On the Riemann Hypothesis, Complex Scalings and

Logarithmic Time Reversal

(Carlos Castro Perelman, Journal of Geometry and Physics 129

(2018) 133-141)
• Riemann’s outstanding hypothesis that the nontrivial complex
zeros of the zeta-function ζ(s) are of the form sn = 12 ± iλn , is one
of most important open problems in pure mathematics.
• ζ(s) has trivial zeros only at the negative even integers
ζ(−2n) = 0
• the zeros of zeta are deeply connected with the distribution of
primes.
The Riemann zeta function is defined for complex s with real part
greater than 1 by the absolutely convergent infinite series
∞
X 1 Y 1
ζ(s) = = −s
, Real(s) > 1.
ns p
1 − p
n=1

The Riemann hypothesis discusses zeros outside the region of

convergence of this series and Euler product. To make sense of the
hypothesis, it is necessary to analytically continue the function to
give it a definition that is valid for all complex s.
• Dirichlet η function, or alternating zeta function,
∞
2 X (−1)n+1
η(s) = (1 − s ) ζ(s) = =
2 ns
n=1

∞
x s−1
Z
1
dx
Γ(s) 0 ex + 1
• the series on the right converges not just when the real part of s
is greater than one, but more generally whenever s has positive
real part.
• Thus, this alternative series extends the zeta function from
Re(s) > 1 to the larger domain Re(s) > 0, excluding the zeros
s = 1 + i(2πn/ ln(2)) of 1 − 2/2s .
• The zeta function can be extended to these values, as well, by
taking limits, giving a finite value for all values of s with positive
real part except for a simple pole at s = 1.
Physical approaches to the RH
• Hilbert-Polya Conjecture
Hilbert-Polya’s suggestion that a physical way to verify RH could
be attained if the nontrivial zeros of ζ(s) could be given the
spectrum of an operator 2I + iH where H is self-adjoint.
Consequently, the operator H might have an interpretation as a
Hamiltonian of a physical system and, therefore, the key to the
proof of the RH may have been coded in physics
• String Theory
A physical interpretation of the location of the nontrivial Riemann
zeta zeroes which corresponds to the presence of tachyonic
resonances/tachyonic condensates in bosonic string scattering
amplitudes was found in C. C (IJGMMP 3, no.2 (2006))
• If there were zeros outside the critical line violating the RH these
zeros do not correspond to any poles of the string scattering
amplitude
Let us begin with the one-dimensional differential operators

d dV
D1 = − + + k, t ∈ R + (1)
d ln t d ln t
where k is an arbitrary parameter. The eigenvalues s can be
complex-valued in general, and its eigenfunctions are

ψs (t) = t −s+k e V (t) . (2a)

• D1 is not self-adjoint since it is an operator that does not admit

an adjoint extension to the whole real line characterized by the real
variable t.
The conjugation operation C acting on the eigenfunctions is
defined as
∗ +k ∗ +s
ψs (t) = t −s+k e V (t) → ψs ∗ (t) = t −s e V (t) = t −s
ψs (t).
(2b)
one learns that the conjugation operation C can also be recast as a
scaling transformation of ψs (t) by t-dependent (local) scaling
factors
 ∗ +s ∗ +s)
t −s = e (−s (ln t)
=e 2 i Im(s) (ln t)
= e iθs (t) (2c)

which amounts to a t-dependent phase rotation θs (t) which is

proportional to the imaginary part of s and to ln t.
• We also define the “mirror” operator to D1 as follows,

d dV (1/t)
D2 = − + k. (3)
d ln t d ln t
that is related to D1 by the substitution t → 1/t
• D2 is not self-adjoint either. The eigenfunctions of the D2
operator are Ψs ( 1t ), with the same eigenvalue s
1 1
D2 ψs ( ) = s ψs ( ) (4)
t t
Out of the infinity of possible choices for V (t), one may choose,
for example, V (t) which is related to the Bernoulli string spectral
counting function, and given by the Jacobi theta series as follows
∞
2tl
X
e 2V (t)
= e −πn = 2ω(t l ) + 1. (5)
n=−∞

where l is another real parameter introduced corresponding to the

scaling exponent t l in eq-(5).
The related theta function defined by Gauss is given by
∞
2 /x
X
G (1/x) = e −πn = 2ω(1/x) + 1. (6)
n=−∞

∞ 2
e −πn x . The Gauss-Jacobi series obeys the
P
where ω(x) =
n=1
relation
 1 √
G ( ) = x G (x). (7)
x
resulting from the Poisson re-summation formula.
Setting e 2V (t) = G (t l ) = G (x), where x ≡ t l ; using the
transformations (7), and when the parameters l, k are constrained
to obey the condition l = 4(2k − 1), one can show that the
eigenfunctions of the D2 operator Ψs ( 1t ), satisfy the key relation
1
ψs ( ) = ψ1−s (t), log (1/t) = −logt. (8)
t
A little algebra reveals that the pair of mirror Hamiltonians given
by

HA = D2 D1 , HB = D1 D2 (9a)
when l = 4(2k − 1), satisfy
1 1
HA ψs (t) = s(1 − s)ψs (t). HB ψs ( ) = s(1 − s)ψs ( ). (9b)
t t
due to the relation Ψs (1/t) = Ψ1−s (t) based on the modular
√
properties of the Gauss-Jacobi series, G ( x1 ) = x G (x).
Therefore, despite that HA , HB are not Hermitian they have the
same spectrum s(1 − s) which is real-valued only in the critical line
and in the real line.
The inner product is defined as follows
Z∞
dt
hψs1 |ψs2 i = ψs∗1 ψs2 , ψs (t) = t −s+k e V (t) (10a)
t
0

Based on this definition, the inner product of two eigenfunctions of

D1 is
Z∞
hψs1 |ψs2 i = e 2V (t) t −s12 +2k−1 dt; s12 ≡ s1∗ + s2 (10b)
0
A regularization of the integral (10a) can be attained by removing
the zero n = 0 mode of the Gauss-Jacobi series. Upon doing so
and performing the change of variables x = t l , it leads to
Z∞  
2 2(−s12 +2k)
−1 2 2 2
e 2V x 2l dx = Z (2k − s12 ) = Z [s] (10c)
l l l l
0

where we have denoted

2
s12 = s1∗ + s2 = x1 + x2 + i(y2 − y1 ), s = (−s12 + 2k) (10d)
l
 Z∞ X
∞
2
e −πn x x s/2−1 dx =
0 n=1
Z ∞
x s/2−1 ω(x)dx
0
Z ∞
1
= + [x s/2−1 + x (1−s)/2−1 ] ω(x)dx
s(s − 1) 1

= Z (s) = Z (1 − s),
(11)
• where the completed zeta function is defined as
s
Z (s) ≡ π −s/2 Γ( ) ζ(s). (12)
2
• and which obeys the functional relation Z (s) = Z (1 − s).
It is very important to emphasize that the removal of the zero
mode n = 0 of the Gauss-Jacobi theta series G (x), in the
evaluation of the inner product in eqs-(10), invalidates the
transformation property of G (x) of eq. (7) under x → 1/x, and
consequently
1
ψs ( ) = ψ1−s (t) (8)
t
and eqs-(9) will no longer be satisfied due to the eventual loss of
CT invariance.
For this reason we shall introduce next a different family of theta
series which avoids the removal of the zero mode.
• The completed zeta function Z [s] results from the evaluation of
the Mellin transform of the Gauss-Jacobi theta series
Another Theta Series
There is a two-parameter family of theta series Θj,m (t) that yield
well defined inner products without the need to extract the zero
mode n = 0 divergent contribution. We found that the two
parameter family of theta series related to a different choice for
V (t) is given by

n=∞ √ 2 l
X
e 2Vj,m (t) l
= Θj,m (t ) ≡ n 2m
H2j (n 2πt l ) e −πn t , (14)
n=−∞

m = 1, 2, . . . ; j = 0, 1,2, . . . Due to the weighted theta series

in eq-(14) the zero mode n = 0 does not contribute to the sum in
eq-(14), since m is a positive integer,
• the Mellin transform of Θj,m (t l ) = Θj,m (x) (x = t l ) after
exploiting the symmetry of the even-degree Hermite polynomials
was provided given by Mark Coffey (2006)
 n=∞
Z ∞
2 X 2m √ 2
[ n H2j (n 2πx) e −πn x ] x s/2−1 dx =
0 l
n=1
2 s
− (8π)j Pj (s) π −s/2 Γ( ) ζ(s−2m); Re s > 1+2m, m = 1, 2, ....
l 2
(15)

The polynomial pre-factor in (15) is given in terms of a

terminating Hypergeometric series

(2j)! s 1
Pj (s) = (8π)−j (−1)j F
2 1 (−j, ; ; 2). (16)
j! 2 2

• The polynomial Pj (s) has simple zeros on the critical line

Re s = 12 , obeys the functional relation Pj (s) = (−1)j Pj (1 − s)
and in particular Pj (s = 12 ) = 0 when j = odd. It is only when
j = even that Pj (s = 12 ) 6= 0.
• Despite this singular behavior at x = 0, the doubly improper
integral (15) converges when Re s > 1 + 2m
• It is then via the analytical continuation of the zeta function to
the region Re s < 1 + 2m as found by Riemann in his celebrated
paper that allows us to establish a one-to-one correspondence
between the nontrivial zeta zeros inside the critical strip, and the
zero inner products :
• zeros of zeta ↔ zero inner products < Ψs |Ψs 0 >= 0.
A Poisson re-summation formula for Θj,m (x) (14) leads to the
important relation

(−1)j 1
√ Θj,m ( ) = Θj,m (x). (17)
x x
which allows us to show that only when j = even one can
implement the T transformations of the new eigenfunctions
Ψsj,m (t) = t −s+k e Vj,m (t) of D1 , and corresponding to the weighted
theta series Θj,m (t l ) of eq-(14), giving

j,m 1 j,m
T Ψj,m
s (t) = Ψ s ( ) = Ψ1−s (t) (18)
t
this relationship requires that one must have

j = even, l = 4(2k − 1) (19)

Therefore, the eigenfunctions and eigenvalues of the pair of
Hamiltonians are

HA Ψj,m j,m
s (t) = s(1 − s) Ψs (t)

1 j,m 1
HB Ψj,m
s ( ) = s(1 − s)Ψs ( ) (20)
t t
subjected to the conditions in eq-(19).
We explicitly inserted the superscripts j, m in
Ψj,m
s (t) = t
−s+k e Vj,m (t) to denote the j, m dependence in the

definition of V (t) given by eq-(14).

In what follows we shall omit the superscripts j, m for convenience.
Defining the state

1
ΨCT
s (t) ≡ C T Ψ s (t) = C Ψ s ( ) = C Ψ1−s (t) = Ψ1−s ∗ (t)
t
(21)
one finds that ΨCT s (t) is also an eigenfunction of HA with an
eigenvalue s ∗ (1 − s ∗ ) :

HA | ΨCT
s (t) > = HA CT | Ψs (t) > = HA | Ψ1−s ∗ (t) > =

s ∗ (1−s ∗ ) | Ψ1−s ∗ (t) > = s ∗ (1−s ∗ ) CT | Ψs (t) > = (Es )∗ | ΨCT

s (t) > .
(22)
where we have defined (Es )∗ = s ∗ (1 − s ∗ ).
CT linear versus CT anti-linear

If the CT action on s(1 − s) Ψs is linear :

CT s(1 − s) Ψs = s(1 − s) CT Ψs

instead of antilinear :

CT s(1 − s) Ψs = s ∗ (1 − s ∗ ) CT Ψs

, and if there is CT symmetry

< Ψs | [HA , CT ] | Ψs > = 0 ⇒

< Ψs | HA CT | Ψs > − < Ψs | CT HA | Ψs > =

(Es )∗ < Ψs | CT | Ψs > − Es < Ψs | CT | Ψs > =
(Es∗ − Es ) < Ψs | CT | Ψs > = 0. (23)
Similar results follow for the HB operator. From eq-(23) one has
two cases to consider
• Case A : If the pseudo-norm is null

< Ψs | CT | Ψs > = 0 ⇒ (Es − Es∗ ) 6= 0 (24)

then the complex eigenvalues Es = s(1 − s) and Es∗ = s ∗ (1 − s ∗ )
are complex conjugates of each other.
• In this case there exists the possibility that there are quartets of
non-trivial Riemann zeta zeros (off the critical line) given by
sn , 1 − sn , sn∗ , 1 − sn∗ .
• Case B : If the pseudo-norm is not null :

< Ψs | CT | Ψs > 6= 0 ⇒ (Es − Es∗ ) = 0 (25)

then the eigenvalues are real given by

Es = s(1 − s) = Es∗ = s ∗ (1 − s ∗ ) = Real

and which implies that s = real (location of the trivial zeta zeros)
and/or s = 21 + iρ (location of the non-trivial zeta zeros).
• In this case the RH would be true and the non-trivial Riemann
zeta zeros are given by sn = 12 + iρn and 1 − sn = sn∗ = 12 − iρn .
We are going to prove next why Case A does and cannot occur,
therefore the RH is true because we are left with case B.
It remains to prove, when l = −2, k = 14 , t l = x, and
s12 = s1∗ + s2 = s1∗ + (1 − s1∗ ) = 1, that

j,m
< Ψj,m j,m j,m
s (t) | CT | Ψs (t) > = < Ψs (t) || Ψ1−s ∗ (t) > =

n=∞
Z ∞
2 X 2m √ −πn2 x
2(−s12 +2k)
−1
[ n H2j (n 2πx) e ]x 2l dx =
0 l
n=1
2 1 1 1
− (8π)j Pj (s = ) π −1/4 Γ( ) ζ( −2m) 6= 0; j = even, m = 1, 2, 3, ..
l 2 4 2
(26)
Hence, one arrives at a definite solid conclusion : Because
1
ζ( − 2m) 6= 0, m = 1, 2, ....
2
and
1
Pj ( ) 6= 0
2
when j = even in eq-(26), then

< Ψs | CT | Ψs > 6= 0

so this rules out case A in eq-(1.24), and singles out case B in

eq-(25) leading to the conclusion that
• Es = s(1 − s) = real ⇒ s = 12 + iρ (and/or s = real), which is
the location of the non-trivial zeta zeros (if the RH is true) and
trivial zeta zeros, respectively.
If CT anti-linear

If the CT action is antilinear:

CT (Es |Ψs >) = (Es )∗ (CT |Ψs >)

then
[H, CT ] |Ψs > = (Es∗ − Es∗ ) CT |Ψs > = 0
is trivially zero so that [H, CT ] = 0.
If, in addition, |Ψs > is CT -invariant

CT |Ψs > = |Ψ1−s ∗ > = |Ψs >

it then leads to the condition 1 − s ∗ = s ⇒ s = 21 + iρ , and such
that s must be confined to the critical line.
• To sum up, antilinearity, and CT invariance of the states Ψs
yields that s = 12 + iρ.
• We are not done yet. We still have to analyze other inner
products.
• In particular, the inner product of the states Ψj,ms (t) with the
states Ψ 1 +2m (t), when m = 1, 2, ...., is given by
2

< Ψj,m
1
+2m
(t) | Ψj,m
s (t) > = −
2

2 s + 2m
(8π)j Pj (s + 2m) π −(s+2m)/2 Γ( ) ζ(s). (27)
l 2
to arrive at this result above requires performing the change of
variables t l = x, and fixing uniquely the values l = −2; k = 41
obeying the required constraint l = 4(2k − 1).
Therefore, for each given value of j, m, the non-trivial zeta zeros
sn = 12 ± ρn are in a one-to-one correspondence with the states
Ψj,m j,m
sn (t) orthogonal to the states Ψ 1 +2m (t) in eq-(27) :
2

< Ψj,m
1
+2m
(t) | Ψj,m
sn (t) > =
2

2 sn + 2m
− (8π)j Pj (sn +2m) π −(sn +2m)/2 Γ( ) ζ(sn ) = 0; m = 1, 2, 3, ....
l 2
(28)
Let us analyze the scenario in case the RH were not true. Given
any real number s 0 = 12 + ξ ∈ R, such that ξ > 2m, and a
complex number s = 12 + iλ ∈ L residing in the critical line, let us
imagine that the zero inner product

< Ψsj,m j,m

0 (t) | Ψs (t) > = < Ψj,m
1
+ξ
(t) | Ψj,m
s (t) > =
2

2 s +ξ
− (8π)j Pj (s + ξ) π −(s+ξ)/2 Γ( ) ζ(s + ξ − 2m) = 0 (29)
l 2
as well as its complex conjugate

< Ψsj,m j,m j,m

0 (t) | Ψs ∗ (t) > = < Ψ 1
+ξ
(t) | Ψj,m
s ∗ (t) > =
2

2 ∗ −(s ∗ +ξ)/2 s∗ + ξ
j
− (8π) Pj (s +ξ) π Γ( ) ζ(s ∗ +ξ−2m) = 0 (30)
l 2
generate other nontrivial zeros off the critical line given by
z ≡ s + ξ − 2m; z ∗ ≡ s ∗ + ξ − 2m, respectively.
• By symmetry, 1 − z ∗ = s − ξ + 2m, and 1 − z = s ∗ − ξ + 2m
should also be another pair of complex conjugate (putative) zeros
off the critical line, since the number of zeros off the critical line
must appear in quartets resulting from the symmetry property of
the completed zeta function Z (s) = Z (1 − s).
• Since m are Z, and from the fundamental integral in eq-(15),
one can infer the pair of complex conjugate zeros s − ξ + 2m, and
s ∗ − ξ + 2m, cannot be obtained from an orthogonality condition
among Ψjm jm
s (t) and Ψs 0 (t), for any s located in the critical line, and
for any s 0 = 12 + ξ located in the real line (ξ > 2m), respectively.
• Therefore, only half of the quartet of putative zeros off the
critical line could be found from the orthogonality conditions.
One could still have the following inner products

< Ψj,m
1
+ξ 0 (t) | Ψj,m
s (t) > =
2

2 0 −(s+ξ 0 )/2 s + ξ0
j
− (8π) Pj (s + ξ ) π Γ( )ζ(s + ξ 0 − 2m) (31)
l 2
of the states Ψsj,m (t) with another state Ψj,m
1
+ξ 0
(t) associated to a
2
different value of ξ0 6= ξ, and such that

ζ(s + ξ 0 − 2m) = ζ(s − ξ + 2m) = 0 (32)

This requires performing the identification

s+ξ 0 −2m = s−ξ+2m ⇒ ξ + ξ 0 = 4m, ξ > 2m > 0, 0 < ξ 0 < 2m

(33)
so one has found the sought-after missing pair of orthogonal states
Ψjm
1
+ξ 0 (t), Ψjm
s (t) which generates the putative zero off the critical
2
line given by s − ξ + 2m, plus its complex conjugate.
• But in this case one would have to choose two different states,

Ψjm
1
+ξ
(t), Ψjm
1
+ξ 0
(t)
2 2

in order to evaluate the inner products with the states Ψj,m

s (t),
where s lives in the critical line.
• The energy eigenvalues Es = s(1 − s) associated with these two
states are 14 − ξ 2 , 14 − (ξ 0 )2 , respectively.
• However, the fact that the above energy eigenvalues are not the
same is very problematic if one wishes to label the states
Ψjm
1
+ξ
(t); Ψjm
1
+ξ 0
(t) as two degenerate states which are both
2 2
orthogonal to the states Ψjm
s (t).
• From the de la Vallee-Poussin theorem that there are no zeros
when the real part of s is 0, 1, one can infer that

1 1 1 1
2m− < ξ 0 < 2m < ξ < 2m+ ⇒ −ξ 2 < −(ξ 0 )2 (34)
2 2 4 4

so that the state Ψjm

1
+ξ
(t) has lower energy than the state
2
Ψjm
1
+ξ 0
(t). Furthermore,
2

1 1
< Ψ 1 +ξ0 (t)jm |Ψjm
1
+ξ
(t) > ∼ Pj ( + ξ + ξ 0
) ζ( + ξ + ξ 0
− 2m) 6= 0
2 2 2 2
(35)
jm jm
so that the states Ψ 1 +ξ (t), Ψ 1 +ξ0 (t) are not orthogonal.
2 2
 0
• Another possibility is to look at the inner products Ψjm
s (t) with
0
Ψjm
1
+ξ 0
(t), where m 6= m0 .
2
The orthogonality condition yields in this case the relation

s + ξ 0 − 2m0 = s − ξ + 2m ⇒ ξ + ξ 0 = 2m + 2m0 ;

ξ > 2m > 0, 0 < ξ 0 < 2m0 (36)

• However, in this case one would be choosing eigenfunctions

0 jm0
Ψjm
s (t), Ψ 1
+ξ 0
(t) of another different operator D10 resulting from
2
the different value of m0 6= m, and which leads to a different
weighted theta series in eq-(14).
• Therefore, in order to generate the quartets of putative zeros off
the critical line one would be forced to look at the orthogonality
conditions of Ψjm s (t) with respect to two states which are not
degenerate in energy,
• or involving eigenfunctions of many different operators
D1 , D10 , D100 , · · · associated with many different functions
Vj,m (t), Vj,m0 (t), Vj,m00 (t), · · · , instead of focusing on the
orthogonality conditions involving eigenfunctions Ψj,m s (t) of only
one operator D1 , associated to only one function Vj,m (t).
Conclusion

• Using the weighted Θ series one recaptures the nontrivial zeros

from the orthogonality conditions involving inner products of states
of the same energy Ψjm1
+2m
(t), with the states Ψ j,m
s (t), for
2
1
s= 2 + iλ residing in the critical line,

< Ψj,m
1
+2m
(t) | Ψj,m
sn (t) > ∼ ζ(sn ) = 0 (37)
2

1
when sn = 2 ± iλn .
• The analytical continuation of the zeta function to the region
inside the critical strip was required in order to establish the
one-to-one correspondence among the nontrivial zeta zeros and the
zero inner products.
Reminder about Theta series

• If one had used the states based on the Gauss-Jacobi theta series
∞
2tl
X
−s+k
ψs (t) = t e V (t)
, e 2V (t)
= e −πn
n=−∞

one would have had to regularize the ill defined inner products by
removing the zero modes in eqs-(10).
• In doing so, it would invalidate the transformation property of
the Gauss-Jacobi theta series G (x) of eq-(8) under x → 1/x, and
such that eqs. (9) will no longer be satisfied due to the eventual
loss of CT invariance.
• Therefore we do have a blessing in disguise when we had to
dismiss and bypass these states based on the Gauss-Jacobi theta
series because otherwise
Remarks on Inner Products

• When s1 , s2 reside in the critical line, the inner products

1 1
< Ψj,m
s1 (t) | Ψj,m
s2 (t) > ∼ P j ( +i(ρ2 −ρ1 )) ζ( +i(ρ2 −ρ1 )−2m)
2 2
(38)
1
are zero when Pj ( 2 + i(ρ2 − ρ1 )) = 0, since the polynomial Pj (s)
has simple zeros on the critical line.
• Are inner products positive definite ?
The norm of these states is not null
1 1
< Ψj,m
s1 (t)|Ψ j,m
s1 (t) > = < Ψj,m
s2 (t)|Ψj,m
s2 (t) > ∼ P j ( ) ζ( −2m) 6= 0
2 2
(39)
1
since Pj ( 2 ) 6= 0 when j = even.
• This is very relevant if one wishes the states to belong to
Banach, Hilbert spaces. Thus, by setting l = −2 in eqs-(15,26) it
will impose constraints on the values of j, m in order to have
positive-definite norms of Ψj,m 1
s1 (t) when s = 2 + iλ.
• The trivial zeta zeros ζ(−2n) = 0 for n = 1, 2, 3, · · · appear
when one takes the inner products
< Ψj,m j,m
s (t)|Ψ 1 −s (t) > ∼ ζ(−2m) = 0, when s is real valued.
2
Relation to Other Work
The connection of our work with the Berry-Keating Hamiltonian
and the recent work of Bender, Brody and Muller (BBM) . The
Berry-Keating Hamiltonian

xp + px d 1 d 1
HBK = = −i(x + ) = −i( + ) (40)
2 dx 2 dlnx 2
up to an overall numerical factor, is a very special case of the
operator D1 in eq-(1) for a simple choice of the potential, and
after identifying x with t.
The momentum operator p on the half-line is not Hermitian but
the combination (symmetric operator) xp + px defined by
Berry-Keating is Hermitian, in the interval (0, ∞).
The PT symmetric Hamiltonian studied by BBM was defined in
terms of a similarity transformation

HBBM = ∆ HBK ∆−1

where the action of the operators ∆−1 , ∆ are

∆−1 f (x) = (1 − e −ip )f (x) = f (x) − f (x − 1)

∞
−ip −1 −1
X (−ip)n
∆f (x) = [(1 − e )] f (x) = (−ip) Bn f (x)
n!
n=0

where Bn are the Bernoulli numbers.

Since the eigenfunctions of HBK are x −z with eigenvalues i(z − 21 ),
it follows that the eigenfunctions of HBBM
P∞ are then −zgiven in terms
of the Hurwitz zeta function ψz (x) = n=1 (x + n) , with the
same eigenvalues i(z − 12 ), and such that ψz (x = 0) = ζ(z).
References
B. Riemann, On the number of prime numbers less than a given
quantity, Monatsberichte der Berliner Akademie, November, 1859.
(Translated by D. R. Wilkins, 1998)
A. A. Karatsuba, S. M. Voronin, The Riemann zeta function.
[Translated from the Russian by Neal Koblitz ] (Berlin-New York,
Walter de Gruyter Pub., 1992);
S. J. Patterson, An introduction to the theory of the Riemann zeta
function (Cambridge, Univ. Press, 1988);
H. M.Edwards, Riemann’s Zeta Function ( New York, Dover Pub.,
2001);
E. C. Titchmarsh, The theory of the Riemann zeta-function
(Oxford, Clarendon Press, 1986).
M. Berry and J. Keating, SIAM Review, 41, no. 2 (1999) 236.
M. V. Berry and Z. V. Lewis, Proc. Roy. Soc. Lond. A 370
(1980) 459.
M. Berry, ” H = xp and the Riemann zeros” in Supersymmetry
and Trace Formulae ” Chaos and Disorder , ed. J. Keating, D.
Khmeltnitskii and I. Lerner, Kluwer 1999.
M. Coffey, Theta and Riemann ξ function representations from
Harmonic Oscillator eigensolutions, Phys. Letts A 362 (2007)
352; Special functions and the Mellin transforms of Laguere and
Hermite functions, [arXiv : math-ph/0612085].
C. Bender, D. Brody and M. Muller, “Hamiltonians for the zeros of
the Riemann Zeta Function”, Phys. Rev. Lett, 118 (2017)
130201.
R. Aschhheim, C. Castro and K. Irwin, “The search for a
Hamiltonian whose Energy Spectrum coincides with the Riemann
Zeta Zeros” International Journal of Geometric Methods in
Modern Physics Vol. 14, No. 6 (2017) 1750109.
C. Castro, Int. J. Geom Methods in Mod Phys 3, no.2 (2006)
D. Schumayer and D. A. W. Hutchinson, “Physics of the Riemann
Hypothesis” , Rev. Mod. Phys 83 (2011) 307 [arXiv: 1101.3116].
H. Wu and D. W. L. Sprung, Phys. Rev. E 48 (1993) 2595.
A. Connes, Selecta Mathematica (New Series) 5 (1999) 29.
G. Sierra, “The Riemann zeros as spectrum and the Riemann
hypothesis” arXiv : 1601.01797.
G. Sierra and P. Townsend, ”Landau Levels and Riemann Zeros”
arXiv : 0805.4079 [ math-ph] .
G. Sierra, xp with interactions and the Riemann zeros, [arXiv:
math-ph/0702034].
G. Sierra, New Journal of Physics 10 (2008) 033016.
A. LeClair, “Riemann Hypothesis and Random Walks: the Zeta
case”, arXiv : 1601.00914.
G. Franca and A. LeClair, “A theory for the zeros of Riemann Zeta
and other L-functions” areXiv : 1407.4358.