## Abstract

The interest in the Klein–Gordon equation with different potentials has increased in recent years due to its possible applications in Cosmology, Hadron Physics and High-Energy Physics. In this work, we investigate the solutions of the Klein–Gordon equation for bosons under the influence of an external potential by using the Feshbach–Villars method. We present detailed results for two cases: the Coulombic potential and the harmonic potential. For the latter case, we studied the effects of self-interacting particles by adopting a mean-field approach. We show that our results converge smoothly to the solution of the Schrödinger equation for the same systems as the relativistic effects diminish.

## 1 Introduction

The Klein–Gordon (KG) equation has been investigated by various methods due to its application in different fields. The case of a scalar field has an increasing interest because of its connections with the axion, a hypothesized particle that may have a relevant role in Astrophysics and in studies of Dark-Matter [1,2,3,4].

In Astrophysics, the investigation of a possible bosonic phase in massive stars has triggered the study of the solutions of the relativistic equation with an external potential. The possibility of a condensate regime inside the stars has been considered. Several methods have been used to solve the equation for different potentials [5, 6]. In the case of self-interacting particles, the problem becomes particularly difficult, and in Ref. [7], the authors claim that it is not possible to find a solution for this case. Nevertheless, they investigated the axion star and the possibility of the formation of a condensate.

Systems formed by interacting bosons are also of interest in High-Energy Physics and in Hadron Physics, where the possibility of a gluonic field in the condensate regime has been considered [8,9,10]. With the advent of colliders such as the EIC, FAIR and NICA, with the capability to investigate the hadron structure in detail, we can expect important advances in our understanding of the complex hadronic structure. The effects of the interaction of the components of the system with an external potential have been extensively studied in connection with the Bose–Einstein condensate [11, 12]. The interaction can favor or disfavor the formation of the condensate. The correlations and interactions among the particles do not prevent the formation of the condensate [13] and can lead to non-extensive forms of the condensate described in terms of the Tsallis statistics [14, 15]. The characteristics of the non-extensive condensate have been described in Ref. [16, 17].

In this work, we investigate the solutions of the KG equation for a system of bosons under the influence of an external potential, in particular for the Coulombic potential and for the harmonic potential, and then, we include the self-interaction through a mean-field approach. We investigate the ground-state solution and obtain the chemical potential for the case of the external harmonic potential, and with the self-interaction regulated by the scattering length. We use the Feshbach–Villars formalism, which allows us to evaluate the transition from the relativistic to the non-relativistic regimes. This is particularly useful in the case of self-interacting particles in a harmonic potential, for which the Schrödinger equation is known [18, 19]. We show that our results with the KG equation smoothly approach those for the Schrödinger equation as the systems become non-relativistic.

## 2 The Feshbach–Villars formalism

The KG equation for free particles is

This equation can be written in a convenient form by using the Feshbach–Villars formalism [20] (see also Ref. [21]). In this case, we define the spinor

where \(\varphi (t,\mathbf {r})\) and \(\chi (t,\mathbf {r})\) are wave-functions related to \(\psi (t,\mathbf {r})\) by

In the Feshbach–Villars formalism, the KG equation is written for the spinor as a Schrödinger-like equation, i.e.,

where

In the equation above \(\sigma _i \; (i=1,2,3)\) are the Pauli matrices, and \(\hat{\mathbf {p}} = -i \hbar \mathbf {\nabla }\). The conserved charge and current are given by

respectively. The spinor representation is equivalent to the coupled differential equations

The Feshbach–Villars formalism becomes especially convenient to study the non-relativistic regime, in which the relativistic wave equation reduces to the solution of the Schrödinger equation. We will study below some solutions of the KG equation for different potentials.

The time independent form of the KG equation is

and the corresponding solution of the time-dependent equation, Eq. (4), is

## 3 Introduction of the chemical potential

In the present work, we are interested in the investigation of the Bose–Einstein condensate formed in a relativistic system. We include the chemical potential, by assuming that the ground-state energy coincides with the chemical potential. The stationary-state solution is

where ± corresponds to the positive(negative) chemical potential solution. Then, the original KG equation, Eq. (1), leads to a time independent KG equation which writes

whose solution in the spinor representation is

where the spinor

has the amplitudes

Notice that \([\varphi _0^\pm (\mathbf {p})]^2 - [\chi _0^\pm (\mathbf {p})]^2 = \pm 1\), and \(\mu \equiv E_p = \sqrt{p^2 c^2 + m^2 c^4}\).

## 4 Introduction of an external potential

The Bose–Einstein condensation can happen even in the presence of an external potential. To study the effects of an external potential, let us solve the time independent KG equation

in different cases. This equation has been obtained and studied in the literature in several contexts, see, e.g., Refs. [22,23,24,25,26,27,28,29,30]. It can be derived by applying the correspondence principle in quantum mechanics: \(E \rightarrow i\hbar \frac{\partial }{\partial t}\), \(\mathbf {p} \rightarrow -i\hbar \mathbf {\nabla }\); to the relativistic expression for the energy \(E = \sqrt{ \mathbf {p}^2 c^2 + m^2 c^4} + V(\mathbf {r})\). Using the Feshbach–Villars formalism, we obtain the relation

with \(\psi = \varphi + \chi \). Then, we find from Eq. (16) that the spinor components \(\varphi \) and \(\chi \) can be computed from the wave function \(\psi \) as follows:

In the massless case, one has to consider \(m=0\) in Eq. (15), and replace \(m \rightarrow \mathcal N\) in Eqs. (16) and (17), where \(\mathcal N\) is an arbitrary real nonzero parameter [31].

The Schrödinger equation can be obtained from the KG equation (15) by considering the non-relativistic limit, i.e., \(c \rightarrow \infty \). To see that, let us define the energy with the rest-mass energy subtracted, \(\bar{\mu }\), as:

Then, we can write the KG equation (15) in a form similar to the Schrödinger equation, but with an explicit relativistic correction. This is

which is completely equivalent to Eq. (15). Then, the Schrödinger equation is obtained from the KG equation in the limit \(|V(\mathbf {r}) - \bar{\mu }| \ll m c^2\). Notice that in this limit one has from Eq. (17) that \(\varphi (\mathbf {r}) \simeq _{{c \rightarrow \infty }} \psi (\mathbf {r})\) and \(\chi (\mathbf {r}) \simeq _{{c \rightarrow \infty }} 0\).

### 4.1 Coulombic potential

Let us study the case of a Coulombic potential

We consider Eq. (15) in spherical coordinates, and assume \(\ell = 0\) (zero angular momentum), obtaining the equation

This equation can be written in a simpler form by introducing the reduced radial wave function:

Then one gets

The solution of Eq. (23) with the potential of Eq. (20) can be obtained analytically as a linear combination of two independent solutions, i.e.,

where \(L_{n}^a(z)\) and *U*(*a*, *b*, *z*) are the generalized Laguerre polynomial and the confluent hypergeometric function, respectively, while we have defined \(\Delta \equiv \sqrt{m^2 c^4 - \mu ^2}\) and

The solution at short distances behaves as:

with \(+(-)\) standing for the dominant behavior of the \(L_n^a(z)(U(a,b,z))\) solution. The physical (regular) solution corresponds to \(r^{\beta _+}\), so that the coefficient \(C_-\) should vanish and then we should consider \(C_2=0\) in Eq. (24). From the behavior of the *L*-solution at large distances, one finds that the square-integrable condition for the wave function leads to the exact result for the eigenvalue problem of Eq. (23) which we will denote by \(\mu _n^{{\text {KG}}}\), corresponding to the solutions of the equation:

leading to the chemical potential

Notice that the form of Eq. (28) implies that \(|\mu _n^{{\text {KG}}}| < mc^2\), so that the eigenfunction associated with the eigenvalue \(\mu _n^{{\text {KG}}}\) behaves as \(u(r) \simeq _{{r \rightarrow \infty }} e^{-\kappa r}\) with \(\kappa \equiv \Delta _n/(\hbar c) > 0\), thus corresponding to a bound state. To consider the non-relativistic limit, we make *c* to be large enough. Then, we have

We can compare our results in the non-relativistic limit with the results obtained with the Schrödinger equation. In this latter case, the eigenvalues are

Let us define \(\delta \mu _n\) as the difference between the result for the eigenvalue obtained with the KG equation (after subtracting the rest-mass energy), and the non-relativistic result of Eq. (30), i.e.,

In the limit of \(c \gg 1\), the expression of Eq. (28) has the following expansion:

so that one can easily identify the correction \(\delta \mu _n\) at order \({\mathscr {O}}(c^{-2})\) from a comparison with Eq. (31). Before displaying the results, it is convenient to define the dimensionless variables

Then, the dimensionless eigenvalues

are functions only of the parameter \(\lambda \). In particular, the decomposition of Eq. (31) writes in dimensionless units as:

The non-relativistic limit corresponds to \(\lambda \rightarrow 0\), or equivalently \(c \rightarrow \infty \). We display in Fig. 1 (left and middle panels) the dimensionless reduced wave function \(\bar{u}(\bar{r})\), as well as the spinor components \(\bar{u}_\varphi (\bar{r}) \equiv \sqrt{\frac{\hbar c}{mc^2} } \cdot r \varphi (r)\) and \(\bar{u}_\chi (\bar{r}) \equiv \sqrt{\frac{\hbar c}{mc^2} } \cdot r \chi (r)\). We display also in Fig. 1 (right) the behavior of \(\delta M_n\) as a function of \(\lambda \). Notice that , as it is already clear from Eq. (35), showing that the result of the Schrödinger equation is recovered in the non-relativistic limit.

Finally, let us note that the eigenvalue \(\mu _n^{{\text {KG}}}\) is real only for real values of \(\beta _{+}\), and this happens for

i.e., for \(\lambda < 1/2\). One can see that in the regime \(g > \hbar c/2\), the asymptotic behaviors of the reduced wave function are

so that while it is a square-integrable function, it is not a regular solution at \(r \rightarrow 0\).

### 4.2 System of *N* interacting particles

Another interesting case that we will study is a system of *N* interacting particles under the influence of an external potential. In the non-relativistic limit, such a system is usually described in terms of the Gross–Pitaevskii nonlinear Schrödinger equation, which is given by

In Ref. [18] (see also [19]), this system was studied with an harmonic external potential and a mean-field potential for the interaction term given in terms of a scattering-length, *a*, that is,

Here, we study this case in the relativistic regime and evaluate its non-relativistic limit comparing our results with the results from Ref. [18]. In the relativistic case, from Eq. (23) and using the potential of Eq. (39), one has

Let us define the dimensionless variables

Using Eq. (18) one finds \(\beta = 1/\alpha + \bar{\beta }\). Then, Eq. (40) can be written in the equivalent form:

The non-relativistic limit corresponds to \(\alpha \rightarrow 0\), or equivalently \(c \rightarrow \infty \). In the non-relativistic case, we define the normalization as \(N = \int d^3r |\psi (r)|^2 = 4\pi \int _0^\infty dr \, r^2 |u(r)|^2\) where *N* is the number of atoms or partons. The corresponding normalization for the dimensionless wave function is

The length \(\ell _s\) is the characteristic length of the system. In the relativistic case, it is better to define the normalization in terms of the conserved charge, i.e.,

which for the dimensionless spinor components of the wave function \(\Phi _{f}(x) \equiv \sqrt{8\pi |a|} \cdot r f(r) \,, \; f = \varphi , \chi \), leads to

Obviously \(\displaystyle Q \mathop {\longrightarrow }_{c \rightarrow \infty } N\) and \(\displaystyle q \mathop {\longrightarrow }_{c \rightarrow \infty } n\). We show in Fig. 2 the results of \(\bar{\beta }\) and *q* by solving the Gross–Pitaevskii nonlinear equation in the relativistic case, Eq. (42), and compared with the non-relativistic result \((\alpha =0)\) of Ref. [18]. Even though the curves of Fig. 2 contain all the relevant information, it would be interesting to show explicitly the dependence of \(\bar{\beta }\) with the scattering length *a*, by using that \(\bar{\beta }(q) = \bar{\beta }\left( Q |a|/\ell _s\right) \). To do so, we display in Fig. 3 the results of \(\bar{\beta }\) as a function of *Q* for several values of \(|a|/\ell _s\) (left panel), and as a function of \(|a|/\ell _s\) for several values of *Q* (middle panel). One can see that for fixed \(\bar{\beta }\), the larger the value of |*a*|, the smaller the value of *Q*. This can be easily understood by noticing that the relation \(\bar{\beta }= \bar{\beta }(q)\) given by Fig. 2 (middle) can be inverted to give the one-valued function \(q = q(\bar{\beta })\), so that using Eq. (45) one finds \(Q = \ell _s q(\bar{\beta })/|a|\). Notice that the normalization of \(\bar{\beta }\) given by Eq. (41) does not contain any dependence in *a*, so that the dependence of \(\bar{\mu }\) with *a* is proportional to the dependence of \(\bar{\beta }\) with *a*, i.e., \(\bar{\mu }(a) = \hbar \omega \cdot \bar{\beta }\left( Q |a|/\ell _s \right) \). Finally, we display in Fig. 3 (right) the dependence of *N*/*Q* with the parameter \(\alpha \) for \(\Phi ^\prime (0) = 2.5\), where *N* is computed in the relativistic case by using Eq. (43) with \(\Phi (x) = \Phi _\varphi (x) + \Phi _\chi (x)\) the solution of Eq. (42). Notice that in the non-relativistic limit \((\alpha \rightarrow 0)\) both quantities are equal, but in the relativistic regime one finds \(N < Q\).

## 5 Discussion and conclusions

Concluding, in this work we analyzed the time independent solutions of the nonlinear Klein–Gordon equation. We used the Feshbach–Villars approach with potential and analyzed the behavior of the system under different external potentials, namely the Coulombic potential and the harmonic potential.

With the introduction of the chemical potential as the eigenvalue of the Hamiltonian operator for the ground-state, we obtained the relevant conditions for the formation of the Bose–Einstein condensate in a relativistic system. The use of the Feshbach–Villars formalism allowed us to easily investigate the transition of the system from the relativistic to the non-relativistic regimes. We compare the results obtained for the relativistic system with a mean-field potential and an harmonic external potential, with those for the non-relativistic system under the same conditions. We observe a smooth transition of several parameters as the relativistic effects diminish, for both the Coulombic potential and the harmonic potential.

The effects of the interaction between the particles of the system, described by the scattering length in a mean-field approach, are analyzed, and in particular, we study the variation of the chemical potential with the scattering length.

The study of the nonlinear Klein–Gordon equation to investigate condensation of relativistic particles invites us to pursue extensions of this work. One possibility is to use the same formalism of the present work to investigate condensation in the relativistic hadronic matter, and another option is to study a modified form of the Gross–Pitaevskii equation with fractional derivatives.

## Data Availability Statement

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

## References

C.J. Hogan, M.J. Rees, Axion miniclusters. Phys. Lett. B

**205**(2–3), 228–230 (1988)R. Friedberg, T.D. Lee, Y. Pang, Scalar soliton stars and black holes. Phys. Rev. D

**35**(12), 3658 (1987)J. Eby, M. Leembruggen, L. Street, P. Suranyi, L. Wijewardhana, Global view of QCD axion stars. Phys. Rev. D

**100**(6), 063002 (2019)J. Eby, M. Leembruggen, P. Suranyi, L. Wijewardhana, Stability of condensed fuzzy dark matter halos. J. Cosmol. Astropart. Phys.

**2018**(10), 058 (2018)A.M. Grundland, E. Infeld, A family of nonlinear Klein-Gordon equations and their solutions. J. Math. Phys.

**33**(7), 2498–2503 (1992)J. Eby, M. Leembruggen, L. Street, P. Suranyi, L. Wijewardhana, Approximation methods in the study of boson stars. Phys. Rev. D

**98**(12), 123013 (2018)J. Eby, P. Suranyi, L. Wijewardhana, Expansion in higher harmonics of boson stars using a generalized Ruffini-Bonazzola approach. Part 1. Bound states. J. Cosmol. Astropart. Phys.

**2018**(04), 038 (2018)A.H. Rezaeian, Semi-inclusive photon-hadron production in pp and pA collisions at RHIC and LHC. Phys. Rev. D

**86**(9), 094016 (2012)J. Jankowski, D. Blaschke, M. Spaliński, Chiral condensate in hadronic matter. Phys. Rev. D

**87**(10), 105018 (2013)H. Fujii, Y. Tsue, Quark condensate in a medium and pion-nucleon sigma term based on a QCD-like theory. Phys. Lett. B

**357**(1–2), 199–203 (1995)M.O. Mewes, M.R. Andrews, N.J. van Druten, D.M. Kurn, D.S. Durfee, C.G. Townsend, W. Ketterle, Collective excitations of a Bose-Einstein condensate in a magnetic trap. Phys. Rev. Lett.

**77**(6), 988–991 (1996). https://doi.org/10.1103/PhysRevLett.77.988V. Bagnato, D. Kleppner, Bose-Einstein condensation in an external potential. Phys. Rev. A

**35**(10), 4354–4358 (1987). https://doi.org/10.1103/PhysRevA.35.4354O. Savchuk, Y. Bondar, O. Stashko, R.V. Poberezhnyuk, V. Vovchenko, M.I. Gorenstein, H. Stoecker, Bose-Einstein condensation phenomenology in systems with repulsive interactions. Phys. Rev. C

**102**, 035202 (2020). https://doi.org/10.1103/PhysRevC.102.035202A.R. Plastino, C. Tsallis, Dissipative effects in nonlinear Klein-Gordon dynamics. EPL (Europhysics Letters)

**113**(5), 50005 (2016). https://doi.org/10.1209/0295-5075/113/50005L.E. Guerrero, J.A. Gonzalez, in Long-range interacting solitons: pattern formation and nonextensive thermostatistics, Physica A 257 (1–4), 390–394 (1998) , vth Latin American Workshop on Non-Linear Phenomena/11th MEDYFINOL Conference on Statistical Physics of Dynamic and Complex Systems, Canela, Brazil, Sep 28–Oct 03, 1997. https://doi.org/10.1016/S0378-4371(98)00165-4

E. Megías, V. Timóteo, A. Gammal, A. Deppman, Bose-Einstein condensation and non-extensive statistics for finite systems. Phys. A Stat. Mech. Appl.

**585**, 126440 (2022). https://doi.org/10.1016/j.physa.2021.126440A.K. Rajagopal, R.S. Mendes, E.K. Lenzi, Quantum statistical mechanics for nonextensive systems: prediction for possible experimental tests. Phys. Rev. Lett.

**80**(18), 3907–3910 (1998)A. Gammal, T. Frederico, L. Tomio, Improved numerical approach for the time-independent Gross-Pitaevskii nonlinear Schrödinger equation. Phys. Rev. E

**60**(2), 2421 (1999)S.K. Adhikari, Numerical study of the spherically symmetric Gross-Pitaevskii equation in two space dimensions. Phys. Rev. E

**62**(2), 2937–2944 (2000)H. Feshbach, F. Villars, Elementary relativistic wave mechanics of spin 0 and spin 1/2 particles. Rev. Mod. Phys.

**30**(1), 24 (1958)P. Alberto, S. Das, E.C. Vagenas, Relativistic particle in a box: Klein-Gordon versus Dirac equations. Eur. J. Phys.

**39**(2), 025401 (2018)F. Dominguez-Adame, Bound states of the Klein-Gordon equation with vector and scalar Hulthen type potentials. Phys. Lett. A

**136**, 175–177 (1989). https://doi.org/10.1016/0375-9601(89)90555-0G. Chen, Z.D. Chen, Bound stated solutions of the Klein-Gordon equation with Hartmann potential and recursion relations. Acta Phys. Sin.

**54**(6), 2524–2527 (2005)X.A. Zhang, K. Chen, Z.L. Duan, Bound States of the Klein-Gordon equation and Dirac equation for ring-shaped non-spherical oscillator scalar and vector potentials. Chin. Phys.

**14**(1), 42–44 (2005)C.Y. Chen, Exact solutions of the Dirac equation with scalar and vector Hartmann potentials. Phys. Lett. A

**339**, 283–287 (2005)C.Y. Chen, F.L. Lu, D.S. Sun, The relativistic bound states of Coulomb potential plus a new ring-shaped potential. Acta Phys. Sin.

**55**(8), 3875–3879 (2006)M.C. Zhang, Z.B. Wang, Bound states of the Klein-Gordon equation and Dirac equation with the Manning-Rosen scalar and vector potentials. Acta Phys. Sin.

**55**(2), 521–524 (2006)R.J. Ducharme, Exact solution of the Klein-Gordon equation for the hydrogen atom including electron spin (2010). http://arxiv.org/abs/1006.3971

L.K. Sharma, P.V. Luhanga, S. Chimidza, Potentials for the Klein-Gordon and Dirac equations. Chiang Mai J. Sci.

**38**(4), 514–526 (2011)G. Bi, Y. Bi, New relativistic wave equations for two-particle systems (4 2012). http://arxiv.org/abs/1204.4876

A.J. Silenko, Zitterbewegung of bosons. Phys. Part. Nucl. Lett.

**17**(2), 116–119 (2020). https://doi.org/10.1134/S1547477120020193

## Acknowledgements

The work of E M is supported by the project PID2020-114767GB-I00 financed by MCIN/AEI/10.13039/501100011033, by the FEDER/Junta de Andalucía-Consejería de Economía y Conocimiento 2014-2020 Operational Program under Grant A-FQM-178-UGR18, by Junta de Andalucía under Grant FQM-225, and by the Consejería de Conocimiento, Investigación y Universidad of the Junta de Andalucía and European Regional Development Fund (ERDF) under Grant SOMM17/6105/UGR. The research of E M is also supported by the Ramón y Cajal Program of the Spanish MCIN under Grant RYC-2016-20678. M J T is supported by FAPESP grant 2021/12954-5. V S T is supported by FAEPEX (grant 3258/19), FAPESP (grant 2019/010889-1) and CNPq (grant 306615/2018-5). A D is partially supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq-Brazil), grant 304244/2018-0, by Project INCT-FNA Proc. No. 464 898/2014-5 and supported by FAPESP grant 2016/17612-7.

### Open Access

This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

## Funding

Funding for open access charge: Universidad de Granada / CBUA.

## Author information

### Authors and Affiliations

### Corresponding author

## Rights and permissions

**Open Access** This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

## About this article

### Cite this article

Megías, E., Teixeira, M.J., Timoteo, V.S. *et al.* Nonlinear Klein–Gordon equation and the Bose–Einstein condensation.
*Eur. Phys. J. Plus* **137**, 325 (2022). https://doi.org/10.1140/epjp/s13360-022-02511-2

Received:

Accepted:

Published:

DOI: https://doi.org/10.1140/epjp/s13360-022-02511-2