**Journal of Modern Physics
**Vol.5 No.14(2014), Article ID:49260,9 pages
DOI:10.4236/jmp.2014.514135

Brief Note on a Scalar Quantum Field with Finite Lifetime in a Lorentz Invariant Non-Rectangular Euclidean Space

Carmen Tornow

Institute of Planetary Research, DLR, Berlin, Germany

Email: carmen.tornow@dlr.de

Copyright © 2014 by author and Scientific Research Publishing Inc.

This work is licensed under the Creative Commons Attribution International License (CC BY).

http://creativecommons.org/licenses/by/4.0/

Received 17 April 2014; revised 14 May 2014; accepted 11 June 2014

ABSTRACT

A not necessary rectangular Euclidean space (NoNRES) is constructed, in which one obtains a generally Lorentz invariant scalar product for the low energy sector (LES). This sector is defined for energies below the Planckian limit. If the energy is zero, the NoNRES becomes rectangular and due to the Lorentz invariance, it is applicable for the complete LES of the theory. In contrast to the usual Minkowski space the metric of the NoNRES depends on the kinetic energy of the observed quantum particles. It is assumed that this metric may be useful to derive the scattering cross-section of the corresponding quantum field theory. This assumption is related to the occurrence of divergent loop momentum integrals caused by including the infinite energy range above the Planckian limit (high energy sector or HES). Due to its energy dependence, the metric in both energy sectors differs. In the HES, it depends on the effective dimension of the NoNRES. This dependency results from fluctuations of the space-time above the Planckian limit. Even if they are not part of the theory (as they would be in quantum gravity), these fluctuations should not be ignored. The effective dimension decreases if the energy of the considered particle increases. Since this is true for the HES only, the ultraviolet divergences of loop integrals seem to vanish without distorting the results of the LES. The mechanism is illustrated by calculating the tadpole integral occurring for a simple self-interacting scalar quantum field (with the Higgs mass as example). One obtains a finite contribution for the integral and consequently for the lifetime of the scalar particle.

**Keywords:**Lorentz Invariance, Scalar Field, Renormalisation, Lifetime

1. Introduction

Ultraviolet (UV) divergences in quantum field theories (QFTs) without regularization
and renormalization occur because of short-distance interactions between the elementary
particles. These interactions are considered in the high energy range of loop-momentum
integrals. A well-known example for this behavior is the one loop-momentum integral
of the scalar QFT with a j^{4} potential term to simulate self-interacting.
Although, this theory must be seen as an artificial test model, we know that a scalar
field is related to the Higgs boson [1] [2] . If one assumes a very small coupling constant for the
j^{4} potential, the dressed propagator D(x_{1}, x_{2})
is given by the one-loop approximation and consists of the free field propagator
D(x_{1}, x_{2}), and the tadpole diagram (Figure
1).

An analysis of the momentum integral corresponding to the tadpole diagram shows
that its divergence degree is +2. In order to calculate the corrected mass and lifetime
of the self-interacting scalar field a regularization and renormalization procedure
is needed. In general, the divergences in loop-momentum integrals originate from
the fact, that special QFTs, e.g. quantum electrodynamics, are low energy approximations
of a complete theory of all possible quantum particle interactions, including gravity
[3] . In this study it is asked whether there is
chance to simplify the regularization and renormalization procedure. As a first
step to seek an answer to this question, the upper dressed propagation operator
defined in Figure 1 is computed with a modified
method. This method requires a split of the integration range using the Planckian
mass,
where E_{P} = 1.22 × 10^{19} GeV is the Planck
energy. The related split value is. Accordingly, one gets for the spatial frequency
k = 2p/l or if we consider the particle masses of real and virtual particles k =
2pl where l is the inverse Compton length l = m_{0}×c/ħ• a low
energy sector (LES) with, i.e. Lorentz invariance of measurable
(scalar) values, and

• a high energy sector (HES) with, i.e. Lorentz symmetry is broken.

Actually, between the energy sector of standard QFT and the HES an appropriate transition
region might exist, since it is speculated that the effects related to the violation
of Lorentz invariance (e.g. polarization independent refractivity
[4] ) can already be detected at energies much lower than E_{P}
= m_{P}c^{2}. Nonetheless, it will be shown in section 3 that for
the simple scalar quantum field a sharp separation works well.

Because an interval split is useless if one integrates over both energy intervals with the same method, an integration measure needs to be derived that differs between LES and HES. This measure shall reflect the lower retrievable information content of the HES compared to the LES. The next two sections explain how this measure is derived and implemented mathematically. In Section 4, the modified method of computing loop integrals is applied to the tadpole integral illustrated in Figure 1. Lastly, the applicability of this integration method is discussed and further steps to establish this technique are outlined.

2. Invariance of Lorentz Transformation in the Non-Rectangular Euclidean Space

It is assumed that the integration measure directly related to the geometry of Minkowski space (MS) may not be appropriate for the calculation of the high energy range of momentum integrals. Therefore, an alternative, Lorentz invariant, geometry is constructed. It turns out that a non-rectangular space with Euclidean signature and an energy dependent angle between its space and time axes can be used. This non-rectangularity is useful to obtain an energy dependent integration measure. In this space, which is abbreviated by NoNRES (see section 3), one has general transformations given by

(1)

between an inertial system S moving with the velocity v = V_{j}e_{j}
and an inertial system S‘ moving with a velocity. The index j is arbitrary
but fixed. Further one sets,
,
for k ¹ j. As for the MS the scalar product
of vector
is Lorentz invariant, i.e.,

(2)

Figure 1. Two Feynman diagrams (right-hand side) giving the first order dressed propagator (left-hand side) of the weakly self-interacting scalar field, j.

where the matrix
describes the Lorentz transformation (LT) and g^{a}^{b} (g¢^{m}^{n})
is the metric tensor in the NoNRES. From Equation (2) we obtain

(3)

The parameter s_{j} depend on the velocity ratios
and
where j = 1, 2, 3 is arbitrary but fixed. If one assumes for s_{j} and w_{j}
a relation typical for the LT, all the non-zero elements of
can be written according to Equation (1) in the following way

(4)

All off-diagonal elements of the metric g^{0j}
in the NoNRES are related to the angle between the axis x_{j}
and x_{0}, which is influenced by the
ratio b_{j}. Based on these assumptions
one obtains for the elements of g^{m}^{n}^{}

(5)^{}

The equivalent approach follows for the metric g'^{m}^{n} in S'
using the coefficients
and. When S is an inertial system for a resting body
its metric g^{m}^{n} = d^{m}^{n} with d^{m}^{n}
= 1 if m = n and d^{m}^{n} = 0 if m ¹ n. An expression that
relates r_{j} and
to s_{j} can be derived inserting the Equations (4) and (5) into Equation
(3):

(6)

Because the absolute value occurs in Equation (4) there are two solutions for Equation (6)

(7)

We note that_{
} and. The relations
in Equation (7) correspond to the addition theorems of the velocity ratios b_{j}
and. To show this, we need to express
and
by b_{j}. It follows from Equation (5)
that r_{j} = 0 and l_{j} = 1 for a system at rest, S. Equation
(7) simplifies to

(8.1)

where the index 0 indicates that one inertial system, S, is at rest. The two plots in Figure 2 present the inverse function, i.e., the curves of and versus the parameters and. It is interesting to recognize the increase of on the left-hand plot and the decrease of on the right-hand plot.

This behaviour ensures that the metric given by Equation (5) is real. Now, S' is at rest and S is the inertial system for a moving body. We set, in Equation (7) and obtain

Figure 2. Influence of
velocity ratio_{
} and_{
}^{ }(abscissa) on metric
parameter
(solid line) and
(dashed line) for
and.

(8.2)

Inserting Equations (8.1) and (8.2) into the relation for^{
} and
in Equation (7) it can be shown that

(9)

(10)

which are indeed the addition theorems of the velocity ratios known from special
relativity in the MS, where parameters_{
} and_{
} are equal to b_{j} and, while for
and
_{ }one_{ }finds
and. Note, that matrix
is identical to the LT. In addition, timeand space-like regions can be identified
in the NoNRES, which are related to both of the solutions in Equation (7). The first
relation in Equations (7) and (9) is the addition theorem of velocities lower than
c and the second relation is the addition theorem of velocities higher than c. The
index j, that can have values between 1 and 3, gives the three matrices
in Equation (4) that form the Lorentz group in the NoNRES. In special relativity,
a fundamental parameter is the “Eigenzeit” following from the approach

(11)

It agrees with the “Eigenzeit” in the MS and verifies the identity between
and b_{j} (the plus index is suppressed)

(12)

Using this “Eigenzeit” the energy-mass relation can be derived. For the particle energy E we obtain

(13)

where the vector
is the four-velocity, and the vector p = m_{0}u is the four-momentum, if
the particle rest mass is m_{0}. The energy-mass relation in Equation (13)
is well-known from special relativity in the MS. Using this equation we express
the metric parameter l_{j} as

(14)

For a particle nearly at rest with E ~ m_{0}c^{2}, the NoNRES becomes
almost rectangular, i.e., l_{j} ~ 1 and r_{j} ~ 0. From Equation
(14) it is clear that the angle between the time and the relevant space axis in
S depends on the particle energy, just as required at the beginning of this section.
If the particle energy of the quantum field influences the metric attached to it,
the NoNRES cannot be globally an Euclidean space, R^{n}, where n = 4. It
seems rather to be a manifold, M^{n}, that is locally isomorphic to R^{n}.
The corresponding local positions in the NoNRES are infinitesimal small regions
in space and time given by the interaction events between the quantum fields.

In the following section we derive the effective dimension of the NoNRES and describe the consequences of the geometry for this number. The effective dimension is a key ingredient for the desired integration measures allowing a low-weighted consideration of the high energy sector.

3. The Effective Dimension of the NoNRES

Using Equation (2) we can transform any inertial system in the NoNRES into a rectangular
Euclidean space. However, doing so requires a continuous space-time with a well-defined
metric. For energies in the Planckian range space-time fluctuations occur [4] (which may cause the
polarization independent refractivity [5] ) and this condition is not fulfilled. Since Equation
(2) can be understood as a symmetry, where the related group is the Lorentz group,
the emergence of the non-continuous space-time structure leads to a violation of
Equation (2) and thus to a broken Lorentz symmetry. In the upper energy range of
the LES and in the HES fluctuations have to be expected, if a virtual (off-shell)
particle with an energy near the Planckian range interacts with a real (on-shell)
or another virtual particle. They cause uncertainties for the physical parameters
containing expressions like g^{m}^{n}k_{m}k_{n}
or with similar terms. Because of Equation (13) the value of r_{j} approaches
1 if the particle energy tends to infinity. Accordingly, the angle between the space
and time axis decreases to 0˚. Moreover, all the other spatial coordinate ranges
are correspondingly shrunk compared to the ranges of x_{0} and x_{j},
since the factor
^{ }tends to infinity. As a result, the
information on the dynamics of the physical parameters is overwhelmed by the metric
fluctuations. Due to a broken Lorentz symmetry caused by the non-continuous space-time,
the non-rectangularity of the NoNRES becomes manifest in the HES. In the LES it
appears to be a transformation effect which does not seriously destroy the accuracy
of the theory. Consequently all scalar values can be calculated for a 4-dimensional
rectangular space in the LES. In order to consider this geometrical behaviour of
our special type of space-time in its name, it is called not necessary rectangular
Euclidean space. The information loss for real high quality measurements (i.e. measurements
with finite but sufficient accuracy) can be quantified. An appropriate measure to
describe the mixing degree of quantum states is von Neumann’s entropy S_{N}
given by [6]

(15)

where j is the density matrix of an ensemble of quantum mechanical states q_{k}.
Assuming that the density matrix fulfils the eigenvalue equations: fu_{a}
= e_{a}u_{a}, we can simplify Equation (15) an write:

(16)

where e_{a} and (u_{k})_{a} are the eigenvalues and −vectors
of the d × d density matrix^{ }f. If the d eigenvalues {e_{a}} with
a^{ }=^{ }0,^{ }d − 1 are normalized as given in Equation
(16) they can be linked to the probabilities of the states q_{k}. Then,
the von Neumann entropy in Equation (16) agrees with Shannon’s entropy and the related
density matrix represents an ensemble of pure states. If one sets for_{ }
f the metric_{ }g of the NoNRES, Shannon’s entropy seems to be useful to
derive an effective dimension, dim(g), for the NoNRES even if g in Equation (5)
depends on classical observables and not on the equivalent quantum operators (compare
to Equation (14)). Using the entropy in Equation (16), one computes the effective
dimension dim(g) by the following expression

(17)

where the bar indicates normalization of the eigenvalues of the metric. The notation dim[g(E)] =: dim(E) is used to underline the energy dependency of the effective dimension which is caused by the energy dependency of the metric. If dim(E) = 4, the information content of the four vector components is measurable without any loss due to the finite accuracy of the measured data. If dim(E) < 4 uncertainties resulting from measurement errors can increase up to infinity if the smallest eigenvalue approaches zero [7] . According to Equation (5) the eigenvalues of the metric in their normalized form are

(18)

If r_{j} = 0, i.e., l_{j} =^{ }1, the eigenvalues are ē_{a}
= 1/4 for all a. Thus, the effective dimension dim(E) = 4 and the space-time corresponds
to an rectangular Euclidian space, i.e. a coordinate system for a resting particle.
For a moving particle one obtains r_{j} > 0, i.e., l_{j} < 1 leading
to dim(E) < 4. Inserting Equation (14) into Equation (18) gives

(18.1)

(18.2)

(18.3)

with
(note that the x-axis in Figure 3 presents q =
E/E_{0},) and max(Q) = min(q) = 1. For the usual light elementary particles
with masses equivalent to the GeV-energy scale, one obtains for the LES, i.e.
and therefore, ē_{0} ~ 1 and ē_{a}
~ 0 for a > 0. Next, one gets following expression for the entropy S_{N}
according to Equation (16)

(16.1)

or

(16.2)

In Figure 3 the shapes of the effective dimensions,
dim(E), versus the energy ratio,^{ }E/E_{0}, for three scalar particles
with different rest energies_{ }are shown. The

(17')

For the boundary region of LES one finds that
(note: lim_{Q}_{®}_{0}(Q^{Q})
= 1) which corresponds to
where
(see Figure 3). To facilitate a comparison of the
shape of dim(E) for a particle with Higgs mass (left plot of
Figure 3) with that of hypothetical super-heavy particles the other two
plots are presented. The soft decrease in the middle and right plot of Figure 3 results from the much higher masses of these hypothetical
particles.

In contrast to the shapes of dim(E) for the super-heavy particles, the information
loss calculated for the particle with Higgs mass reaches its limit dim(E) = 1 for
an increasing energy very rapidly. Nevertheless, a comparison of the magnitudes
of q in the three plots shows, that the different shapes become similar to each
other if the number of magnitudes would be the same in all plots of Figure 3. As a consequence, dim(E) differs between particles
of different mass while dim(q) is a universal parameter. If Lorentz symmetry exists
and the light speed is constant for the complete LES, the total energy of this sector
cannot be higher than the Planckian limit, m_{P}c^{2}. Note, that
in the left plot of Figure 3 the maximum energy
is max(E) ~ 7 ´ 10^{5} E_{H} (with E_{H} = m_{H}c^{2})
and much smaller than the maximum energy in the middle and right plots (with max(E)
~ E_{P}). For an off-shell Higgs particle the dimensional limit is reached
long before the quantum fluctuations of the metric become important, the Lorentz
symmetry is intact and the LT is applicable even if dim(E) » 1. For off-shell
super-heavy particles,

E_{0} = 1.26 × 10^{2} GeV E_{0} = 9.50 × 10^{16}
GeV E_{0} = 1.20 × 10^{18} GeV

Figure 3. Loss of information, given by the dimension dim(E), between two corresponding inertial
systems for a measuring device and a particle. This loss is shown for three particles
with different masses in individual plots. Since dim(E) depends on the particle
energy, an energy ratio, q = E/E_{0} is introduced. The rest energy, E_{0}
= m_{0}c^{2}, in the left plot results from the Higgs mass [1] , m_{0} = m_{H}. In the middle and right
plot, E_{0} corresponds to the mass of hypothetical super-heavy particles
close to the Planckian scale. The minimum energy E on the x-axis of each plot varies,
since the rest masses differ, but min(q) = 1. The upper boundary value of q is q^{*}
= E_{P}/E_{0} and varies in its value due to the different particle
masses..

especially for a particle with a mass m_{0} ~ m_{P}/10 (right plot
of Figure 3) the quantum fluctuations of the metric
have started already while dim(E) > 1 and the transformation between inertial systems
is not allowed anymore due to the broken Lorentz symmetry.

In the Introduction a transition region between the sector of the standard model is mentioned. For the simple scalar field this transition region needs not to be considered. However, the shapes of dim(E) for the super-heavy particles suggest the existence of some transition region.

4. The Interacting Scalar Particle in the NoNRES

Now, the derived integration measure is used for an interacting scalar quantum field,
j. We consider the 4-momentum p_{m} of a free scalar particle and calculate
the energy-momentum relation in the NoNRES

(19)

which leads to the Klein-Gordon operator À = ¶_{a}¶^{a}
+^{ }1 where the inverse Compton length l = m_{0}×c/ħ is used
to scale the field j ® j/l and the axes x_{m}
® lx_{m}. For a self-interacting scalar
field, j, the action integral A[j, J]
is given by

(20)

where the potential term is V[j]
= w/4!×j^{4}, J is some background field and dim is the dimension of the
particle related space-time. In contrast to the MS, the metric factor g^{1/2}
of the NoNRES, occurring in the action, gives 1. Accordingly, the path integral
can be written as exp(−A/ħ), unlike exp(−iA/ħ) in the
MS. The Green’s function, D(x), of a free scalar field with V[j] = 0 can be calculated from

(21)

where k_{n} ® k_{n}/l and ie is
a regularization term for the denominator if k^{2} = 1. The connected 2-point
Green’s function (compare to the dressed propagator in
Figure 1) for the weakly self-interacting field in the one-loop approximation
is given by

(22)

The Fourier transform of Equation (22) can be expressed as

(23)

Since the interaction given by Equation (20) is sufficiently weak one can write for Equation (23)

(24)

where the scaled corrected mass and lifetime of the weakly self-interacting field, j, result from

(25)

In the MS D(0) is UV divergent. However, in the NoNRES we obtain a finite expression
as shown below. Considering dim = dim(E) the integral for D(0) in Equation (23),
which depends on k^{2} only, is given by

(26)

where the integration range of k extends from^{ }null_{ }to infinity.
From this equation the sought integration measure, which is d^{dim}^{(E)}k,
can be read off. In the special case of a k^{2} -dependent integration function,
it simplifies to. The integration range is split into
the two energy sectors as explained in the Introduction. They go from the k = 0
to E_{p}/m_{0}c^{2} and from k = E_{p}/m_{0}c^{2}
to ¥. If the energy E^{ }< E_{P}
the physical space is modeled by a 4-dimensional rectangular Euclidean space due
to the intact Lorentz symmetry and the applicability of the corresponding LT. If
E ~ E_{P} the LT is not allowed anymore. However, dim(E) has reached the
limit dim(E) = 1 long before (see Figure 3), thus,
we can find a jump function for the effective dimension of the local NoNRES written
as

(27)

Consequently, we obtain for Equation (26)

(28)

Both terms are finite and give

(29)

Obviously, the effective dimension, dim(E), acts as a “regularization-like parameter” influencing the integration result in the high-energy range. However, in contrast to usual regularization procedures in the MS, dim(E) leads to a convergent result without specific normalization concepts.

Because of scaling we substitute for k^{*} the ratio of Planckian to particle
rest energy, i.e., k^{*} = E_{P}/m_{0}c^{2}. Since
the generator of the massive scalar field, j, is assumed to be the Higgs particle,
it results m_{0}c^{2} º m_{H}c^{2} = 126 GeV
[1] [2] , where m_{H} is
its mass and c is the light speed, as has already been used to calculate the shape
of dim(E) presented in Figure 3. Therefore, we
obtain for the corrected mass, m^{*}, and the un-scaled lifetime, t_{L},
derived from Equation (25), the following results

(30)

(31)

The coupling constant should be w £ 10/Js to guarantee the applicability of a first order approximation to re-arrange Equation (23). If the particle mass is not the estimated Higgs mass, but, e.g., the electron mass, we had to require a much lower upper boundary for w or Equation (24) would be invalid. The theory described by Equation (20) is true for massive scalar particles. If one likes to consider massless particles, e.g. photons, together with Higgs scalars, the standard model of the electro-weak and strong interaction needs to be slightly modified to ensure its Lorentz invariance in the NoNRES.

5. Conclusions

The metric, g^{m}^{n}, in the NoNRES, which depends on the particle
energy, E, allows to derivate an effective dimension, dim(E), which is the key parameter
of the integration measure d^{dim}^{(E)}k. In contrast to the NoNRES,
the effective dimension of the MS always gives dim_{MS}(E) = 4 for all values
of E, i.e., it is independent of the particle energy. The described integration-method
constructed for the NoNRES involves the high energy range of the specific QFT more
appropriately than the MS, since the available information content of the observed
process is considered and quantified. In addition, the NoNRES can be constructed
for space-times of arbitrary dimensions. The results obtained for Equations (30)
and (31) are quite reasonable. In contrast to the mass correction, the life time
of the particle is mainly influenced by the result of the high energy sector, i.e.,
the large imaginary part of D(0) in Equation (29).

In this study a self-interacting scalar particle is considered. In order to develop
a real QFT in the NoNRES we have to include particles with a larger coupling constant
(i.e. higher order loop integrals needs to be taken into account) and particles
with a spin larger than 0. Then, QED (quantum electrodynamics) can be studied to
model the interaction between electrons by means of photon exchange. In comparison
to Equation (20), the QED action integral is more complex and contains the Abelian
gauge field, A_{a}, and a spinor field, y. An extension of our integration
method to QED would open the access to real measurements.

Acknowledgements

The author would like to cordially thank Uwe Motschmann from the Institute for Theoretical Physics at the Technical University Braunschweig for many fruitful discussions.

References

- Aad, G., et al. [ATLAS Collaboration] (2012) Physics Letters B, 716, 1-29. http://dx.doi.org/10.1016/j.physletb.2012.08.020
- Chatrchyan, S., et al. [CMS Collaboration] (2012) Physics Letters B, 716, 30-61. http://dx.doi.org/10.1016/j.physletb.2012.08.02s1
- Delamotte, B. (2004) American Journal of Physics, 72, 170-184. http://dx.doi.org/10.1119/1.1624112
- Erdmenger, J., Meyer, R. and Park, J.-H. (2007) Physical Review Letters, 98, Article ID: 261301. http://dx.doi.org/10.1103/PhysRevLett.98.261301
- Gharibyan, V. (2012) Physical Review Letters, 109, Article ID: 141103. http://dx.doi.org/10.1103/PhysRevLett.109.141103
- Balian, R. (2003) Seminaire Poincare, 2, 13-27.
- Twomey, S. (2002) Introduction to the Mathematics of Inversion in Remote Sensing and Indirect Measurements, Courier Dover Publications.