Scribd
Upload
10 views

Circe Compactification in String Theory

This document discusses computing an integral in type IIB string theory related to circle compactification. It presents the integral to be computed, applies an unfolding trick to evaluate it, and expresses the result using Dedekind eta functions and Eisenstein series. Several steps of the calculation are shown.

Uploaded by

underscoreiw
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
10 views

Circe Compactification in String Theory

This document discusses computing an integral in type IIB string theory related to circle compactification. It presents the integral to be computed, applies an unfolding trick to evaluate it, and expresses the result using Dedekind eta functions and Eisenstein series. Several steps of the calculation are shown.

Uploaded by

underscoreiw
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

R4 threshold in type IIB string theory

February 16, 2023

Circle compactification
The integral we have to compute is

d2 τ
Z
2 Γ1,1 (R)Zcirc (a, τ ) (1)
F τ2

where
√

m2
+n2 R2
X
Γ1,1 (R) = τ2 e−πτ2 R2 e2iπτ1 mn (2)
(m,n)∈Z2

We note that in this case the unregulated integral is completely finite and there-
fore the regulator is not strictly needed. Nevertheless, as we shall see, adding
it is useful in order to compare the new result with the two dimensional case.
Applying the unfolding trick we get
Z
1 π
P1 (a) = Γ1,1 (R) = (R + R−1 )
a F 3a
2 ∞ dτ2
Z 2
 ∞ 2 (3)
X −πτ2 m +n2 R2
X − πr
P2 (a) = 3/2
e R2 δ mn,0 e a2 τ2 ,
a 0 τ r=1
2 (m,n)∈Z2

Note that the Kronecker delta function is the result of the τ1 integral.
We also point out that there is no need to use the primed version of the P ’s as
we already noted that in this case there are no divergences.
Computing the second integral and summing in r we get
∞ ∞ ∞
X 1 X  2πm
 X  2πnR

P2 (a) = 4 −4 log 1 − e− aR − 4 log 1 − e− a , (4)
r=1
r m=1 n=1

where we note the presence of a logarithmic divergence in the first term: this
is a direct consequence of adding Zcirc to the original integral. Nevertheless we
note that this divergence is completely independent of a and therefore we can
ignore it, in analogy of what happened in the two dimensional case.

1
Putting everything together we have
∞ ∞
π −1
X 
− 2πm
 X  2πnR

P (a) = (R + R ) − 4 log 1 − e aR −4 log 1 − e− a
3a m=1 n=1 (5)
  iR   i 
= − 4 log η η
a aR
where we see manifestly the R → 1/R symmetry as expected, since the integral
we starting with is invariant under that transformation.
We also note that the bit coming from P1 combines with the logarithms giving
us a Dedekind eta function.
Applying the regulator and using
d
−4πi log η(τ ) = G2 (τ ), (6)
dτ
P∞
with G2 (τ ) = 2ζ(2) + 2(2πi)2 n=1 σ(n)e2πinτ , we get

Aρ ∂  
2
a2 P (ρa) − P (a)
1 + ρa ∂a
( ) (7)
Aρ 1  iR  1  i  R  iR  1  i 
= RG2 + G2 − G2 − G2 .
1 + ρa2 π a R aR ρ aρ ρR aRρ

Partially summing the Eisenstein series, we can rewrite everything in the fol-
lowing way

Aρ ∂   π R + R−1
2
a2 P (ρa) − P (a) =
1 + ρa ∂a 3 1 + ρa2
∞
(
2πAρ X 1 2
 πk  1
2 πk
 
+ csch − csch (8)
1 + ρa2 ρR aρR R aR
k=1
)
R  πkR   πkR 
+ csch2 − Rcsch2 .
ρ aρ a

It should be noted from eq (7) that, again, the two results coming from P1
and P2 recombine in Eisenstein series. From (7) we can retrieve the P1 result
partially summing the series, as we can se from the first term in the right hand
side of eq (8).
...........

You might also like