the h21 data set to zero,

19
 ~
5~ 1 K 3K
(1) 2 K− K (2 ) 2 2 4+ 4
Hµ = e 4 4 ∂ µ Im S , (3.9a) Hµ = e ∂ µ Im C0 ; (3.9b)
3 3

also, the string coupling constant is

~
K K
−
ϕˆ
e = 2e 2 2
. (3.9c)

Here S and C0 are the N = 1 superfields which form the dilaton hypermultiplet; the four
dimensional dilaton has been generalized to
1 ~
φ = e − K . (3.10)
2
~
3K K
+
Substituting equations (3.9) into equation (3.7) gives, after a Weyl rescaling g µν → 2e 4 4
g µν (to
go to the D = 4 Einstein metric),

2 2
3 ∫
d 4x − g { 2e ~
K +K
~ ~
ν e(1Λ) AµΛ ∂ µ Im C0 − 2e K + Kν m(1)Λ AµΛ ∂ µ Im C0 −
~ ~ ~
e 2 Kν e(Λ2 ) AµΛ ∂ µ Im S + e 2 Kν m(2 )Λ AΛµ ∂ µ Im S . } (3.11)

This can be recognized as the interaction terms of the vector potentials with charged fields e S and
e C0 . Hence, completing the square with the kinetic terms for the hypermultiplets gives (with an
appropriate numerical rescaling of ν e(i()m )Λ )

~
{ ( ~ 2 ~
) ( ~ 2
S = ∫ d 4 x − g 8e K + K ν e(1Λ) AµΛ − ν m(1)Λ AΛµ + ∂ µ Im C0 + 2e 2 K ν e(Λ2 ) AµΛ − ν m(2 )Λ AΛµ + ∂ µ Im S + ... . ) }
(3.12)

From this equation, it is seen that Im C0 carries electric (magnetic) charges ν e(1Λ) (ν m(1)Λ ) and that
Im S carries electric (magnetic) charges ν e(Λ2 ) ν m(2 )Λ . ( )
With regard the compactification of IIB on a Calabi-Yau, now, the attention is restricted to an
h11 = 1 , h21 = 0 Calabi-Yau. The (uncomplexified) moduli space therefore is one-dimensional, and
corresponds to the choice of metric; specifically a conformal factor eσ . Furthermore, as RR fields
are suppressed in string perturbation theory, and because only the structure of the dilaton multiplet
is of interest, it will be convenient to take

lˆ = 0 ; Bˆ ij = 0 ; Dµˆνˆσˆτˆ = 0 . (3.13)

The equations of motion are usually written in terms of the fields

1 + iλˆ 1 − e −ϕˆ ∂ µˆψˆ

ψˆ = = −ϕˆ
, (3.14a) Pˆµˆ = , (3.14b)
1 − iλˆ 1 + e 1 − ψˆ ∗ψˆ

Qˆ µˆ =
(
Im ψˆ∂ˆ µˆψˆ ∗
,
) (3.14c)
Hˆ ˆ ˆ ˆ − ψˆHˆ µ∗ˆνˆρˆ
Gˆ µˆνˆρˆ = µνρ ; Hˆ µˆνˆρˆ = Hˆ µ(ˆ1ν)ˆρˆ + iHˆ µ(ˆ2νˆ)ρˆ . (3.14d)
1 − ψˆ ∗ψˆ 1
( ∗
1 − ψˆ ψˆ 2 )

20
The equation of motion that will be most interesting is

(∇ µˆ )
− iQˆ µˆ Gˆνµˆρˆˆ = Pˆµˆ Gˆν∗ˆρˆµˆ . (3.15)

Equation (3.15) is satisfied trivially on the Calabi-Yau. After performing a 4-dimensional Weyl
rescaling g µν → e −3σ g µν , equation (3.15) becomes (on the spacetime)

 ψ ∗∂ µψ − ψ∂ µψ ∗  3σ µ ∂ µψ
∇µ + e Gνρ = e 3σ Gνρ∗µ . (3.16)

 2 1 −ψ ψ∗
( 
 )
1 −ψ ψ∗

Subtracting ψ times the complex conjugate of equation (3.16), from equation (3.16), gives

 µ ∗µ 
e 3σ Gνρ − ψGνρ  ,
1
(1 −ψ ψ ) ∇
∗ 2
µ
 1

(3.17)
 1(− ψ ∗
ψ 2
 )
which is satisfied by introducing a complex scalar field D such that

Gµ − ψGµ∗
∂ µ D = e 3σ 1
, (3.18)
(1 −ψ ψ )∗ 2

where, as in equation (3.8)

Gµνρ = ε µνρ σ Gσ . (3.19)

The other equation of motion that is used is

9 σˆτˆ 3
Rˆ µˆνˆ = 2 Pˆ( µˆ Pˆνˆ∗) + Gˆ ( µˆ Gˆν∗ˆ )σˆτˆ − gˆ µˆνˆ Gˆ νˆσˆτˆGˆν∗ˆσˆτˆ . (3.20)
4 16

By substituting the Calabi-Yau part of this equation into the space-time part of the equation, the
four-dimensional action
1 9 2
S = ∫ d 4 x − g  R + Pµ + 3(∂ µσ ) + Gµ  (3.21)
2 2

2 4 

can be deduced. Alternatively, this can be found, almost by inspection, via dimensional reduction of
the NSD action of equation (3.1). There is a space-time dependent conformal factor of eσ in the
Calabi-Yau metric; hence − gˆ = e 3σ − g and so to remain in the Einstein frame required the
Weyl rescaling of the four dimensional metric g µν → e −3σ g µν . This is the same Weyl rescaling
used in the derivation of equation (3.17) and the reason for it. To obtain the standard quaternionic
geometry, we make the field redefinitions,

1
σ+ ϕ 3 2
Z = −ie 2
; (3.22a) C0 = i Im D ; (3.22b)
4
1
3σ − ϕ ~
φ =e 2
; (3.22c) φ = Re D ; (3.22d) and S = φ + iφ . (3.22e)

21
Define also

 i  ~ i
K = − ln − (Z − Z ) ; (3.23a) K = − ln[S + S ] ; (3.23b) Ν 00 = R00 = (Z − Z )3 ; (3.23c)
 8  32
Dµ C0 = ∂ µ C0 ; (3.23d); and Dµ S = ∂ µ S . (3.23e)

Then, the scalar part of the action of equation (3.21) becomes

{ ~ ~
S = ∫ d 4 x − g K ZZ ∂ µ Z∂ µ Z + K SS Dµ SD µ S + K SC0 Dµ SD µ C0 +
~ ~
K C0 S Dµ C0 D µ S + K C0C0 Dµ C0 D µ C0 , (3.24) }
~
where the subscripts on K , K denote differentiation. We note that C0 is pure imaginary.

B. Hartle-Hawking Wave-function in the mini-superspace sector of physical

superstring theory. [4]

Suppose we consider the compactification of type IIB superstrings to two dimensions on a Calabi-
Yau threefold M times a 2-sphere M × S 2 . Turn on 5-form fluxes for the RR2 5-form field strength
to be
F5 = F3 ∧ ω , (3.25)

where ω is a unit volume form on S 2 , and F3 is a 3-form on M . Choosing an integral basis of

{
magnetic/electric H 3 (M ) as α I , β J } I =0,...h 2 ,1
, we write

(
F3 = ∑ p I α I + q I β I . ) (3.26)
I

It is possible to write down a superpotential whose extremization leads to the condition for (2, 2)
supersymmetry in d = 2:
W = ∫ 2 F5 ∧ Ω , (3.27)
M ×S

hence
W =∫
M ×S 2
∑ (p α
I
I
I )
+ qI β I ∧ ω ∧ Ω , (3.27b)

where Ω is the holomorphic 3-form on the Calabi-Yau three-fold. To deduce this superpotential,
we note that this superpotential is consistent with the tension (BPS mass in 1 dimension) of the
domain wall D3 brane which wraps a 3-cycle in the Calabi-Yau and changes the F5 flux.
The condition for extremization of W and preserving supersymmetry is

2
We remember that the RR (Ramond-Ramond) states in type I and type II superstring theories, are the bosonic closed
string states whose left- and right-moving parts are fermionic. These include p-form potentials C p , with p taking all
odd values in the IIA string and all even values in the IIB string.

22
 DW = 0 . (3.28)

The complex structure of the Calabi-Yau are field variables and we can look for extrema of W with
respect to their variation. We denote variation of Ω in arbitrary direction of H 2,1 ( X ) by δΩ . The
supersymmetry condition (3.28) is equivalent to

∫ M ×S 2
δΩ ∧ F5 = ∫ δΩ ∧ F3 = 0 . (3.28b)
M

Since F3 is also real, this implies that F3 ∈ H 3,0 + H 0,3 . Using the fact that there is only one element
in H 3, 0 represented by Ω , and using the reality condition for F3 we deduce that

F3 = Re(CΩ ) ,

for some complex number C . In other words

(
p I = Re CX I , ) qI = Re(CFI ) , (3.29)

where
X I = ∫ Ω, FI = ∫ I Ω,
AI B

hence
W =∫
M ×S 2
∑ [Re(C ∫
I
AI
) ( ) ]
Ω α I + Re C ∫ I Ω β I ∧ ω ∧ Ω ,
B
(3.29b)

and (AI , B I ) are 3-cycles on M that are dual to the 3-forms (α I , β I ). With the complex structure of
the Calabi-Yau satisfying (3.29), supersymmetry is preserved with a suitable choice of metric in 2
spacetime dimensions, i.e. the AdS2 metric. For such a complex structure, the superpotential W is
not zero, but it is proportional to ∫ Ω ∧ Ω . In addition, the size of S 2 is also determined by the
M
supersymmetry condition, and we find that

( )
Area S 2 = πCC ∫ Ω ∧ Ω . (3.30)
M

This is exactly the content of the attractor mechanism. To case it in the standard description of the
black hole attractor, we consider D3 branes wrapping q I times on AI and p I times on B I . This
gives rise to a supersymmetric black hole in four dimensions, whose BPS3 mass M BPS is given by

2
2
M BPS = K −1 W , (3.31)

where the exponentiated Kahler potential K is given by

3
We remember that the BPS (Bogomolny-Prasad-Sommerfield) state is a state that is invariant under a non-trivial
subalgebra of the full supersymmetry algebra. Such states always carry conserved charges, and the supersymmetry
algebra determines the mass of the state exactly in terms of its charges.

23
 ( )
K = i X I FI − X I FI = −2 Imτ IJ X I X J , (3.32)

and W is the superpotential (3.27), which can also be expressed as

W = q I ∫ Ω − p I ∫ I Ω = q I X I − p I FI . (3.33)
AI B

Furthermore, the variation of the action

π
S =− (K (X , X ) + 2iW ( X ) − 2iW (X )) = − π (∫ Ω ∧ Ω + ∫ (Ω + Ω ) ∧ F ), 3 (3.34)
4 4

in arbitrary directions of H 3 (M ) , reproduces the attractor equations (3.29).

Consider type IIB superstring compactified on the Calabi-Yau 3-fold M times S 2 × S 1 . There is a
natural Euclidean solution to the classical equations of motion which develops from this spacelike
section. It is the geometry M × S 2 × H 2 / Z , where H 2 is the hyperbolic disk, i.e. the Euclideanized
AdS2, with the metric
ds 2 = dρ 2 + e 2 ρ dτ 2 , (3.35)

and the Z quotient periodically identifies τ ≈ τ + β . If we view ρ as an Euclideanized time, the

geometry M × S 2 × H 2 / Z describes an Euclidean time evolution of type II string compactified on
M × S 2 × S 1 . This is how we were originally led to the metric (3.35).
Let us consider a natural notion of a “mini-superspace” where we view ρ as an Euclideanized time.
Among relevant light modes are the complex moduli of the Calabi-Yau three-fold denoted by
z i (i = 1,..., h 2,1 ) , which are in vector multiplets in four dimensions. The gravity multiplet also
produces some scalar fields upon compactification on S 2 × S 1 . One is the radius R of S 2 . Another
scalar field is related to how the S 1 is fibered over S 2 . The radius R and the chemical potential ϕ
naturally combine with the complex structure moduli z of the Calabi-Yau to make a “large moduli
space” with coordinates X I (I = 0,1,..., h 2,1 ) . More explicitly, choose any holomorphic section
X 0I ( z ) over the complex structure moduli space and define

1/ 2
 W (X 0 ) 
X = i 2 Re 
I iϕ
 X 0I . (3.36)
 K ( X 0 , X 0 )W ( X 0 ) 

Upon compactification on S 1 , each gauge field becomes equivalent to a pair of massless scalar
fields – one is the Wilson line of the gauge field along the S 1 , and the other is the dual magnetic
potential around the S 1 ,
~ ~
φ I = ∫ 1 A I , φ I = ∫ 1 AI , (3.37)
S S

~ ~
where AI is the dual of A I in four dimensions. One can also think of φI as the dual of the massless
gauge field in three dimensions. By definition, they couple to the charges ( p I , q I ) of the black hole
as
(q φ I + p Iφ~I )
e ∑I I
i
. (3.38)

24
 ( ~
)
Combining with X I defined in (3.36), a set of four scalar fields X I , X I , φ I , φI for each I gives
bosonic components of supermultiplets. The dependence of the wave-function on φ I , φI is also
~
( )
~ ~
simple. Since φ I is dual to AI and φI is dual to A I in three dimensions, we have

~
dφ J ~ dφ J
GIJ = ∫ 2 FI = q I , G IJ
= ∫ 2 F I = pI , (3.39)
dρ S dρ S

~ ~
where GIJ is the metric in the kinetic term for the gauge fields, and F I = dA I , FI = dAI . These
~
equations means that, when we quantize the theory along the ρ direction, (φ I , φI ) are canonically
~
conjugate to (qI , p I ) . Therefore, the wave-function depends on φ , φ as ( )
(q φ I + p Iφ~I )
Φ p ,q X , X ; φ , φ = e ∑ I I
( )
~
Ψ p ,q ( X , X ) , (3.40)
i

~
( )
if it is an eigenstate of the fluxe quantum numbers ( p, q ) . This φ , φ dependence is also expected
( ) ~
from the fact that φ , φ are electric and magnetic static potentials for the black hole charges (3.38).
Now we consider the original Hartle-Hawking wave-function for a three-sphere S 3 . The saddle
point computation of the wave-function on S 3 can be viewed as filling it with a 4-dimensional ball
with the S 3 as its boundary, and this leads to the action S E in the Euclidean ball,

1
SE ≈ − ,
Λ

where Λ is the cosmological constant, and the mini-superspace wave-function behaves as

 1
Ψ ≈ exp −  .
 Λ

1
In the present context Λ < 0 , and −
Λ
( )
≈ Area S 2 ≈ Sentropy , so we may expect that

∫ dXdX Ψp,q (X , X ) ≈ exp(Sentropy ) , (3.41)

2

namely the wave-function is normalized by the exponential of the entropy. It is natural since the
string partition function on the full space M × S 2 × H 2 / Z should give the black hole entropy. Thus,
at least semi-classically we expect (3.41) to hold. In view of our discussion following (2.11) a
natural guess for the probability measure is

 π π 
ψ p ,q ( X , X ) ≈ exp − K − i (W − W ) ,
2
(3.42)
 4 2 

hence
 π π 
∫ dXdX exp− 4 K − 2 i(W − W ) ≈ exp(S entropy ), (3.41b)

25
where W ( X ) depends on the flux determined by ( p, q ) as in (3.33). Indeed, the right-hand side is
peaked at the attractor value and its value is given by the exponential of the entropy. To see this in
terms of the physical variables, the complex structure moduli z i and the radius R of the S 2 , we
can substitute (3.36) into (3.42) and find

[
Ψ p ,q ≈ exp − π (R 2 − M BPS R ) .
2
] (3.43)

Extremizing this with respect to the complex structure moduli z i gives the attractor equation and
extremizing with respect to R gives

π 2 
 = exp(S entropy ) ,
2
Ψ p ,q extremum ≈ exp M BPS
4 

reproducing the expected result. The eq. (3.42) captures essential aspects of the wave-function.
This same wave function will give a semi-classical approximation to the topological string partition
function, which we will argue gives the exact answer for the Hartle-Hawking wave function
including all string loop corrections.
The wave-function Ψ p ,q ( X , X ) should satisfy the Wheeler-De Witt equation. In mini-superspace
the WDW equation corresponds to the quantization of the attractor flow for a black hole with
charges p I and qI . Consider a ten-dimensional Euclidean metric of the form,

ds 2 = e 2U + 2 ρ dτ 2 + e −2U dρ 2 + e −2U dΩ 2 + dsCY

2
, (3.44)

where τ is the Euclideanized time direction compactified on S 1 , ρ is the radius coordinate, dΩ 2

2
is the metric on a two-sphere of unit radius, and dsCY is the metric on the internal Calabi-Yau three-
fold. Note that e −U is the radius of the S 2 , and the AdS 2 geometry is realized when U is constant.
Since we are interested in BPS configurations and since the supercharges preserved by the
background square to become the translation along the τ direction, we assume that the scale factor
eU and the complex moduli z i (i = 1,..., h 2,1 ) of the Calabi-Yau three-fold are independent of τ . In
this case, we have a one-dimensional system along the ρ direction described by the effective action,

1 ∞ 
ρ  dU 
2
dz i dz j 
S eff = ∫ dρe  + 1 + g ij + e 2U M BPS
2
( ( )
z , z ) + 4 g ij ∂ i M BPS ∂ j M BPS  , (3.45)
2 −∞  dρ  dρ dρ 

where
W ( z )W ( z )
M BPS = . (3.46)
K ( z, z )

We regard ρ as the Euclidean time of the system, which flows from ρ = +∞ to − ∞ . Since the
effective action (3.45) can be written as

1 ∞  dU   dz i
2

S eff = ∫ dρe ρ  + 1 − eU M BPS  + g ij  − 2eU g im ∂ m M BPS 
2 −∞  dρ   dρ 

26
  dz j 
 − 2 e U
g jn
∂ M  + (total derivative), (3.46b)
 dρ n BPS 
 

the BPS equations are

dU dz i
= −1 + eU M BPS ( z , z ) , = 2eU g ij ∂ j M BPS ( z, z ) . (3.47)
dρ dρ

The signs on the right-hand side of these equations are chosen so that they are compatible with the
initial condition at ρ → ∞ , which we regard as the infinite past in the Euclidean time.
The equations (3.47) can be combined into a single equation on the large moduli space. To write
down such an equation, we start with a holomorphic section X 0I ( z )(I = 0,1,..., h 2,1 ) over the moduli
space of complex structure. They make projective coordinates of the moduli space, and as such
there is a freedom to rescale these coordinates. We define the exponentiated Kahler potential K 0
and the superpotential W0 for these coordinates as

K 0 = −2 Im τ IJ X 0I X 0J , W0 = q I X 0I − p I FI ( X 0 ) .

We then combine the scale factor eU in the metric (3.44) and the complex moduli z i into a single
set of coordinates X I defined by

1/ 2
 W 
X I = 2ie −U  0  X 0I . (3.48)
 K 0W0 

Note that the right-hand side of (3.48) is invariant under rescaling of X 0I . Moreover

K (X , X ) = −2 Im τ IJ X I X J = 2e −U ( )2

is the diameter squared of the S 2 . Thus, the large moduli space parametrized by X I combines the
complex moduli z i and the radius of the S 2 .
Using X I , the attractor flow equations (3.47) can be written as a single equation

IJ
dX I  i 
= XI +  ∂ JW (X ) . (3.49)
dρ  Imτ 

Let us show that (3.49) is equivalent to (3.47). If we multiply X J Imτ JI to both sides of (3.49), the
left-hand side becomes

dX J dU e −2U  dz i dz i 
X I Imτ IJ = 2e −2U −  ∂ i M BPS − ∂ i M BPS  .
dρ dρ M BPS  dρ dρ 

On the other hand, the right hand-side becomes

27
 1
X I Imτ IJ X J + iX I ∂ I W (X 0 ) = − K ( X , X ) + iW (X ) = −2e −2U + 2e −U M BPS .
2

Combining them together, we obtain

dU 1  dz i dz i 
−  ∂ i M BPS − ∂ i M BPS  = −1 + eU M BPS .
dρ 2M BPS  dρ dρ 

The real part of this equation is precisely the first of (3.47):

dU
= −1 + eU M BPS . (3.50)
dρ

The imaginary part gives

dz i dz i
∂ i M BPS = ∂ M BPS . (3.51)
dρ dρ i

Similarly multiplying ∂ i X 0J Imτ IJ to both sides of (3.49) and using (3.50) and (3.51), we find

dz i
= 2eU g ij ∂ j M BPS . (3.52)
dρ

Moreover, (3.52) implies (3.51) since both sides of (3.51) are now equal to − 2eU g ij ∂ i M BPS ∂ j M BPS .
Therefore, (3.49) for X I defined by (3.48) is equivalent to the standard BPS equations (3.47).
In (3.49), a general BPS solution can be easily expressed. Taking the real and imaginary parts of
this equation, one finds

 dX I   dF 
Re X I −  = p I , Re FI − I  = qI .
 dρ   dρ 

A general solution to this is then

Re X I = p I + c I e ρ , Re FI = qI + d I e ρ ,

where (c I , d I ) are integration constants specified by the initial condition at the infinite past ρ = ∞ .
Whatever initial condition one chooses there, X I at the infinite future ρ → −∞ are fixed to be at
the attractor value,
Re(X I ) → p I , Re(FI ) → q I .

It is useful to write the BPS equation (3.49) as

dX I 1
Im(τ IJ ) = − ∂ I K + i∂ IW . (3.53)
dρ 2

28
Now, we will use this equation to obtain the supersymmetric version of the WDW equation.
The supersymmetric WDW equation is equivalent to the quantum version of the BPS equation
(3.49) and its complex conjugate. To understand the quantum version of the BPS equation, we note
that the metric in the X space implied by the effective action (3.45) is almost given by Im(τ IJ )
since
X I Im(τ IJ )X J = −2 R 2 , Di X I Im(τ IJ )X J = X I Im(τ IJ )D j X J = 0 , Di X I Im(τ IJ )D j X J = 2 R 2 g ij ,
(3.54)
where
Di X I = K∂ i (K −1 X I ) , (
D j X J = K∂ j K −1 X J , )
and g ij = ∂ i ∂ j ln K . We note that Imτ IJ has one negative sign in the direction of H (3, 0 ) whereas it
gives the standard positive definite metric in the H (2,1) direction. Flipping the sign in the H (3, 0 )
direction gives what is denoted by Im Ν IJ in the supergravity literature, which is the metric derived
from the effective action (3.45). In the semi-classical approximation, flipping of the sign of the
metric can be done by a suitable contour deformation in a functional integral. Thus, we will use
Imτ IJ as our metric and the corresponding quantization rule is

π dX J ∂
Im(τ IJ ) →− .
4 dρ ∂X J

Given this rule, the quantum version of the BPS equation (3.49) is

 ∂ π π 
 I
− ∂ I K + i ∂ I W  Ψ p ,q = 0 , (3.55)
 ∂X 8 4 

and similarly for the complex conjugate equation

 ∂ π π 
 I
− ∂ I K − i ∂ I W  Ψ p ,q = 0 . (3.56)
 ∂X 8 4 

We denote the operator appearing in the constraint (3.55) C I and the one in (3.56) by C I , so that in
terms of the state Ψ p ,q the constraints are C I Ψ p ,q = 0 and C I Ψ p ,q = 0 respectively. Imposing
both constraints is sufficient to determine the entire wave function. One finds in this way

π π 
Ψ p ,q ( X , X ) = exp  K (X , X ) + (iW ( X ) − iW ( X )) .
8 4 

One can use the BPS constraints also in a way that leads to a description of the covariant
supersymmetric gradient flow, also known as the attractor flow, in terms of a holomorphic wave
function ψ p ,q ( X ) . This wave function can be obtained by imposing first the constraint (3.55). This
reduces Ψ ( X , X ) essentially to a holomorphic function. The second condition (3.56) then fixes
ψ p ,q ( X ) to be given by
π
i Wp ,q ( X )
ψ p ,q ( X ) = e 2
. (3.57)

29
We note, in particular, that ψ 0,0 ( X ) = 1 . In the remaining part of this chapter we will provide
evidence that this wave function coincides with the semi-classical approximation obtained from the
topological string partition function, which we propose to be the exact Hartle-Hawking wave
function including all string loop corrections.
We abbreviate the operators appearing in the constraints (3.55) and (3.56) by C I and C I
respectively. We want to impose C I on the ket state Ψ and its conjugate C J| on the bra state Ψ .
Notice that C J| differs from C J in the sign of the derivative. The Dirac bracket is defined as

[X I
,XJ ]
Dirac [ ]
= X I , X J − ∑ X I , C K| [ ] [C 1, C ] [C , X ] ,
| L
J

K ,L K L

where the denominator should be read as the inverse matrix. Here we only wrote commutators that
we know are non-vanishing. For the constraints one finds

π π
[C K ]
, C L| = −
4
∂K ∂LK =
2
Im τ KL ,

while the commutators of the constraints with the coordinates give

[X I
]
, C K| = δ I K , [C , X ] = δ
L
J
L
J
.

Inserting this in to the definition for the Dirac bracket leads to the following commutation relations
for X I and X J
IJ
2 1 
[X I
,X J
]
Dirac =   . (3.58)
π  Imτ 

The Hilbert space for (3.58) can be represented by holomorphic wave-functions ψ X I of X I , with ( )
the inner product defined by
π
K
ψ 1 ψ 2 = ∫ dXdXe 4 ψ 1 ( X )ψ 2 ( X ) . (3.59)

The relation between the wave function in the real and complex polarization can be found at the
semi-classical level by applying standard canonical transformation techniques. In classical
mechanics canonical transformations can be described with the help of a generating function. In our
case this generating function should depend on one of the real and one of the complex coordinates.
Since we are interested in transforming wave functions ψ ( X ) to ψ (χ ) the appropriate choice is to
use a function S ( X , χ ) of the real coordinate χ I and the complex coordinate X I . It is determined
by requiring that the following canonical transformation

χ I = Re(X I ) , η I = Re(FI ) , (3.59b)

takes the form

1 ∂S ( X , χ ) 2  1  ∂S ( X , χ )
IJ
I
ηI = , X =   .
iπ ∂χ I iπ  Imτ  ∂X J

30
After a little algebra one finds

iπ  I 1 1 
S(X , χ ) =  χ FI ( X ) − F0 ( X ) + F0 (2 χ − X ) . (3.60)
2  2 2 

This leads to the equations

1 1
ηI = FI ( X ) + FI (2 χ − X ) , X I = 2χ I − X I ,
2 2

which are equivalent to (3.59b). The relation between the semi-classical wave function
ψ ( X ) = e g ( X ) and the corresponding wave function ψ (χ ) = e f ( χ ) in the real polarization is

i
S ( X ,χ )
ψ ( X ) = ∫ dχe π
ψ (χ ) , (3.61)

where it is understood that the right-hand side is computed in the saddle point approximation.
Now we want propose an exact such wave function which agrees in the semi-classical limit with the
wave function discussed before. We will argue that the state with no flux can be identified with the
topological string wave function:
ψ 0, 0 = ψ top .

That this relation could hold presupposes that topological string partition function also corresponds
to a wave function associated to quantizing H 3 (M ) . Moreover this is in agreement with the fact
that at least semi-classically the topological string partition function describes the Hartle-Hawking
wave function in the real polarization. We argue that the semi-classical result

ψ p ,q = O p ,q ψ 0 , 0

also holds with the expected factor

 iπ 
O p ,q = exp − W p ,q + c.c. (3.62)
 4 

interpreted as an operator acting on the topological string Hilbert space.

Semi-classically, the entropy of the BPS black hole obtained by wrapping D3 branes around cycles
of M is given by the area (3.30) of the horizon. The resulting quantum corrected entropy formula
can be concisely expressed as
S BH (q, p ) = F ( p, φ ) + ∑ qI φ I , (3.63)
I

where
i
F ( p, φ ) = Ftop ( X ) + Ftop (X ) , X = p+ φ . (3.64)
π

and
∞
Ftop ( X ) = ∑ Fg ( X )
g =0

31
is the full topological string partition function.
Moreover the quantum corrected attractor equations also take the simple form:

∂
qI = − F ( p, φ ) . (3.65)
∂φ I

At the attractor point, the string perturbation expansion is an asymptotic expansion for large black
hole charges. Since (3.63) takes the Legendre transformation from φ to q , the number of states
Ω( p, q ) of the black hole with finite charges p I , qI is given by Laplace transformation of the
topological string partition function

2
 i 
Ftop  p + φ 
− q I φ I + F ( p ,q )
Ω( p, q ) = ∫ dφe = ∫ dφe − qI φ I
e  π 
. (3.66)

More precisely, the conjecture states that Ω( p, q ) given by (3.66) is the Witten index for the
quantum Hilbert space of the black hole.
We note also that the expression (3.66) for the number of states can be written in a nice way as a
Wigner function. Namely, by taking the contour of the φ integral in (3.66) along the imaginary axis
as φ = −iπχ , one gets
Ω( p, q ) = ∫ dχe iπqχψ top ( p − χ )ψ top ( p + χ ) , (3.67)

where
Ftop ( χ )
ψ top (χ ) = e (3.68)

is the exact topological string partition function.

Now we can relate this result to the normalization of the Hartle-Hawking wave function.

Let ψ p,q denote the state we obtain upon doing the path-integral on the right in a fixed flux sector.
The above consideration leads to the statement that

Ω( p , q ) = ψ p , q ψ p , q .

Now let us write the expression (3.66) in invariant form

Ω( p, q ) = ψ top e iπ (qχ − pη ) ψ top ,

where χ and η are to be regarded as operators, and the state ψ is defined by the wave function
ψ ∗ (− χ ) . Here, the state on the left is not simply the complex conjugate but it contains a minus sign
due to the time reversal in the attractor flow equation. Next we note that the attractor relations
(3.59b) imply that q I χ I − p Iη I = Re W ( X ) , and hence

32
 −π
i (W ( X )−W ( X ))
Ω( p, q ) = ψ top e 2
ψ top .

It follows that if we identify

π
−i (W ( X )−W ( X ))
ψ p ,q = e 4
ψ top , (3.69)

or in the wave function form as

1  i 
− qI φ I + Ftop  p I + φ I 
ψ p ,q (φ ) = eI 2  π 
,

then
Ω( p, q ) = ψ p ,q ψ p ,q = ∫ dφ I ψ p ,q φ I ( ) 2
(3.69b)

exactly as expected. Moreover the form of the wave function (3.69) is exactly consistent with the
semi-classical reasoning which led to (3.62). The fact that the wave functions for both the IIA and
IIB side would lead to the same state is clear once we recall that the internal part of the Calabi-Yau
and thus the mini-superspace is identical for both cases where a D2 brane IIA instanton is playing
the role of D3 brane of IIB. We find this a highly non-trivial evidence for our conjecture for the
exact Hartle-Hawking wave function.
The relation between the wave function ψ top ( X ) and the exact topological string partition function
ψ top (χ )
i
S ( X ,χ )
ψ top ( X ) = ∫ dχe π ψ top (χ ) (3.70)

is a semi-classical formula. It describes the loop corrections to the Hartle-Hawking wave function to
all orders in perturbation theory.
The partition function of topological string theory can be computed perturbatively around a given
background by writing X I = Z I + x I , and treating the perturbation x I as coupling constants on the
worldsheet. The coordinates x I used by the perturbative topological string are a linearization of the
“curved” X I coordinates. The relation with (χ ,η ) is

χ I = Re(Z I + x I ) η I = Re(FI (Z ) + τ IJ x J ) , (3.71)

where τ IJ = ∂ I ∂ J F0 (Z ) is determined by the background. These are just the attractor equations
(3.59b) linearized around X I = Z I . In this way the topological string avoids the normal ordering
problems but at the cost of a background dependence. The partition function ψ top (x; Z , Z ) is related
to the background independent wave function ψ top (χ ) via the Bargman transform

1
iS ( x , χ ;Z , Z )
ψ top (x; Z , Z ) = det Im τ 2
∫ dχ e ψ top (χ ) , (3.72)

with

33
 π π
S (x , χ ; Z , Z ) = τ IJ (Z )χ I χ J + πχ I (x I + Z I ) + (x I
+ Z I )( xI + Z I ) ,
4 4

where indices are lowered with Imτ IJ (Z ) . This expression is a linearization of the generating
function S ( X , χ ) defined in (3.60). The topological partition function ψ (x; Z , Z ) thus gives in a
certain sense a linearized description of the Hartle-Hawking wave function ψ top ( X ) .

4. p-Adic Models in the Hartle-Hawking proposal. [5]

Ordinary and p-adic quantum mechanics can be unified in the form of adelic quantum mechanics

(L2 (Α), W (z ),U (t )) . (4.1)

L2 (Α ) is the Hilbert-space on Α , W ( z ) is a unitary representation of the Heisenberg-Weyl group

on L2 (Α ) and U (t ) is a unitary representation of the evolution operator on L2 (Α ) . The evolution
operator U (t ) is defined by

U (t )ψ ( x ) = ∫ Κ t ( x, y )ψ ( y )dy = ∏ ∫ Κ t(v ) ( xv , yv )ψ (v ) ( yv )dyv . (4.2)

Α Qv
v

The eigenvalue problem for U (t ) reads

U (t )ψ αβ ( x ) = χ (Eα t )ψ αβ ( x ) , (4.3)

where ψ αβ are adelic eigenfunctions, Eα = (E∞ , E2 ,..., E p ,...) is the corresponding adelic energy,
indices α and β denote energy levels and their degeneration. Any adelic eigenfunction has the
form

ΨS ( x ) = Ψ∞ ( x∞ )∏ Ψ p (x p )∏ Ω x p ( ),p x ∈ Α , (4.4)
p∈S p∉S

where Ψ∞ ∈ L2 (R ) , Ψ p ∈ L2 (Q p ) are ordinary and p-adic eigenfunctions, respectively. The Ω -

function, that is defined from the following formula

Ω x ( ) = 1,
p x p ≤ 1; Ω x ( ) = 0,
p x p >1, (4.4b)

is an element of the Hilbert space L2 (Q p ) , and provides convergence of the infinite product (4.4).
A suitable way to calculate p-adic propagator Κ p ( x' ' , t ' ' ; x' , t ') is to use Feynman’s path integral
method, i.e.

x '',t ''  1 t '' 

Κ p ( x ' ' , t ' ' ; x ' , t ') = ∫ χp− ∫ L(q& , q, t )dt  Dq . (4.5)
x ',t '
 h t'


For quadratic Lagrangians it has been evaluated in the same way for real and p-adic cases, and the
following exact general expression is obtained:

34
 1
 1 ∂2S  1 ∂2S 2  1 
Κ v ( x' ' , t ' ' ; x' , t ') = λv  −  χ v  − S ( x ' ' , t ' ' ; x ' , t ')  . (4.6)
 2 h ∂x ' ' ∂x '  h ∂x ' ' ∂x ' v  h 

With regard the Hartle-Hawking proposal for the wave function of the universe, the p-adic wave
function is given by the integral
( )
Ψ p qα = ∫ dNΚ p qα , N ;0,0 , (4.7)
Gp
( )

where, according to the adelic structure of N , G p = Z p (i.e. N p ≤ 1 ) for every or almost every p .

Models of the de Sitter type

Models of the de Sitter type are models with cosmological constant Λ and without matter fields.
We consider two minisuperspace models of this type, with D = 4 and D = 3 space-time dimensions.
The corresponding real Einstein-Hilbert action is

1 1
S=
16πG ∫M
d D x − g (R − 2Λ ) +
8πG ∫∂M
d D−1 x h K , (4.8)

where R is the scalar curvature of D-dimensional manifold M , Λ is the cosmological constant,

and K is the trace of the extrinsic curvature K ij on the boundary ∂M . The metric for this model is
of the Robertson-Walker type
[
ds 2 = σ D −2 − N 2 dt 2 + a 2 (t )dΩ 2D −1 . (4.9) ]
In this expression dΩ 2D−1 denotes the metric on the unit (D − 1) -sphere,
σ D −2
[
= 8πG / V (D − 1)(D − 2 ) , where V
D −1
]
D −1
is the volume of the unit (D − 1) -sphere.
In the real D = 3 case, the model is related to the multiple-sphere configuration and wormhole
solutions. v -adic classical action for this model is

1   2a ' ' a ' a '2 + a ' ' 2 

S v (a ' ' , N ; a ' ,0 ) =  N λ + λ  −  . (4.10)
2 λ   sinh N λ (
tanh N λ ) ( ) 


( )
Let us note that λ , λ = ΛG 2 , denotes the rescaled cosmological constant Λ . Using (4.6) for the
propagator of this model we have

1/ 2
 2 λ  λ
Κ v (a' ' , N ; a' ,0) = λv  −  χ v (− S v (a' ' , N ; a' ,0)). (4.11)
 sinh N λ ( ) (
 sinh N λ
 ) v

The p-adic Hartle-Hawking wave function is

Ψ p (a ) = ∫ dN
λ p (− 2 N )  N
χ p  − +
λ coth N λ 2 
a ,
( ) (4.12)
1/ 2
N p ≤1
N  2 2 
p

which after p-adic integration becomes

35
 Ψ p (a ) = Ω a p , ( ) λ p ≤ p −2 , p ≠ 2,
1
Ψ p (a ) =
2
( )
Ω a2 , λ 2 ≤ 2 −4 , p = 2. (4.13)

The de Sitter model in D = 4 space-time dimensions may be described by the metric

 N2 2  2G
ds 2 = σ 2  − dt + q(t )dΩ 32  , σ2 = , (4.14)
 q(t )  3π

2G  N 2 2 
i.e. ds 2 =  − dt + q(t )dΩ32  , (4.14b)
3π  q(t ) 

1 t ''  q& 2 
and the corresponding action S v [q ] = ∫ dtN  − 2
− λq + 1 , where λ = 2ΛG / (9π ) .
2 t'  4N 
 q ' '−q ' 
For N = 1 , the equation of motion q&& = 2λ has solution q (t ) = λt 2 +  − λT t + q ' , where
 T 
q ' ' = q (t ' ') , q ' = q (t ') and T = t ' '−t ' . Note that this classical solution resembles motion of a
particle in a constant field and defines an algebraic manifold. The choice of metric in the form
(4.14) yields quadratic v -adic classical action

T (q ' '−q ')

2
λ2T 3
S v (q ' ' , T ; q ' ,0 ) = − [λ (q '+ q ' ') − 2] − . (4.15)
24 4 8T

According to (4.6), the corresponding propagator is

λv (− 8T )
Κ v (q ' ' , T q ' ,0) = 1/ 2
χ v (− S v (q ' ' , T q ' ,0 )) . (4.16)
4T v

We obtain the p-adic Hartle-Hawking wave function by the integral

λ p (− 8T )  λ2T 3 T q2 
Ψ p (q ) = ∫ dT χ p  − + (λq − 2 ) + , (4.17)
T p ≤1
4T
1/ 2
 24 4 8T 
p

( )
and as a result we get also Ω q p function with the condition λ = 4 ⋅ 3 ⋅ l , l ∈ Z p . The above Ω -
functions allow adelic wave functions of the form (4.4) for both D = 3 and D = 4 cases. Since
λ p ≤ p −2 in (4.13) for all p ≠ 2 , it means that λ cannot be a rational number and consequently
the above the de Sitter minisuperspace model in D = 3 space-time dimensions is not adelic one.
However D = 4 case is adelic, because λ = 4 ⋅ 3 ⋅ l is a rational number when l ∈ Z ⊂ Z p .

5. p-Adic and Adelic wave functions of the Universe. [6]

In the Vladimirov-Volovich formulation, p-adic quantum mechanics is a triple

36
 (L (Q ), W (z ),U (t )),
2 p p p (5.1)

( )
where W p ( z ) corresponds to W p αxˆ , β kˆ defined in the following equation

1 
( )
Wv αxˆ, βkˆ = χ v  αβ  χ v − βkˆ χ v (− αxˆ ) .
2 
( ) (5.1a)

Adelic quantum mechanics is a natural generalization of the above formulation of ordinary and p-
adic quantum mechanics: (L2 ( A), WA ( z ), U A (t )) . In complex-valued adelic analysis it is worth
mentioning an additive character
χ A ( x ) = χ ∞ ( x∞ )∏ χ p (x p ) , (5.2)
p

a multiplicative character
s s s
x A = x∞ ∞ ∏ x p , s ∈C , (5.3)
p
p

and elementary functions of the form

ϕ P ( x ) = ϕ ∞ ( x∞ )∏ ϕ p (x p )∏ Ω x p
p∈P p∉P
( ), p
(5.4)

where ϕ ∞ ( x∞ ) is an infinitely differentiable function on R and x∞ ∞ ϕ ∞ ( x∞ ) → 0 as x∞ ∞ → ∞ for

n

any n ∈ {0,1,2,...} , ϕ p (x p ) are some locally constant functions with compact support, and

( ) = 1,
Ω xp p
xp p
≤ 1, ( ) = 0,
Ω xp p
xp p
> 1. (5.5)

All finite linear combinations of elementary functions (5.4) make the set L( A) of the Schwartz-
Bruhat adelic functions. The Fourier transform of ϕ ( x ) ∈ L( A) , which maps L( A) onto L( A) , is

ϕ~ ( y ) = ∫ ϕ ( x )χ A ( xy )dx , (5.6)
A

where χ A (xy ) is defined by (5.2) and dx = dx∞ dx2 dx3 ... is the Haar measure on A . A basis of
L2 ( A(P )) may be given by the corresponding orthonormal eigenfunctions in a spectral problem of
the evolution operator U A (t ) , where t ∈ A . Such eigenfunctions have the form

ψ P ( x, t ) = ψ ∞ ( x∞ , t∞ )∏ψ p (x p , t p )∏ Ω x p
p∈P p∉P
( ), p
(5.7)

where ψ ∞ ∈ L2 (R ) and ψ p ∈ L2 (Q p ) are eigenfunctions in ordinary and p-adic cases, respectively.

( ) is an element of L (Q ) , defined by (5.5), which is invariant under transformation of an
Ω xp p 2 p

evolution operator U p (t p ) and provides convergence of the infinite product (5.7).

p-Adic and adelic minisuperspace quantum cosmology is an application of p-adic and adelic
quantum mechanics to the cosmological models, respectively. In the path integral approach to

37
standard quantum cosmology, the starting point is Feynman’s path integral method. The amplitude
to go from one state with intrinsic metric h'ij and matter configuration φ ' on an initial hypersurface
Σ' to another state with metric h' 'ij and matter configuration φ ' ' on a final hypersurface Σ' ' is
given by the path integral

[ ]
K∞ (h' 'ij , φ ' ' , Σ' ' ; h'ij , φ ' , Σ') = ∫ χ ∞ (− S ∞ g µν , Φ )D∞ g µν D∞ Φ (5.8)

over all four-geometries g µν and matter configurations Φ , which interpolate between the initial
and final configurations. In (5.8) S ∞ [g µν , Φ ] is an Einstein-Hilbert action for the gravitational and
matter fields. To perform p-adic and adelic generalization we make first p-adic counterpart of the
action using form-invariance under change of real to the p-adic number fields. Then we generalize
(5.8) and introduce p-adic complex-valued cosmological amplitude

[ ]
Kp (h' 'ij , φ ' ' , Σ' ' ; h'ij , φ ' , Σ') = ∫ χ p (− S p g µν , Φ )Dp g µν Dp Φ . (5.9)

The standard minisuperspace ground-state wave function in the Hartle-Hawking (no-boundary)

proposal is defined by functional integration in the Euclidean version of

ψ ∞ [hij ] = ∫ χ ∞ (− S ∞ [g µν , Φ ])D∞ g µν D∞ Φ , (5.10)

over all compact four-geometries g µν which induce hij at the compact three-manifold. This three-
manifold is the only boundary of the all four-manifolds. Extending Hartle-Hawking proposal to the
p-adic minisuperspace, an adelic Hartle-Hawking wave function is the infinite product

ψ A (q ) = ∏ ∫ χ v (− S v [g µν , Φ ])Dv g µν Dv Φ , (5.11)
v

where path integration must be performed over both, Archimedean and non-Archimedean
geometries. If an evaluation of the corresponding functional integrals for a minisuperspace model
yields ψ (qα ) in the form (5.7), the such cosmological model is a Hartle-Hawking adelic one.
Now we consider the approach consists in the following p-adic proposal for the Hartle-Hawking
type of the wave function:

ψ ∞ (q ) = ∑ ∏ ∫ χ p (− S p [g µν , Φ ])Dp g µν Dp Φ , (5.12)
a.m. p

where summation is over algebraic manifolds.

The de Sitter minisuperspace model in D = 4 space-time dimensions is the Hartle-Hawking adelic
one. Namely, according to the Hartle-Hawking proposal one has

ψ v (q ) = ∫ Kv (q, T ;0,0 )dT , v = ∞,2,3,..., p,..., (5.13)

where
−
1
 λ2T 3 T q2 
Kv (q' ' , T ; q' ,0 ) = λv (− 8T ) 4T v 2 χ v − + (λq − 2 ) +  (5.14)
 24 4 8T 

38
is the kernel of the v -adic evolution operator. The functions λv (a ) have the properties

λv (a ) v = 1, λv (b 2 a ) = λv (a ), λv (a )λv (b ) = λv (a + b )λv (ab(a + b )) . (5.15)

Employing the p-adic Gauss integral

1
 β2 
χ p (αx 2 + βx )dx = λ p (α ) 2α
−
∫Qp p
2 χ p  − , α ≠ 0 ,
 4α 
(5.16)

one can rewrite p-adic version of (5.13) in the form

 λ2T 3  λq 1  
ψ p (q ) = ∫ dxχ p (qx )∫ DTχ p − + − − 2 x 2 T  . (5.17)
Qp
 24  4 2  

Taking the region of integration to be T p

≤ 1 one obtains

 λq 1  λ2
ψ p (q ) = ∫ dxχ p (qx )Ω − − 2x2 ,

≤ 1. (5.18)
Qp
 4 2 p 24 p

An evaluation of the integral (5.18) yields

(
ψ p (q ) = exp iπδ 1q δ p2 Ω q p ,
2
)( ) λ2
24
≤ 1, (5.19)
p

where δ ab is the Kronecker symbol. With regard ψ ∞ (q∞ ) , the result depends on the contour of
integration and has an exact solution

 
 1   1 − λq∞ 
ψ ∞ (q∞ ) = exp  Ai 2 
, (5.20)
 3λ  ( λ )
 2 3 

that can be rewritten also

(2λ )2 / 3 ln[ψ (q )] = 1 , (5.20b)

Ai (1 − λq∞ )
∞ ∞
3λ

where Ai( x ) is the Airy function. Thence, we obtain an adelic wave function for the de Sitter
cosmological model in the form

p
(
ψ A (q ) = ψ ∞ (q∞ )∏ exp iπδ 1q δ p2 Ω q p
2
) ( ), p
λ2
24
≤ 1. (5.21)
p

39
The necessary condition that a system can be regarded as the adelic one is the existence of p-adic
( )
ground state Ω qα p (α = 1,2,..., n ) in the way

∫ q 'α p ≤1
( )
Kp qα'' , T ; qα' ,0 dqα' = Ω qα'' ( ) p (5.22)

for all p but a finite set P . For the case of de Sitter model one obtains

λ2
( )
ψ p (q ) = Ω q p , T p ≤ 1,
24
≤ 1, p ≠ 2 ,
p

1 λ2
ψ p (q ) = Ω( q 2 ), T 2 ≤ , ≤ 1, p = 2 , (5.23)
2 24 2

what is in a good agreement with the result (5.21) obtained by the Hartle-Hawking proposal.

6. Number Theory: On Some Equations Concerning the Riemann Zeta Function.

A. The Goldston-Montgomery Theorem [7]

In the chapter “Goldbach’s numbers in short intervals” of Languasco’s paper “The Goldbach’s
conjecture”, is described the Goldston-Montgomery theorem.

THEOREM 1

Assume the Riemann hypothesis. We have the following implications: (1) If 0 < B1 ≤ B2 ≤ 1 and
1 X B1
F (X ,T ) ≈ T log T uniformly for ≤ T ≤ X B2 log 3 X , then
2π log 3 X

X
1 1
∫ (ψ (1 + δ )x ) − ψ (x ) − δ (x ) dx ≈ 2 δX
2 2
log , (6.1)
1
δ

1 1
uniformly for B2
≤ δ ≤ B1 .
X X
X
1 1
∫ (ψ ((1 + δ )x ) −ψ (x ) − δx ) dx ≈ 2 δX
2 2
(2) If 1 < A1 ≤ A2 < ∞ and log uniformly for
1
δ
1 1 1
1 / A1 3
≤ T ≤ 1 / A2 log 3 X , then F (X ,T ) ≈ T log T uniformly for
X log X X 2π

40
 T A1 ≤ X ≤ T A2 .

Now, for show this theorem, we must to obtain some preliminary results .

Preliminaries Lemma. (Goldston-Montgomery)

Lemma 1.

+∞
We have f ( y ) ≥ 0 I (Y ) = ∫ e f (Y + y )dy = 1 + ε (Y ) . If R(y) is a Riemann-
−2 y
∀y ∈ R and let
−∞
integrable function, we have:

b
b 
∫a R ( y ) f (Y + y )dy =  ∫ R( y )dy (1 + ε ' ( y )) .
 
a 

Furthermore, fixed R, ε ' (Y ) is little if ε ( y ) is uniformly small for Y + a − 1 ≤ y ≤ Y + b + 1 .

Lemma 2.

Let f (t ) ≥ 0 a continuous function defined on [0,+∞ ) such that f (t ) << log 2 (t + 2 ) .

If
T
J (T ) = ∫ f (t )dt = (1 + ε (T ))T log T ,
0
then
∞ 2
 sin ku  π  1
∫0  u  f (u )du =  2 + ε ' (k )k log k ,

with ε ' (k ) small for k → 0 + if ε (T ) is uniformly small for

1 1
2
≤ T ≤ log 2 k .
k log k k

Lemma 3.

Let f (t ) ≥ 0 a continuous function defined on [0,+∞ ) such that f (t ) << log 2 (t + 2) . If

∞ 2
 sin ku  π  1
I (k ) = ∫   f (u )du =  + ε ' (k ) k log , (6.2) then
0
u  2  k

T
J (T ) = ∫ f (t )dt = (1 + ε ' )T log T , (6.3)
0

1 1
with ε ' small if ε (k ) ≤ ε uniformly for ≤ k ≤ log 2 T .
T log T T

41
Lemma 4.

4 X i (γ − γ ')
Let F ( X , T ) := ∑
γ γ
0< , '<T 4 + (γ − γ ')
2
. Then (i) F ( X , T ) ≥ 0 ; (ii) F ( X , T ) = F (1 / X , T ) ; (iii) If

The Riemann hypothesis is preserved, then we have

 1  1  log log T 
F ( X , T ) = T  2 log 2 T + log X  + O 

X 
 2π log T
 

uniformly for 1 ≤ X ≤ T .

Lemma 5.

Let δ ∈ (0,1] and a (s ) =

(1 + δ )s − 1 . If c(γ ) ≤ 1 ∀y we have that
s

2 2
c(γ ) c(γ )
+∞ +∞
 2 1 
∫ a(it ) ∑γ 1 + (t − γ )2 dt = ∫ γ∑≤Z a(1 / 2 + iγ )1 + (t − γ )2 dt + O δ 2 log3 δ  + O Z log3 Z 
2

−∞ −∞

1
for Z > .
δ

For to show the Theorem 1, there are two parts. We go to prove (1).
We define
T 2
X iγ
J ( X , T ) = 4∫ ∑ dt .
0 γ 1 + (t − γ )
2

Montgomery has proved that J ( X , T ) = 2πF ( X , T ) + O (log 3 T )and thence the hypothesis
1 1
F (X ,T ) ≈ T log T is equal to J ( X , T ) = (1 + o(1))T log T . Putting k = log(1 + δ ) , we have
2π 2
2
2  sin kt 
a(it ) = 4  .
 t 

For the Lemma 2, we obtain that

∞ 2
X iγ π  1 π  1
∫ a(it ) ∑γ 1 + (t − γ )2 dt =  2 + o(1)k log k =  4 + o(1)δ log δ
2

1 3 1
for ≤T ≤ log 2 .
1 δ δ
δ log 2
δ

For the Lemma 5 and the parity of the integrand, we have that

42
 2
+∞
X iγ π  1
∫−∞ γ∑≤Z a ρ 1 + (t − γ )2 dt =  2 + o(1)δ log δ
( ) (a)

1 1
if Z ≥ log 3 .
δ δ

X iγ
From the S (t ) = ∑ a (ρ )
γ ≤Z 1 + (t − γ )
2
we note that the Fourier’s transformed verify that

Sˆ (u ) = π ∑ a(ρ )X iγ e(− γu )e
−2π u
.
γ ≤Z

From the Plancherel identity, we have that

2
+∞
2  1
∫ γ∑ a(ρ )X e(− γu ) du =  + o(1)δ log .
iγ − 4π u
e
−∞ ≤Z π  δ

For the substitution Y = log X , − 2πu = y we obtain

2
+∞
1
∫ γ∑ a(ρ )e
iγ (Y + y )
dy = (1 + o(1))δ log
−2 y
e . (b)
−∞ ≤Z δ

Using the Lemma 1 with R( y ) = e 2 y if 0 ≤ y ≤ log 2 and R( y ) = 0 otherwise, and putting

x = eY + y we have that
2
2X
3  1
∫ γ∑ a(ρ )x dx =  + o(1)δX 2 log .
ρ

X ≤Z 2  δ

Substituting X with X 2 − j , summarizing on j, 1 ≤ j ≤ K , and using the explicit formula for ψ ( x )

with Z = X log 3 X we obtain

X
1 1
∫ (ψ ((1 + δ )x ) −ψ (x ) − δx ) dx = 2 (1 − 2 + o(1))δX 2 log
2 −2 K
.
X 2− K
δ

Furthermore, we put K = [log log X ] and we utilize, for the interval 1 ≤ x ≤ X 2 − K , the estimate of
Lemma 4 (placing X 2 − K for X ). Thus, we obtain (1).
Now, we prove (2).
We fix an real number X 1 . Making an integration for parts between X 1 and X 2 = X 1 log 2 / 3 X 1 we
obtain, remembering that for hypothesis we have
X
1 2 1
∫1 (ψ ((1 + δ )x ) −ψ (x ) − δx ) dx ≈ 2 δX log δ ,
2

43
 X2
1  1
∫ (ψ ((1 + δ )x ) −ψ (x ) − δx ) x dx =  + o(1)δX 1−2 log .
2 −4
that (c)
X1 2  δ

Utilizing the estimate, valid under the Riemann hypothesis

X
2
∫ (ψ ((1 + δ )x ) −ψ (x ) − δx ) dx << δX
2 2
log 2 ,
1
δ

we obtain analogously as before that

∞
1  1
∫ (ψ ((1 + δ )x ) −ψ (x ) − δx ) x
2 −4
dx << δX 2−2 log 2 = o δX 1−2 log  . (d)
X2
δ  δ

Now, summarizing (c) and (d) and multiplying the sum for X 12 we obtain

∞
 x 2 X 12  1
∫1 min X 12 , x 2 (ψ ((1 + δ )x ) −ψ (x ) − δx ) x dx = (1 + o(1))δ log δ .
2 −2

Putting X 1 = X , Y = log X , x = eY + y and using the explicit formula for ψ ( x ) with Z = X log 3 X ,
we obtain the equation (b).

B. On the study of the behaviour of the argument of the Riemann function ζ (s ) with the
1
condition that s lies on the critical line s = + it , where t is real. [8]
2

We introduce the known functions S (t ) , S1 (t ) .

Definition 1. For real t , not equal to the imaginary part of a zero of ζ (s ) ,

1 1 
S (t ) = arg ζ  + it  ,
π 2 
1 
where arg ζ  + it  is obtained by continuous extension of arg ζ (s ) along the polygonal line
2 
starting at the point s = 2 (arg ζ (2 ) = 0 ) , going to the point s = 2 + it and then to the point
1
s = + it . If t is the imaginary part of a zero of ζ (s ) then
2

1
S (t ) = lim {S (t + δ ) + S (t − δ )} .
δ →0 2

Definition 2. For positive t the function S1 (t ) is defined by

t
S1 (t ) = ∫ S (u )du .
0

Definition 3. The following function is known as Selberg’s function:

44
 ησ (s ) = ∑ λνν − s .
ν <x

Furthermore, we have the following four theorems.

Theorem A. With H = T 27 / 82+α , 0 < α < 0.001 , T ≥ T1 (α ) > 0 , and 1 / 2 ≤ σ ≤ 1 we have

(
N (σ , T + H ) − N (σ , T − H ) = O HT −0.1α (2σ −1) log T , )
where N (σ , T ) is the number of zeros of the Riemann zeta function in the rectangle Re s ≥ σ ,
0 < Im s ≤ T , and the constant in the symbol O depends only on α .

Theorem B. Suppose that H = T 27 / 82+α , 0 < α < 0.001 , T ≥ T1 (α ) > 0 , and k is a natural number.
Then the following asymptotic formula holds:

T +H (2k )! H (log log T )k + O(H (log log T )k −1/ 2 ) ,

∫ S (t ) dt =
2k

k!(2π )
T 2k

where the constant in the symbol O depends only on α and k .

Theorem C. Suppose that H = T 27 / 82+α , 0 < α < 0.001 , T ≥ T1 (α ) > 0 , and k is a natural number.
Then the following asymptotic formula holds:

∫
T +H

T
(
S1 (t ) dt = ck H + O H log −1 T ,
2k
)
where ck is a positive constant depending only on k , and the constant in the symbol O depends
only on α and k .

Theorem D. Let H = T 27 / 82+α , 0 < α < 0.001 . Then there are positive numbers T1 = T1 (α ) and
A = A(α ) such that for T ≥ T1 the function S (t ) changes sign in the interval (T , T + H ) no fewer
than K times,

[ (
K = H (log T ) exp − A log log T .
1/ 3
)]
Now suppose that t ≥ 2 , 2 ≤ x ≤ t 2 and ρ = β + iγ run through the values of zeros of the Riemann
zeta function with the condition
3 β −0.5
x
t −γ ≤ .
log x

Definition 4. We write
1  1 2 
σ x ,t = + 2 max β − , .
2  2 log x 

Theorem 1. If t ≥ 2 and 2 ≤ x ≤ t 2 we have

45
 1 Λ x (n ) sin (t log n )  1 Λ (n )   1 
S (t ) = − ∑ σ x ,t
⋅ + O  σ x ,t −  ∑ σxx ,t +it  + O  σ x ,t −  log t  ,
π n< x 3 n log n  2  n< x3 n   2 

where
Λ x (n ) = Λ(n ) , 1 ≤ n ≤ x ,
x3 x2
log 2 − 2 log 2
Λ x (n ) = Λ(n ) n n , x ≤ n ≤ x2 ,
2
2 log x
x3
log 2
Λ x (n ) = Λ(n ) n , x 2 ≤ n ≤ x3 .
2 log 2 x

Theorem 2. If t ≥ 2 and 2 ≤ x ≤ t 2 we have

1 ∞ 1 Λ x (n )   1 
S1 (t ) = −
π∫ 0.5
log ζ (σ ) dσ +
π
∑ nσ x ,t 2
1 + σ x ,t −  log n  cos(t log n )
log n   2 
n< x3

 1
2
Λ x (n )   1
2

+ O  σ x ,t − 
 2
∑3 nσ x ,t +it  + O σ x,t − 2  log t  .

 n< x   

Now we describe the estimates of the mean deviations of S (t ) and S1 (t ) from the corresponding
original segments of the Dirichlet series.
To prove the following Theorems 3 and 4, on which the proofs of Theorem B and C are based, we
require the following Lemma 1 and 2.

Lemma 1. Suppose that H = T 27 / 82+α , 0 < α < 0.001 , T ≥ T1 (α ) > 0 , x ≥ 2 , k is a positive integer,
1 ≤ y ≤ x 8k , x 3 y 2 ≤ T 0.1α . Then for 0 ≤ ν ≤ 8k we have:

ν 1
 1  σ x ,t −
I (ν ) = ∫
T +H
(
σ x ,t −  y 2 dt = O H (log x ) ,
−ν
)
T
 2

where the constant within the symbol O depends only on α and k .

Proof. Using the definition of σ x,t we obtain

I (ν ) << H (log x ) + R(ν ) ,

−ν
(6.4)

where
1

ν
1  x3 y 2 ( )β−
2
R(ν ) = ∑ β − 
T − H <γ ≤T + 2 H  2 log x
, thence
β > 0.5

46
 1
 1  x3 y 2
ν
( )
β−
2
I (ν ) << H (log x )
−ν
+ ∑ β − 
T − H <γ ≤T + 2 H  2 log x
,
β > 0.5

and summation in the last sum is taken over zeros of ζ (s ) of the form ρ = β + iγ . We present the
terms in R(ν ) in a somewhat different form. Since

ν
 1 3 2 1

ν  1
ν −1 1 β −  x y ( )
β−
2
log x 3 y 2 
ν
1 3 2 1
x 3 y 2 2 du = 
β 2 β
∫1/ 2 log x  u − 2  ( u
)−
−∫ u −  x y ( )
u−
2 du ,
log x 1/ 2 log x  2

it follows that, by increasing somewhat the right-hand side of R(ν ) , we obtain

β  log x 3 y 2  ν
1  3 2 u − 12 ν  1
ν −1
3 2 u−2 
1
R(ν ) ≤ ∑ ∫1/ 2  log x  2 
  u −  x y + (  u )
−  x y du = ( )
T − 2 H <γ ≤T + 2 H  log x  2 
1  1  
ν ν −1
 log x 3 y 2  1 ν  u−  
1
=∫   u−  +  u −   × x 3 y 2 2  (
∑ )
g (γ ; u )du ,
1/ 2
 log x  2  log x  2    T −2 H <γ ≤T + 2 H 

where
g (γ ; u ) = 1 , 1 / 2 < u < β ,
g (γ ; u ) = 0 , u ≥ β .

Further, we have

∑
T − 2 H <γ ≤T + 2 H
g (γ ; u ) = ∑1 = N (u, T + 2 H ) − N (u, T − 2H ) .
T − 2 H <γ ≤T + 2 H
β >u

By Theorem A
  1
−0.2α  u −  
N (u , T + 2 H ) − N (u , T − 2 H ) = O HT  2
log T  ;
 
 

and by the hypothesis of the lemma,

1  1
0.1α  u − 
(x y )
3 2 u− 2
≤T  2
, log x 3 y 2 ≤ log T .

Hence we arrive at an estimate for R(ν ) given by

1 log T  1
ν
ν  1   −0.1α  u − 2 
ν −1  1

R(ν ) << ∫  u −  +  u −   HT (log T )du . (6.5)


1 / 2 log x
 2  log x  2  

47
It is easy to estimate the last integral. In fact, on changing the variable of integration to
 1
v = 0.1α  u −  log T , we obtain
 2

ν  1
0.05α log T (10v )
ν +1
1  1  −0.1α  u − 2  e−v −(ν +1) ∞ ν +1 − v
∫1/ 2  2 
 u −  T du = ∫0 (α log T )ν +1
dv ≤ 10ν +1 (α log T ) ∫0 v e dv =
− (ν +1)
= (ν + 1)!10ν +1 (α log T ) . (6.6)

From (6.5) and (6.6) we obtain

log 2 T −(ν +1) log T

R(ν ) << H (log T ) H (log T ) << H (log x ) ,
−ν −ν
+
log x log x

thence
1  1   −0.1α  u − 2 
ν ν −1  1
 log T  1 ν 
R(ν ) << ∫   u−  +  u −   HT (log T )du <<

1/ 2 log x
 2  log x  2  
log 2 T −(ν +1) log T
H (log T ) H (log T ) << H (log x ) . (6.6b)
−ν −ν
<< +
log x log x

The lemma follows from (6.4) and the last inequality.

Lemma 2. Suppose that H 0 > 1 , 1 < y ≤ H 01/ k , k is a natural number and, for prime p ≤ y ,

log p
α p ≤ c1 , α 'p ≤ c2 .
log y

Then
1 2k 2k
− −it
∫ ∑α
H0
∫ ∑α
H0
p p 2
dt = O(H 0 ) ; '
p p −1−i 2t
= O (H 0 ) .
0 0
p< y p< y

Theorem 3. Suppose that H = T 27 / 82+α , 0 < α < 0.001 , T ≥ T1 (α ) > 0 , k is a natural number,
T α / k < x < H 1/ k . Then
2k
T +H 1 sin (t log p )
∫ S (t ) + ∑ dt = O(H ) ,
T π p< x p

where the constant in the symbol O depends only on α and k .

Theorem 4. Suppose that H = T 27 / 82+α , 0 < α < 0.001 , T ≥ T1 (α ) > 0 , k is a natural number,
T α / k < x < H 1/ k .
Then
2k
1 1 Λ(n )
∫T
T +H
S1 (t ) +
π ∫0.5
∞
log ζ (σ ) dσ −
π
∑ 2
cos(t log n ) dt = O HL−2 k , ( )
n< x n log n

48
where the constant in the symbol O depends only on α and k .

We base the proofs of Theorems 3 and 4 on Theorems 1 and 2 and Lemmas 1 and 2. We give only
the proof of Theorem 3. In Theorem 1 we put x = T α / 60 k and add to both sides of the equation in
this theorem the sum
1 sin (t log p )
∑
π p < x3 p
.

On proceeding to estimates on the right-hand side we then have

1 sin (t log p )  7 
S (t ) + ∑
π p < x3 p
= O ∑ Rj  ,
  (6.7)
 j =1 

where
Λ( p ) − Λ x ( p ) −it Λ x ( p)  1 Λ ( p)
R1 = ∑ p , R2 = ∑ 1− p(0.5−σ x ,t
)
p −it , R3 =  σ x ,t −  ∑ σxx ,t +it ,
p < x3 p log p p< x3 p log p  2  p< x3 p
( )
Λ x p 2 −i 2t ( )(
Λ x p2
)  1 Λ p2 ( )
∑ ∑ 1 − p x ,t p −i 2t , R6 =  σ x ,t −  ∑ 2xσ x ,t +i 2t ,
1− 2σ
R4 = p , R5 =
p < x1.5 p log p p < x1.5 p log p  2  p< x1.5 p
log p
R7 = ∑ ∑ r/2
.
r >2 p r < x3 p

We transform each of the R j , j = 2,5,6,7, to a uniform form (similar to the form of R3 ). First of
all we have, obviously,

 1  1 log p  1
R7 << 1 <<  σ x ,t −  log T , R6 <<  σ x,t −  ∑ <<  σ x ,t −  log T ,
 2   2  p< x1. 5 p  2
 1  1
1
(
R5 << ∑ 1 − p x ,t <<  σ x ,t −  ∑
1− 2σ
) log p
<<  σ x ,t −  log T .
p < x1.5 p  2  p< x1.5 p  2

We now consider R2 . It is easy to see that

σ x ,t Λ x ( p) σ x ,t Λ x ( p)
R2 = ∫ ∑
0.5
p< x3 p
u +it
du ≤ ∫
0.5
∑
p< x 3 p u +it
du .

Further, for 1 / 2 ≤ u ≤ σ x ,t we have

Λ ( p) Λ ( p ) log( xp ) Λ x ( p ) log( xp )
1 1 1 1
u− ∞ −v σ x ,t − ∞ −v
∑3 pxu+it = x 2 ∫u x 2 ∑3 x p v+it dv ≤ x 2 ∫0.5 x 2 ∑ p v+it
dv .
p< x p< x p< x3

Therefore
Λ x ( p ) log(xp )
1 1
 1  σ x ,t − ∞ − v
R2 <<  σ x ,t −  x 2 ∫ x 2
 2 0.5
∑ p v+it
dv .
p < x3

49
From (6.7) and the estimates already obtained we obtain

2k
T +H 1 sin (t log p )  4 
∫ S (t ) + ∑ dt = O ∑ K j  , (6.8)
T π p< x3 p  j =1 

where
2k 2k
T +H Λ( p ) − Λ ( p ) T +H ( )
Λ x p 2 −i 2t
K1 = ∫
T
∑3 p logxp p −it dt , K 2 = ∫
T
∑ p log p
p dt ,
p< x p< x1.5
2k
T +H  1
K 3 = (log T ) ∫
2k
 σ x ,t −  dt ,
T
 2
2k
1  2 k  σ x ,t − 2   ∞ 2 −u Λ ( p ) log( xp ) 
2k  1 1
T +H 
K4 = ∫
T
 σ x ,t −  x 
 2
× ∫ x
 0.5
∑3 x p u+it du  dt .
p< x 

From Lemma 2 we have

K1 = O ( H ) , K 2 = O (H ) ;

from Lemma 1 follows that

  log T  2 k 
K 3 = O H    = O(H ) .
  log x  
 

Applying Cauchy’s inequality to K 4 , we obtain

1/ 2
 T + H 
1/ 2
 T + H  ∞ 1 −u 4k

1  4 k  σ x ,t −   Λ x ( p ) log(xp ) 
4k  1
 
K 4 << ∫  σ x ,t −  x  2  dt 
 T  2 
× ∫  ∫ x 2
T  0.5 ∑3 p u+it du

dt  . (6.9)
  p< x  

By Lemma 1 the first integral in (6.9) is estimated by

I1 << H (log x )
−4 k
. (6.10)

Further, using the Holder inequality, we obtain

4k 4 k −1
 ∞ 1 −u Λ ( p ) log( xp )   ∞ 1 −u  ∞
1
−u Λ x ( p ) log(xp )
 x2 ∑3 x p u+it du  ≤  ∫ x 2 du  ∑
 ∫0.5 ∫ x2 du =
 p< x  
0 .5

0.5
p < x3 p u +it
4k
Λ ( p ) log( xp )
1
− 4 k +1 ∞ −u
= (log x ) ∫ 0.5
x 2
∑3 x p u+it du .
p< x

The second integral in (6.9) is therefore estimated by

50
 4k
Λ ( p ) log( xp )
1
− 4 k +1 T + H ∞ −u
I 2 << (log x ) ∫ ∫ T 0.5
x 2
∑3 x p u+it dtdu =
p< x

4 k +1 ∞
1
−u  T +H Λ x ( p ) log(xp ) 
 ∑
= (log x ) ∫  ∫T
dt du << H (log x )
2 4k
x (6.11)
0.5
p < x3 p
u +it 2
log x 
 

(we have again used Lemma 2, estimating the last integral with respect to t ).

Therefore from (6.9)-(6.11) we obtain

K 4 << H .

From (6.8) and the estimates K j << H , j = 1,2,3,4, it follows that

2k
T +H sin (t log p )
1
∫ S (t ) + ∑ dt = O(H ) ,
T π p< y p

where y = T α / 20 k . Now let T α / k < x < H 1/ k . Going over to inequalities, using the relations already
proved and Lemma 2, we have

2k 2k 2k
T +H 1 sin (t log p ) T +H 1 sin (t log p ) T +H sin (t log p )
∫ S (t ) + ∑ dt << ∫ S (t ) + ∑ dt + ∫ ∑ dt << H
T π p< x p T π p< y p T
y≤ p< x p

which is was required to prove.

C. The P-N Model (Palumbo-Nardelli model) and the Ramanujan identities. [9]

Palumbo (2001) ha proposed a simple model of the birth and of the evolution of the Universe.
Nardelli (2005) has compared this model with the theory of the strings, and translated it in terms of
the latter obtaining:

 R 1 1 
− ∫ d 26 x g − − g µρ g νσ Tr (Gµν G ρσ ) f (φ ) − g µν ∂ µ φ∂ν φ  =
 16πG 8 2 
∞
 2
2 
1
2κ 10
1/ 2 1 ~ 2 κ
= ∫ 2 ∫ d 10 x(− G ) e − 2Φ  R + 4∂ µ Φ∂ µ Φ − H 3 − 102 Trν F2  ,
2 g10
( ) (6.12)
0  

A general relationship that links bosonic and fermionic strings acting in all natural systems.
It is well-known that the series of Fibonacci’s numbers exhibits a fractal character, where the forms
5 −1
repeat their similarity starting from the reduction factor 1 / φ = 0,618033 = (Peitgen et al.
2
1986). Such a factor appears also in the famous fractal Ramanujan identity (Hardy 1927):

51
 5 −1 5
0,618033 = 1 / φ = = R (q ) + , (6.13)
2 3+ 5  1 q f 5 (−t ) dt 
1+
2
exp ∫ 1/ 5

4/5 
 5 0 f ( −t ) t 

 
 
3 5
and π = 2Φ −  R ( q ) + , (6.14)
20  3+ 5  1 q f (−t ) dt  
5



1+
2
exp ∫ 1/ 5
 5 0 f (−t ) t  

4/5  

5 +1
where Φ= .
2

Furthermore, we remember that π arises also from the following identity (Ramanujan’s modular
equations and approximations to π ):

12 ( )(
 2 + 5 3 + 13  ) 24   10 + 11 2   10 + 7 2  
π= log   , (6.14a) and π= log   +   .
130 2 142   4   4 
      
(6.14b)

From (6.14b), we have that

π 142
24 = . (6.14c)
  10 + 11 2   10 + 7 2  
ln   +  
  4   4 
    

We remember that the “Ramanujan function” (an elliptic modular function that satisfies the
“conformal symmetry”) has 24 “modes” that correspond to the physical vibrations of a bosonic
string.
The introduction of (6.13) and (6.14) in (6.12) provides:






 R 1 1
− ∫ d x g −
26
⋅ − g µρ g νσ Tr (Gµν Gρσ ) f (φ ) +
16G   8
  
 3 5 
 2Φ −  R ( q ) + 
 20  3+ 5  1 q f (−t ) dt  
5

  1+
2
exp ∫
 5 0 f −t
1/ 5
(

t 4 / 5  
)
 

52
  
 
1 µν ∞ R 3 5 ⋅
− g ∂ µ φ∂ν φ ] = ∫ 2 ⋅ 2Φ − R(q) +
2 0 κ 20  3 + 5  1 q f ( −t ) dt  
5
exp ∫ 
11
 1+ 1/ 5 4/5  
 2  5 0 f (−t ) t  
1 ~ 2 κ 112
10 1/ 2
−2 Φ
[
∫ d x(− G ) e R + 4∂ µ Φ∂ Φ −
µ
2
H3 −
 
Trν
 
3  5 
2Φ −  R(q ) +  2 Rg102
20  3+ 5  1 q f (−t ) dt  
5

 1+
2
exp ∫ 1/ 5

4/5  
 5 0 f (−t ) t  

( F ) ],
2
2
(6.15)

which is the translation of (6.12) in the terms of the Theory of the Numbers, specifically the
possible connection between the Ramanujan identity and the relationship concerning the Palumbo-
Nardelli model.

7. On some possible mathematical connections.

In this section we describe some possible mathematical connections between some equations of
arguments above discussed and some equations concerning the Riemann zeta-function, the
Ramanujan’s modular equations and the Palumbo-Nardelli model.

First of all, now we describe the following possible mathematical connections.

If we take the eq. (1.52) of section 1 and the eq. (2.51) of section 2, we note that

1  2 1 ij ˆ 2 
−
16πG4 ∫ d 4 x γˆ 12 R42 − 2 h ij hij +
 R4 4 R42
h ∇ hij  ⇒

1  l2  R  R 
⇒ ∫ C µναβ k1 ( )C µναβ − 6l 2  ψ + k 2 ( ) ψ +  ,
4
d x g  R + (7.1)
16πG4  2  6  6 

hence the possible mathematical connection.

Furthermore, we note that the eqs. (3.70), (3.72) of section 3 and (5.10) of section 5, can be related,
and we obtain the following connections

i 1
S ( X ,χ )
iS ( x , χ ; Z , Z )
∫ dχe π ψ top (χ ) ⇒ det Imτ 2
∫ dχe ψ top (χ ) ⇒
[ ]
⇒ ∫ χ ∞ (− S ∞ g µν , Φ )D∞ g µν D∞ Φ . (7.2)

Also the eqs. (3.70), (3.72), of section 3, (4.17), of section 4, and (5.17) of section 5, can be
related, and we obtain the following interesting connections

53
 i 1
S ( X ,χ )
iS ( x , χ ; Z , Z )
∫ dχe π ψ top (χ ) ⇒ det Imτ 2
∫ dχe ψ top (χ ) ⇒
λ p (− 8T )  λ2T 3 T q2 
⇒∫ dT χ p  − + (λq − 2 ) + ⇒
T p ≤1
4T
1/ 2
 24 4 8T 
p

 λ2T 3  λq 1  
⇒ ∫ dxχ p (qx )∫ DTχ p − + − − 2 x 2 T  . (7.3)
Qp
 24  4 2  

With regard the possible mathematical connections concerning the Ramanujan’s modular equations,
we note that the eqs. (4.17) of section 4 and (5.17) of section 5, can be related with the eq. (6.14c)
of section 6, obtaining the following connection

λ p (− 8T )  λ2T 3 T q2 
∫T p ≤1
dT
4T
1/ 2
χ p  −
 24
+ (λq − 2) + ⇒
4 8T 
p

λT  2 3
 λq 1  
⇒ ∫ dxχ p (qx )∫ DTχ p − + − − 2 x 2 T  ⇒
Q p 24   4 2  
   10 + 11 2   10 + 7 2   
 ln   +   
   4   4  
      λq 1  
⇒ ∫ dxχ p (qx )∫ DTχ p − λ2T 3  + − − 2 x 2 T  . (7.4)
Qp
 π 142  4 2  
 
 

But, remember that the 24 “modes” correspond to the physical vibrations of a bosonic string, it is
possible to obtain the following interesting connection concerning the Palumbo-Nardelli model in
the terms of Number Theory

λ p (− 8T )  λ2T 3 T q2 
∫
T p ≤1
dT
4T
1/ 2
χ p  −
 24
+ (λq − 2 ) + ⇒
4 8T 
p

 λ2T 3  λq 1  
⇒ ∫ dxχ p (qx )∫ DTχ p − + − − 2 x 2 T  ⇒
Qp
 24  4 2  
   10 + 11 2   10 + 7 2   
 ln   +   
   4   4  
    +  λq − 1 − 2 x 2 T  ⇒
⇒ ∫ dxχ p (qx )∫ DTχ p − λ2T 3    
Qp
 π 142  4 2  
 
 

54
 




 R 1 1
− ∫ d x g −
26
⋅ − g µρ g νσ Tr (Gµν Gρσ ) f (φ ) +
16G   8
  
 3 5 
 2Φ −  R ( q ) + 
 20  3+ 5  1 q f (−t ) dt  
5

  1 +
2
exp  ∫
 5 0 f −t
1/ 5

t 4 / 5   ( )
 
 
 
1 µν ∞ R 3 5 ⋅
− g ∂ µ φ∂ν φ ] = ∫ 2 ⋅ 2Φ − R(q) +
2 0 κ 20  3+ 5  1 q f (−t ) dt 
5

exp ∫ 
11
 1+ 1/ 5 4/5  
 2  5 0 f (−t ) t  
1 ~ 2 κ 112
∫ d x(− G ) e R + 4∂ µ Φ∂ Φ −
10 1/ 2
−2 Φ µ
[ 2
H3 −
 
Trν
 
3  5  2
2Φ −  R(q ) + 2 Rg10
20  3+ 5  1 q f (−t ) dt  
5

 1+
2
exp ∫ 1/ 5

4/5  
 5 0 f (−t ) t  

(F ) ]
2
2

(7.5)

Now we describe the possible mathematical connections with some equations concerning the
Goldston-Montgomery theorem and the Riemann zeta-function.
We take the eqs. (1.29) of section 1, (6.3), (6.9) and (6.11) of section 6, then we obtain the
following connections

3N 2Ω 4 R l3  1 1 ij  3 l 2  1 1 
− 
16πG ∫
4 ij ij ij ˆ 2
log = W + d x γˆ h ∂ h + h h − 1 + + h h − h ∇ h
8π 2 l
CFT
 4l 4 y ij
l4
ij
2 R 2 
l 2 2
R
ij
8l 2 2
R
ij

   
1/ 2
 T + H  ∞ 1 −u 4k

 Λ x ( p ) log(xp )  
∑3 p u+it
T
⇒ (1 + ε ')T log T = ∫ f (t )dt ⇒ ∫  ∫ x 2 du dt  <<
0 T  0.5 
  p< x  
4k
Λ ( p ) log(xp )
1
− 4 k +1 T + H ∞ −u
<< (log x ) ∫ ∫
T 0.5
x 2
∑3 x p u +it dtdu =
p< x

4 k +1 ∞
1
−u  T +H Λ x ( p ) log( xp ) 
 ∑
= (log x ) ∫  ∫T
dt du << H (log x ) .
2 4k
x (7.6)
0.5
p< x3 p
u + it 2
log x 
 

With the eqs. (1.50) of section 1, (6.3), (6.9) and (6.11) of section 6, we obtain the following
connections

55
 1 3Ω 4 R42 3 N 2 Ω 4 R
∫ log 4 ⇒
4
− d x γ R + WCFT = − + 2
16πG4 4πG4 8π ρ
1/ 2
 T + H  ∞ 1 −u 4k

Λ x ( p ) log(xp ) 
T
 
⇒∫ f (t )dt = (1 + ε ' )T log T ⇒ ∫  ∫ x 2
T  0.5 ∑3 p u+it du  dt  <<
0   p< x  
4k
Λ ( p ) log(xp )
1
− 4 k +1 T + H ∞ −u
<< (log x ) ∫ ∫
T 0.5
x 2
∑3 x p u +it dtdu =
p< x

1 
4 k +1 ∞ −u T +H Λ x ( p ) log(xp ) 
= (log x ) ∫0.5 2  ∫T ∑ dt du << H (log x ) .
4k
x (7.7)
p u +it
log 2
x 
 p< x 3


With the eqs. (1.53) of section 1, (6.3), (6.9) and (6.11) of section 6, we obtain the following
connections

3N 2Ω4 3N 2Ω 4 2
−
16π 2
+
8π 2
R
log 4 = S − ∑ ∫ d 4 x' γˆ h kl ( x')H kl( p ) ( x') 
ρ
1
2
(p2 + 3 p + 6 ) ( )
p  64πG4 R4
N2 
+ 2 4
(2 p( p + 1)( p + 2)( p + 3) log(ρ / R4 ) + Ψ ( p )) ⇒
256π R4 
1/ 2
 T + H  ∞ 1 −u 4k

T  Λ x ( p ) log(xp )  
⇒ (1 + ε ')T log T = ∫
0
f (t )dt ⇒ ∫  ∫ x 2
T  0.5 ∑3 p u+it du

dt  <<
  p< x  
4k
Λ ( p ) log(xp )
1
− 4 k +1 T + H ∞ −u
<< (log x ) ∫ ∫T 0.5
x 2
∑3 x p u +it dtdu =
p< x

4 k +1 ∞
1
−u  T +H Λ x ( p ) log(xp ) 
 ∑
= (log x ) ∫  ∫T
dt du << H (log x ) .
2 4k
x (7.8)
0.5
p < x3 p
u +it 2
log x 
 

With the eqs. (2.51), (2.54) of section 2, (6.3), (6.9) and (6.11) of section 6, we obtain the following
connections

1  l2  R  R 
∫ Cµναβ k1 ( )C µναβ − 6l 2  ψ + k 2 ( ) ψ +  ⇒
4
d x g  R +
16πG4  2  6  6 
1 1 4
⇒ k ( ) + C ≡ ln 2 ⇒
4 4 l (− )
1/ 2
 T + H  ∞ 1 −u 4k

Λ x ( p ) log(xp ) 
T
 
⇒∫ f (t )dt = (1 + ε ' )T log T ⇒ ∫  ∫ x 2
T  0.5 ∑3 p u+it du  dt  <<
0   p< x  
4k
Λ ( p ) log(xp )
1
− 4 k +1 T + H ∞ −u
<< (log x ) ∫ ∫T 0.5
x 2
∑3 x p u +it dtdu =
p< x

1 
4 k +1 ∞ −u T +H Λ x ( p ) log( xp ) 
= (log x ) ∫0.5 2  ∫T
x ∑ p u + it
log 2
x
dt du << H (log x ) .

4k
(7.9)
 p< x 3


56
Furthermore, with the eqs. (5.17), (5.20b) of section 5, (6.3), (6.9) and (6.11) of section 6, we
obtain the following connections

 λ2T 3  λq 1 2 
∫Qp p ∫ p − 24 +  4 − 2 − 2 x T  ⇒
dxχ (qx ) DTχ

⇒
(2λ )2 / 3 ln[ψ (q )] = 1 ⇒
Ai (1 − λq∞ )
∞ ∞
3λ
1/ 2
 T + H  ∞ 1 −u 4k

T  Λ x ( p ) log(xp )  
⇒ (1 + ε ')T log T = ∫
0
f (t )dt ⇒ ∫  ∫ x 2
T  0.5 ∑3 p u+it du  dt  <<
  p< x  
4k
Λ ( p ) log(xp )
1
− 4 k +1 T + H ∞ −u
<< (log x ) ∫ ∫
T 0.5
x 2
∑3 x p u +it dtdu =
p< x

1 
4 k +1 ∞ −u T +H Λ x ( p ) log( xp ) 
= (log x ) ∫0.5 2  ∫T ∑ dt du << H (log x ) .
4k
x (7.10)
p u + it
log 2
x 
 p< x 3


Conclusion

Hence, in conclusion, also for some mathematical sectors concerning the Hartle-Hawking no
boundary proposal concerning the Randall-Sundrum cosmological scenario, Hartle-Hawking wave-
function in the mini-superspace sector of physical superstring theory and p-adic Hartle-Hawking
wave function, can be obtained interesting and new possible connections between them, String
Theory and some sectors of Number Theory, principally the Ramanujan’s modular equations and
some formulae related to the Riemann zeta function.
Furthermore, also the fundamental relationship concerning the Palumbo-Nardelli model, can be
related with some equations (see eq. (7.5)) regarding p-adic models in the Hartle-Hawking proposal
and p-adic and adelic wave functions of the Universe.

Acknowledgments

I would like to thank Prof. Branko Dragovich of Institute of Physics of Belgrade (Serbia) for the
important and fundamental advices and references that he has give me and his availability and
friendship with regard me. Furthermore, I would like to thank Prof. G. Tasinato of Oxford
University for his friendship and availability and the Prof. A. Palumbo whose advices has been
invaluable for me. In conclusion, I would like to thank also F. Di Noto for useful discussions and
past contributes with regard the Number Theory.

References

[1] S. W. Hawking, T. Hertog and H. S. Reall. – “Brane New World” – arxiv:hep-th/0003052 v3.

[2] A.O. Barvinsky, A. Yu. Kamenshchik, A. Rathke, C. Kiefer. – “Nonlocal Braneworld Action:
An Alternative to Kaluza-Klein Description” – arxiv:hep-th/0206188v3.

57
[3] J. Michelson. – “Compactifications of Type IIB Strings to Four Dimensions with Non-trivial
Classic Potential” – arXiv:hep-th/9610151v1.

[4] H. Ooguri, C. Vafa and E. Verlinde. – “Hartle-Hawking Wave-Function for Flux

Compactifications: The Entropic Principle” – arXiv:hep-th/0502211 v2.

[5] G. S. Djordjevic, B. Dragovich, L. D. Nesic, I. V. Volovich. – “p-Adic and Adelic

Minisuperspace Quantum Cosmology” – arXiv:gr-qc/0105050 v2.

[6] B. Dragovich. – “p-Adic and Adelic Cosmology: p-Adic Origin of Dark Energy and Dark
Matter” – arXiv:hep-th/0602044 v1.

[7] A. Languasco. – “La Congettura di Goldbach” – Tesi Dottorato di Ricerca in Matematica – VI

Ciclo (1995).

[8] A. A. Karatsuba. – “On the Function S (t ) ” – Izvestiya: Mathematics 60:5 901-931 (1996)

[9] Palumbo A, Nardelli M. – “The Theory of String: A Candidate for a Generalized Unification
Model” – CNRSOLAR 122JA2006 – 22.11.2006.
Palumbo A, Nardelli M. – “Water and Life (A contribution to the Langlands Program)” – Boll.
Soc. Natur. Napoli – Nuova Serie – Vol. III – (2005-2006); 115-120.

Finito di Stampare nel mese di Luglio 2007

presso DI. VI. Service – Via Miranda, 50 – 80131 Napoli
Tutti i diritti riservati

58