This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

aainstitutetext: School of Physics and Electronics Science,
Hunan Provincial Key Laboratory of High-Energy Scale Physics and Applications,
Hunan University, Changsha 410082, P. R. China
bbinstitutetext: Lanzhou Center for Theoretical Physics, Key Laboratory of Theoretical Physics of Gansu Province, School of Physical Science and Technology, Lanzhou University, Lanzhou 730000, P. R. Chinaccinstitutetext: School of Physical Science and Technology, Southwest University, Chongqing 400715, P. R. China

Localization of matter fields on a chameleon brane

Yi Zhong c,1    and Ke Yang111Corresponding author. [email protected] [email protected]
Abstract

In this work, we address the localization problem of vector field in the chameleon braneworld and investigate the localization of various matter fields. The conditions for localizing the matter fields are determined. It is found that the zero modes of scalar, vector, and fermion fields can be successfully localized, yet the zero mode of Kalb-Ramond field cannot be localized, which implies that the recovery of standard model fields on the brane. Furthermore, the characteristics of quasi-localized modes of the qq-form fields are analyzed, and the parameter constraints of the model are estimated.

1 Introduction

The existence of hidden extra dimensions in our spacetime, particularly at ultraviolet scales, has been a topic of interest for a long time, starting from the proposal of the Kaluza-Klein (KK) theory in the 1920s. Based on this idea, the braneworld scenario was proposed, in which our universe could be a 3-brane embedded in a higher-dimensional space-time, called the bulk. This perspective offers a novel approach to addressing the gauge hierarchy problem and cosmological constant problem ArkaniHamed:1998rs ; ArkaniHamed:2000eg ; Randall:1999ee ; Randall:1999vf , which are two persistent challenges in particle physics and cosmology. The concept of a thick brane Gremm:2000dj ; Gass:1999gk ; Gremm:1999pj ; Afonso:2006gi extends the ideas of the Randall-Sundrum-2 (RS2) model Randall:1999vf , and the configuration of a domain wall Rubakov:1983bb ; Rubakov:1983bz serves as a topological defect in the bulk.

In the braneworld theory, the localization of matter fields is an important issue, as it not only allows the five-dimensional theory to reduce to a four-dimensional effective theory at low energy scales but also provides observable effects for probing the extra dimensions Bajc2000 ; Oda2000 ; Melfo:2006hh ; Liu:2009ve ; PhysRevD.79.125022 ; Liang2009 ; German:2013sk ; Vaquera-Araujo:2014tia ; Salvio:2007qx . In five-dimensional spacetime, free scalar fields can typically be localized on branes Bajc2000 ; Oda2000 , while fermion fields generally require Yukawa couplings with background scalar fields to achieve localization on branes Melfo:2006hh ; PhysRevD.78.065025 ; Liang2009 ; Alencar2014 ; Vaquera-Araujo:2014tia . However, it is generally challenging to localize vector field on branes in five dimensions. Some researchers have attempted to address this issue by assuming couplings between vector field and scalar field or spacetime geometry, or by considering models such as the Weyl geometry braneworld, de Sitter braneworld, and six-dimensional braneworld Chumbes:2011zt ; Alencar2014 ; Vaquera-Araujo:2014tia . See Refs. Dvali:2000rx ; Akhmedov:2001ny ; Guerrero:2009ac ; Delsate:2011aa ; Cruz:2012kd ; Dimopoulos:2000ej ; Belchior:2023gmr ; Zhao:2017epp for more works on the localization of vector field in braneworld models. Despite the extensive research conducted on this issue, there is still significance in providing a concise and natural new mechanism to effectively localize the vector field on the brane.

On the other hand, the chameleon gravity proposed by Khoury et al. in 2003 can address the dark energy problem PhysRevLett.93.171104 ; PhysRevD.69.044026 . This theory introduces a scalar field, known as the "chameleon" which exhibits mass variation depending on the surrounding matter density. The concept of chameleon gravity is closely related to string theory Brax2004 and scalar-tensor theory. The theory has been extensively tested through observations and experiments Burrage:2017qrf , showing promising prospects for resolving the Hubble constant problem PhysRevD.103.L121302 . In this theory, matter fields do not directly couple to the metric tensor gμνg_{\mu\nu} that describes spacetime curvature. Instead, they couple to a physical metric g~μν=b2(ϕ)gμν\widetilde{g}_{\mu\nu}=b^{2}(\phi)g_{\mu\nu} resulting from a conformal transformation with the chameleon scalar field ϕ\phi. Thus, various matter fields are coupled to the chameleon scalar field ϕ\phi, providing a potential solution to the localization problem of vector field. Moreover, the coupling of scalar fields and fermion fields in matter to the chameleon scalar field ϕ\phi results in novel characteristics within the braneworld model in chameleon gravity. Brane cosmology in chameleon gravity was investigated in Refs. PhysRevD.85.023526 ; Bisabr2017 , where thin brane models were considered. While thick brane model in chameleon gravity has not yet been explored, this study aims to address this gap. We will construct a thick chameleon brane and attempt to solve the localization problem of vector field in the chameleon thick braneworld. Furthermore, we will investigate the localization property of various matter fields in this model.

The layout of the paper is as follows: In Section 2, we construct a chameleon brane model within the background of a Sine-Gordon kink. In Section 3, we explore the localization and quasi-localization of three types of qq-form fields. In Section 4, we examine the localization of a fermion field. Finally, we conclude with a discussion of our findings. Throughout this paper, M,N,M,N,\cdots denote the indices of the five dimensional coordinates, and μ,ν,\mu,\nu,\cdots denote the ones on the brane.

2 Braneworld in chameleon gravity

We start with the following five-dimensional action

S=d5xg[12R12MϕMϕV(ϕ)]+Sm[g~MN,ψi],S=\int d^{5}x\sqrt{-g}\left[\frac{1}{2}R-\frac{1}{2}\nabla_{M}\phi\nabla^{M}\phi-V(\phi)\right]+S_{m}[\tilde{g}_{MN},\psi_{i}], (1)

which describes a canonical scalar field minimally coupled to gravity, and matter fields ψi\psi_{i} coupled to the Jordan frame metric g~MN=b2(ϕ)gMN\tilde{g}_{MN}=b^{2}(\phi)g_{MN}. It can be seen from the above action that the braneworld solution in this set up is the same with the standard case. However, there are some novel features regarding the localization of various matter fields are expected, since the matter fields couple to the Jordan frame metric g~MN\tilde{g}_{MN} instead of the Einstein frame one gMNg_{MN}.

The metric ansatz of a flat brane embedded in a five-dimensional anti-de Sitter (AdS) spacetime is

gMN=a2(z)[ημνdxμdxν+dz2],g_{MN}=a^{2}(z)\left[\eta_{\mu\nu}dx^{\mu}dx^{\nu}+dz^{2}\right], (2)

where a(z)=eA(z)a(z)=e^{A(z)} is the warped factor. Varying the action with respect to the Einstein frame metric gMNg_{MN} and the scalar field ϕ\phi, and let Sm=0S_{m}=0, we obtain the following equation of motion,

3A23A′′\displaystyle 3A^{\prime 2}-3A^{\prime\prime} =\displaystyle= ϕ2,\displaystyle\phi^{\prime 2}, (3)
12A2+2e2AV(ϕ)\displaystyle 12A^{\prime 2}+2\text{e}^{2A}V(\phi) =\displaystyle= ϕ2,\displaystyle\phi^{\prime 2}, (4)
ϕ′′+3Aϕ\displaystyle\phi^{\prime\prime}+3A^{\prime}\phi^{\prime} =\displaystyle= e2AV(ϕ)ϕ.\displaystyle\text{e}^{2A}\frac{\partial V(\phi)}{\partial\phi}. (5)

Assuming the sine-Gordon potential V(ϕ)=34k2[5cos(2ϕ3)3]V(\phi)=\frac{3}{4}k^{2}\left[5\cos\left(\frac{2\phi}{\sqrt{3}}\right)-3\right], the brane solution is solved as Koley:2004at

A(z)\displaystyle A(z) =\displaystyle= 12ln(k2z2+1),\displaystyle-\frac{1}{2}\ln\left(k^{2}z^{2}+1\right), (6)
ϕ(z)\displaystyle\phi(z) =\displaystyle= 3arctan(kz),\displaystyle\sqrt{3}\arctan\left(kz\right), (7)

where the parameter kk is associated with the inverse of the five-dimensional AdS radius. Since the brane world solution is the same with the standard case, the KK gravitons generate a correction, which is proportional to 1/kr3{1}/{kr^{3}}, to the Newtonian potential Csaki2000 ; Bazeia2009 ; Liu:2012rc ; Yang:2022fed . Conversely, a recent experiment has demonstrated that the length scale deviating from the gravitational inverse square law is at most 48μm48\mu\text{m} PhysRevLett.124.051301 . Thus the parameter kk can be roughly estimated to be k103eVk\gtrsim 10^{-3}\text{eV}.

In the thick brane models, the extra dimension is non-compact, and the matter fields are localized on the brane by the warped geometry, which is described by the Jordan frame metric

g~MN=a~2(z)[ημνdxμdxν+dz2],\tilde{g}_{MN}=\tilde{a}^{2}(z)\left[\eta_{\mu\nu}dx^{\mu}dx^{\nu}+dz^{2}\right], (8)

where we have defined an effective warped factor a~(z)=eA~(z)=eB(ϕ)+A(z)\tilde{a}(z)=\text{e}^{\tilde{A}(z)}=\text{e}^{B(\phi)+A(z)} and b(ϕ)=eB(ϕ)b(\phi)=\text{e}^{B(\phi)} for convenience, and all the functions deduced from g~MN\tilde{g}_{MN} are with tilde in the following.

3 Localization of qq-form fields

In this section, we consider the localization of qq-form field XM1M2MqX_{M_{1}M_{2}\ldots M_{q}} on the brane, which represents a scalar field for q=0q=0, a vector field for q=1q=1, and Kalb-Ramond (KR) field for q=2q=2. We will first investigate a generic qq-form field, and then discuss each of the scalar, vector and Kalb-Ramond fields. The action of a massless qq-form field in the bulk is given by

Sq=12(q+1)!d5xg~YM1M2Mq+1YM1M2Mq+1,S_{q}=-\frac{1}{2(q+1)!}\int d^{5}x\sqrt{-\tilde{g}}Y^{M_{1}M_{2}\ldots M_{q+1}}Y_{M_{1}M_{2}\ldots M_{q+1}}, (9)

where g~\tilde{g} is the determinant of the Jordan frame metric g~MN\tilde{g}_{MN}, the indices are also raised and lowered by g~MN\tilde{g}_{MN}, and the strength of the qq-form field is defined as YM1M2Mq+1=[M1XM2M3Mq+1]Y_{M_{1}M_{2}\ldots M_{q+1}}=\partial_{[M_{1}}X_{M_{2}M_{3}\ldots M_{q+1}]}. By varying action (9) with respect to XM1M2MqX_{M_{1}M_{2}\ldots M_{q}}, the equations of motion of the qq-form field can be obtained,

μ1(g~Yμ1μ2μq+1)+z(g~Yzμ2μq+1)\displaystyle\partial_{\mu_{1}}\left(\sqrt{-\tilde{g}}Y^{\mu_{1}\mu_{2}\ldots\mu_{q+1}}\right)+\partial_{z}\left(\sqrt{-\tilde{g}}Y^{z\mu_{2}\ldots\mu_{q+1}}\right) =\displaystyle= 0,\displaystyle 0, (10)
μ1(g~Yμ1μq1z)\displaystyle\partial_{\mu_{1}}\left(\sqrt{-\tilde{g}}Y^{\mu_{1}\ldots\mu_{q-1}z}\right) =\displaystyle= 0.\displaystyle 0. (11)

3.1 Localization conditions of the zero modes

Before we employing the Kaluza-Klein (KK) decomposition, it is essential to eliminate the gauge degrees of freedom of the qq-form field. It is obvious that the action (9) is invariant under the gauge transformation XM1M2MqX~M1M2Mq=XM1M2Mq+[M1ΛM2M3Mq]X_{M_{1}M_{2}\ldots M_{q}}\rightarrow\tilde{X}_{M_{1}M_{2}\ldots M_{q}}=X_{M_{1}M_{2}\ldots M_{q}}+\partial_{[M_{1}}\Lambda_{M_{2}M_{3}\ldots M_{q}]} with ΛM2M3Mq\Lambda_{M_{2}M_{3}\ldots M_{q}} an antisymmetric tensor field. Therefore, one can eliminate the gauge degrees of freedom by assuming Xμ1μq1z=0X_{\mu_{1}\ldots\mu_{q-1}z}=0. Then we employ the KK decomposition for the rest components,

Xμ1μ2μq(x,z)=nX^μ1μ2μq(n)(x)Υn(z)e(2q3)A~2.\displaystyle X_{\mu_{1}\mu_{2}\ldots\mu_{q}}(x,z)=\sum_{n}\hat{X}_{\mu_{1}\mu_{2}\ldots\mu_{q}}^{(n)}(x)\Upsilon_{n}(z)\text{e}^{\frac{(2q-3)\tilde{A}}{2}}. (12)

The corresponding KK decomposition for the strength is

Yμ1μ2μq+1(x,z)\displaystyle Y_{\mu_{1}\mu_{2}\ldots\mu_{q+1}}(x,z) =\displaystyle= nY^μ1μ2μq+1(n)(x)Υn(z)e(2q3)A~2,\displaystyle\sum_{n}\hat{Y}_{\mu_{1}\mu_{2}\ldots\mu_{q+1}}^{(n)}(x)\Upsilon_{n}(z)\text{e}^{\frac{(2q-3)\tilde{A}}{2}},
Yμ1μ2μqz(x,z)\displaystyle Y_{\mu_{1}\mu_{2}\ldots\mu_{q}z}(x,z) =\displaystyle= nY^μ1μ2μq(n)(x)Υn(z)e(2q3)A~2(Υn(z)+(2q3)2A~Υn(z)).\displaystyle\sum_{n}\hat{Y}_{\mu_{1}\mu_{2}\ldots\mu_{q}}^{(n)}(x)\Upsilon_{n}(z)\text{e}^{\frac{(2q-3)\tilde{A}}{2}}\left(\Upsilon^{\prime}_{n}(z)+\frac{(2q-3)}{2}\tilde{A}^{\prime}\Upsilon_{n}(z)\right). (13)

Substituting the above decompositions into the Eqs. (10) and (11), it is easy to obtain the Schrödinger-like equation of the KK modes Υn(z)\Upsilon_{n}(z),

[z2+Vq(z)]Υn(z)=mn2Υn(z),\displaystyle\left[-\partial^{2}_{z}+V_{q}(z)\right]\Upsilon_{n}(z)=m_{n}^{2}\Upsilon_{n}(z), (14)

with the effective potential Vq(z)V_{q}(z) given by

Vq(z)=32q2A~′′+(32q)24A~2.\displaystyle V_{q}(z)=\frac{3-2q}{2}\tilde{A}^{\prime\prime}+\frac{(3-2q)^{2}}{4}\tilde{A}^{\prime 2}. (15)

Therefore, we can solve the zero mode with m02=0m_{0}^{2}=0, yielding

Υ0(z)=N0e(32q)A~2,\displaystyle\Upsilon_{0}(z)=N_{0}\text{e}^{\frac{(3-2q)\tilde{A}}{2}}, (16)

where N0N_{0} is a normalization constant. In order to ensure that the five-dimensional action (9) reduces to a four-dimensional effective action on the brane, the five-dimensional qq-form field has to be localized on the brane, which can be realized by introducing the orthonormal condition

+Υm(z)Υn(z)𝑑z=δmn.\displaystyle\int^{+\infty}_{-\infty}\Upsilon_{m}(z)\Upsilon_{n}(z)dz=\delta_{mn}. (17)

Then the action (9) can be reduced to the following effective action:

Sq=12(q+1)!nd4x(Y^μ1μ2μq+1(n)Y^(n)μ1μ2μq+1+Mn2q+1X^μ1μ2μq(n)X^(n)μ1μ2μq),\displaystyle\!\!\!\!\!\!\!S_{q}\!=\!-\frac{1}{2(q+1)!}\sum_{n}\!\int\!d^{4}x\!\left(\hat{Y}_{\mu_{1}\mu_{2}\ldots\mu_{q+1}}^{(n)}\hat{Y}^{(n)\mu_{1}\mu_{2}\ldots\mu_{q+1}}\!+\!\frac{M_{n}^{2}}{q+1}\hat{X}_{\mu_{1}\mu_{2}\ldots\mu_{q}}^{(n)}\hat{X}^{(n)\mu_{1}\mu_{2}\ldots\mu_{q}}\!\right)\!, (18)

where the indices are raised and lowered by the Minkowski metric ημν\eta_{\mu\nu}. The effective action describes a massless qq-form field (n=0n=0) and a series of massive qq-form fields (n1n\geq 1) . Therefore, the condition that the zero mode is localized on the brane is given by

+Υ02(z)𝑑z=N02+[a(z)b(ϕ)]32q𝑑z=1,\displaystyle\int^{+\infty}_{-\infty}\Upsilon_{0}^{2}(z)dz=N_{0}^{2}\int^{+\infty}_{-\infty}\left[a(z)b(\phi)\right]^{3-2q}dz=1, (19)

or,

+[a(z)b(ϕ)]32q𝑑z<.\displaystyle\int^{+\infty}_{-\infty}\left[a(z)b(\phi)\right]^{3-2q}dz<\infty. (20)

The above condition can be rewritten as [a(z)b(ϕ)]32qzp\left[a(z)b(\phi)\right]^{3-2q}\rightarrow z^{p} in which p<1p<-1. Furthermore, note that a(±)=±1kza(\pm\infty)=\pm\frac{1}{kz}, so the localization condition reduces to

b[ϕ()]=zp32q+1\displaystyle b[\phi(\infty)]=z^{\frac{p}{3-2q}+1} (21)

where p<1p<-1. In addition, to prevent that Eq. (20) vanishes and for the consideration of symmetry, b(ϕ)b(\phi) must be an even function of ϕ\phi.

Next, we focus on the cases of the scalar field (q=0q=0), vector field (q=1q=1), and KR field (q=2q=2), respectively.

3.1.1 Scalar field

For the case q=0q=0, the action (9) represents a massless scalar field Φ(xμ,z)\Phi(x^{\mu},z), of which the KK composition is

Φ(xμ,z)=nφ(n)(xμ)χn(z)e32A~.\displaystyle\Phi(x^{\mu},z)=\sum_{n}\varphi^{(n)}(x^{\mu})\chi_{n}(z)\text{e}^{-\frac{3}{2}\tilde{A}}. (22)

The effective potential reads

Vs(z)=32A~′′+94A~2.\displaystyle V_{\text{s}}(z)=\frac{3}{2}\tilde{A}^{\prime\prime}+\frac{9}{4}\tilde{A}^{\prime 2}. (23)

If the orthonormal condition 𝑑zχmχn=δnm\int dz\chi_{m}\chi_{n}=\delta_{nm} is satisfied, the fundamental five-dimensional action of a free massless scalar filed can be reduced to the four-dimensional action of a massless and a series of massive scalar fields, given by

Ss=n1d4x(12μφ(0)μφ(0)+12μφ(n)μφ(n)+m(n)2φ(n)φ(n)).\displaystyle S_{\text{s}}=-\sum_{n\geq 1}\int d^{4}x\left(\frac{1}{2}\partial^{\mu}\varphi^{(0)}\partial_{\mu}\varphi^{(0)}+\frac{1}{2}\partial^{\mu}\varphi^{(n)}\partial_{\mu}\varphi^{(n)}+m_{(n)}^{2}\varphi^{(n)}\varphi^{(n)}\right). (24)

In order to be consistent with the standard model, localization of the zero mode χ0(z)=eA~\chi_{0}(z)=\text{e}^{\tilde{A}} is essential. From Eq. (21) we can obtain the localization condition is

b[ϕ()]=zr,\displaystyle b[\phi(\infty)]=z^{r}, (25)

with r<23r<\frac{2}{3}.

3.1.2 Vector field

Then we consider the case q=1q=1, the action (9) represents a U(1) gauge vector field. We choose the gauge A5=0A_{5}=0, and make the KK decomposition for the vector field

Aμ(x,z)=naμ(n)(x)αn(z)eA~2.\displaystyle A_{\mu}(x,z)=\sum_{n}a^{(n)}_{\mu}(x)\alpha_{n}(z)\text{e}^{-\frac{\tilde{A}}{2}}. (26)

With the above decomposition, it is easy to obtain the Schrödinger-like equation of the Vector KK modes α(n)(z)\alpha_{(n)}(z):

[z2+Vv(z)]αn(z)=mn2αn(z),\displaystyle\left[-\partial^{2}_{z}+V_{\text{v}}(z)\right]\alpha_{n}(z)=m_{n}^{2}\alpha_{n}(z), (27)

with the effective potential Vv(z)V_{\text{v}}(z) given by

Vv(z)=12A~′′+14A~2.\displaystyle V_{\text{v}}(z)=\frac{1}{2}\tilde{A}^{\prime\prime}+\frac{1}{4}\tilde{A}^{\prime 2}. (28)

Therefore, the massless vector zero mode reads

α0(z)=N0eA~2.\displaystyle\alpha_{0}(z)=N_{0}\text{e}^{\frac{\tilde{A}}{2}}. (29)

where N0N_{0} is a normalization constant. Now the orthonormal condition (17) is written as

+αm(z)αn(z)𝑑z=δmn.\displaystyle\int^{+\infty}_{-\infty}\alpha_{m}(z)\alpha_{n}(z)dz=\delta_{mn}. (30)

Then the action (9) can be reduced to the following effective action:

Sv=n1d4x(14f(0)μνfμν(0)+14f(n)μνfμν(n)+12mn2a(n)μaμ(n)),\displaystyle S_{\text{v}}=-\sum_{n\geq 1}\int d^{4}x\left(\frac{1}{4}f^{(0)\mu\nu}f^{(0)}_{\mu\nu}+\frac{1}{4}f^{(n)\mu\nu}f^{(n)}_{\mu\nu}+\frac{1}{2}m_{n}^{2}a^{(n)\mu}a^{(n)}_{\mu}\right), (31)

which describes a massless vector field and a series of massive vector fields.

From Eq. (21) the localization of the U(1) gauge field on the brane requires

b[ϕ()]=zr,\displaystyle b[\phi(\infty)]=z^{r}, (32)

with r<0r<0.

3.1.3 KR field

At last, we consider the case q=2q=2, and the action (9) represents a massless KR field BMNB_{MN}. The KK decomposition (12) reduces to

Bμν(x,z)=nbμν(n)(x)un(z)eA~2,\displaystyle B_{\mu\nu}(x,z)=\sum_{n}b^{(n)}_{\mu\nu}(x)u_{n}(z)\text{e}^{\frac{\tilde{A}}{2}}, (33)

and the effective potential (15) reads

VKR(z)=14A212A′′.\displaystyle V_{\text{KR}}(z)=\frac{1}{4}A^{\prime 2}-\frac{1}{2}A^{\prime\prime}. (34)

If the orthonormal condition (21) is satisfied, the action of the five-dimensional KR field reduces to

SKR=n1d4x(F(0)μνλFμνλ(0)+F(n)μνλFμνλ(n)+13mn2b(n)μνbμν(n)),\displaystyle S_{\text{KR}}=-\sum_{n\geq 1}\int d^{4}x\left(F^{(0)\mu\nu\lambda}F^{(0)}_{\mu\nu\lambda}+F^{(n)\mu\nu\lambda}F^{(n)}_{\mu\nu\lambda}+\frac{1}{3}m_{n}^{2}b^{(n)\mu\nu}b^{(n)}_{\mu\nu}\right), (35)

which describes a four-dimensional massless KR field and a series of four-dimensional massive KR fields on the brane.

From Eq. (21), we obtain the localization condition of the zero mode, i.e.,

b[ϕ()]=zr,\displaystyle b[\phi(\infty)]=z^{r}, (36)

with r>0r>0.

In order to be consistent with the observations that photons are the carrier of the electromagnetic interaction but KR particles have not been discovered yet on the brane, it is necessary for the vector field to be localized but KR filed to not be localized on the brane. Interestingly, the condition for localizing the vector field conflicts with that of the KR field, and therefore the expected result can indeed be achieved as long as r<0r<0.

In summary, in the chameleon braneworld, the scalar and vector fields can be localized on the brane, while the KR field can not. The localization condition of scalar and vector fields is:

+a(z)b(ϕ)𝑑z<.\int^{+\infty}_{-\infty}a(z)b(\phi)dz<\infty. (37)

It is clear that b(ϕ)b(\phi) can not be an odd function of ϕ\phi, otherwise the integration vanishes. For simplicity, here we assume that b(ϕ)b(\phi) is an even function of ϕ\phi. As an example, the above condition can be simply satisfied by just choosing b(ϕ)=cosp(ϕ3)b(\phi)=\cos^{p}(\frac{\phi}{\sqrt{3}}) with p>0p>0, which is the ansatz we adopted in the rest of the work.

3.2 Quasi-localiztion of qq-form fields

In addition to the above investigation of the bounded zero modes, it is also necessary to consider the mass spectra of the quasi-localized states (i.e., resonances). Since the effective potential (15) is an even function of zz, the resonance Υn(z)\Upsilon_{n}(z) has either an even parity Υe(n)(z)\Upsilon^{(n)}_{\text{e}}(z) or an odd parity Υo(n)(z)\Upsilon^{(n)}_{\text{o}}(z). They respectively satisfy the following boundary conditions,

Υe(n)(0)=1,zΥe(n)(0)=0;\displaystyle\Upsilon^{(n)}_{\text{e}}(0)=1,~{}~{}~{}~{}~{}~{}~{}~{}\partial_{z}\Upsilon^{(n)}_{\text{e}}(0)=0; (38)
Υo(n)(0)=0,zΥo(n)(0)=1.\displaystyle\Upsilon^{(n)}_{\text{o}}(0)=0,~{}~{}~{}~{}~{}~{}~{}~{}\partial_{z}\Upsilon^{(n)}_{\text{o}}(0)=1. (39)

The resonances can be found by adopting the concept of the relative probability of the KK mode Υ(n)(z)\Upsilon^{(n)}(z) with mass mnm_{n}, which was defined in Ref. Liu:2009ve ; PhysRevD.79.125022 :

P(mn2)=zbzb|Υ(n)(z)|2𝑑zzmaxzmax|Υ(n)(z)|2𝑑z.\displaystyle P(m_{n}^{2})=\frac{\int^{z_{b}}_{-z_{b}}|\Upsilon^{(n)}(z)|^{2}dz}{\int^{z_{max}}_{-z_{max}}|\Upsilon^{(n)}(z)|^{2}dz}. (40)

Here 2zb2z_{b} is approximately the width of the thick brane (or the width of the effective potential well Vq(z)V_{q}(z)), and zmax=10zbz_{max}=10z_{b}. By solving the Schrödinger-like equation (14) for a given mn2m_{n}^{2} numerically, the corresponding relative probability PP can be obtained for the even or odd modes. Then each peak in the figure of P(mn2)P(m_{n}^{2}) represents a resonant mode. The life-time τ\tau of a resonant mode can obtained by τ=1Γ\tau=\frac{1}{\Gamma}, where Γ\Gamma is the full width at half maximum (FWHM) PhysRevLett.84.5928 ; PhysRevD.79.125022 . It is more convenient to use the dimensionless quantities m/km/k and kτk\tau, which have been scaled with kk. Next we analyze the resonances of scalar, vector and KR fields, respectively.

Refer to caption
(a) Scalar field
Refer to caption
(b) Vector field
Refer to caption
(c) KR field
Figure 1: The effective potential of the q-form fields for different values of pp.

3.2.1 Scalar field

The effective potential of the scalar field and the relative probability as a function of m2/k2m^{2}/k^{2} are plotted in Fig. 1(a) and Fig. 2 for different values of pp, respectively. Since the maximum of the potential increases with the parameter pp, it can be expected that the relative probability, mass, and life-time of the resonances also increases with the parameter pp, as is shown in Fig. 3. As shown in the figures, the first scalar resonance, which has an old parity, appears only if p1p\gtrsim 1. Since the bound zero mode has an even parity, the first resonance is an odd mode.

Refer to caption
(a) p=2p=2
Refer to caption
(b) p=4p=4
Refer to caption
(c) p=6p=6
Figure 2: The relative probability P(m2/k2)P\left({m^{2}}/{k^{2}}\right) of the even and odd parity modes of scalar field for different values of pp.
Refer to caption
(a) relative probability PmaxP_{max}
Refer to caption
(b) scaled mass m/k{m}/{k}
Refer to caption
(c) scaled life-time kτk\tau
Figure 3: The relative probability PmaxP_{max}, scaled mass m/k{m}/{k}, and scaled life-time kτk\tau of the resonances of scalar field for different values of pp.

3.2.2 Vector field

The resonances of the vector field behave similar with those of the scalar field. The effective potential of the vector field and the relative probability as a function of m2/k2m^{2}/k^{2} are plotted in Fig. 1(b) and Fig. 4 for different values of pp, respectively. And the relative probability, mass, and life-time of the resonances for different values of the parameter pp is shown in Fig. 5. As shown in the figures, the first vector resonance, which processes an old parity, appears when p7p\gtrsim 7.

Refer to caption
(a) p=8p=8
Refer to caption
(b) p=14p=14
Refer to caption
(c) p=18p=18
Figure 4: The relative probability P(m2/k2)P\left({m^{2}}/{k^{2}}\right) of the even and odd parity modes of the vector field for different values of pp.
Refer to caption
(a) relative probability PmaxP_{max}
Refer to caption
(b) scaled mass m/k{m}/{k}
Refer to caption
(c) scaled life-time kτk\tau
Figure 5: The relative probability PmaxP_{max}, scaled mass m/k{m}/{k}, and scaled life-time kτk\tau of the resonances of the vector field for different values of pp.

3.2.3 KR field

The effective potential of the scalar field and the relative probability as a function of m2/k2m^{2}/k^{2} are plotted in Fig. 1(c) and Fig. 6 for different values of pp, respectively. And the relative probability, mass, and life-time of the resonances for different values of the parameter pp is shown in Fig. 7. It can be seen that the resonances of the KR field behave different with those of the scalar and vector fields. As shown in Fig. 1(c), the effective potential possesses a local well only if p5p\gtrsim 5. As shown in the Fig. 7, the first KR resonance appears when p5p\gtrsim 5. Since the zero mode, which has an even parity is not localized on the brane, the first KR resonance has even parity.

Refer to caption
(a) p=3p=3
Refer to caption
(b) p=7p=7
Refer to caption
(c) p=15p=15
Figure 6: The relative probability P(m2/k2)P\left(m^{2}/k^{2}\right) of the even and odd parity modes of the KR field for different values of pp.
Refer to caption
(a) relative probability PmaxP_{max}
Refer to caption
(b) scaled mass m/k{m}/{k}
Refer to caption
(c) scaled life-time kτk\tau
Figure 7: The relative probability PmaxP_{max}, scaled mass m/k{m}/{k}, and scaled life-time kτk\tau of the resonances of the KR field for different values of pp.

3.2.4 Constraints on the resonances

It can be seen that the massive resonance of the vector field occurs only if p7p\gtrsim 7, and its mass mm is approximately m/k2m/k\simeq 2 for p=7p=7. Considering the experimental constraint on the photon mass, mγ<1018eVm_{\gamma}<10^{-18}\text{eV} Amsler2008 , it leads to k1018eVk\lesssim 10^{-18}\text{eV} when p7p\gtrsim 7. However, this contradicts the experimental constraint k103eVk\gtrsim 10^{-3}\text{eV} form the deviation from gravitational inverse-square law. Therefore, it implies that no vector resonance exists on the brane and the parameter pp is constrained as 0<p70<p\lesssim 7.

From Fig. 3, it can be observed that for the resonances of the scalar field, their masses approximate m𝒪(k)m\sim\mathcal{O}(k), and their lifetimes τ\tau range from k1<τ<300k1k^{-1}<\tau<300k^{-1}. Similarly, as shown in Fig. 7, the mass of KR resonance is roughly m𝒪(k)m\sim\mathcal{O}(k), and the lifetime τk1\tau\sim k^{-1}. For instance, if k103eVk\sim 10^{-3}\text{eV}, the masses of the scalar and KR resonances are of order of 103eV10^{-3}\text{eV}. Correspondingly, the lifetimes of scalar resonances range form 1012s10^{-12}\text{s} to 1010s10^{-10}\text{s}, and the lifetime of KR resonances is roughly τ1012s\tau\sim 10^{-12}\text{s}.

4 Localization of fermion field

In this section, we investigate the localization of a spin-12\frac{1}{2} fermion field on the brane. The Dirac structure of the fermion fields in the bulk is described by Γ~M=e~M¯MΓ~M¯\tilde{\Gamma}^{M}=\tilde{e}^{M}_{\bar{M}}\tilde{\Gamma}^{\bar{M}} with {Γ~M,Γ~N}=2g~MN\{\tilde{\Gamma}^{M},\tilde{\Gamma}^{N}\}=2\tilde{g}^{MN}. Here M¯,N¯,\bar{M},\bar{N},\cdots denote the local Lorentz indices and Γ~M\tilde{\Gamma}^{M} are the gamma matrices corresponding to the Jordan frame metric g~MN\tilde{g}^{MN}. In this set up, it is easy to see that all the equations are the same with the standard case in general relativity except that all the quantities deduced from g~MN\tilde{g}^{MN} should be marked by tilde. With the metric (8), we can obtain Γ~M=eBΓM=(eA~γμ,eA~γ5)\tilde{\Gamma}^{M}=\text{e}^{-B}\Gamma^{M}=(\text{e}^{-\tilde{A}}\gamma^{\mu},\text{e}^{-\tilde{A}}\gamma^{5}). We consider the following action of a massless spin-12\frac{1}{2} fermion field coupled to the background chameleon scalar field in a Yukawa-type interaction,

Sf=d5xg~Ψ¯Γ~MD~MΨηΨ¯f(ϕ)Ψ,\displaystyle S_{f}=\int d^{5}x\sqrt{-\tilde{g}}\bar{\Psi}\tilde{\Gamma}^{M}\tilde{D}_{M}\Psi-\eta\bar{\Psi}f(\phi)\Psi, (41)

where f(ϕ)f(\phi) is a function of the chameleon scalar field, and the generic covariant derivative D~M=M+ω~M\tilde{D}_{M}=\partial_{M}+\tilde{\omega}_{M} is defined with the spin connection ω~M=14ω~MM¯N¯Γ~M¯ΓN¯\tilde{\omega}_{M}=\frac{1}{4}\tilde{\omega}_{M}^{\bar{M}\bar{N}}\tilde{\Gamma}_{\bar{M}}\Gamma_{\bar{N}}, with

ω~MM¯N¯=\displaystyle\tilde{\omega}_{M}^{\bar{M}\bar{N}}= 12e~NM¯(Me~NN¯Ne~MN¯)12e~NN¯(Me~NM¯Ne~MM¯)\displaystyle\frac{1}{2}{\tilde{e}}^{N\bar{M}}(\partial_{M}\tilde{e}_{N}^{\bar{N}}-\partial_{N}\tilde{e}_{M}^{\bar{N}})-\frac{1}{2}\tilde{{e}}^{N\bar{N}}(\partial_{M}\tilde{e}_{N}^{\bar{M}}-\partial_{N}\tilde{e}_{M}^{\bar{M}}) (42)
12e~PM¯e~QN¯(Pe~QR¯Qe~PR¯)e~MR¯.\displaystyle-\frac{1}{2}\tilde{{e}}^{P\bar{M}}\tilde{{e}}^{Q\bar{N}}(\partial_{P}\tilde{e}_{Q{\bar{R}}}-\partial_{Q}\tilde{e}_{P{\bar{R}}})\tilde{{e}}^{\bar{R}}_{M}.

Obviously, the fermion field couples to the scalar field not only through the Yukawa interaction, but also through the determinant g~\sqrt{-\tilde{g}}, the Gamma matrix Γ~\tilde{\Gamma} and the generic covariant derivative D~M\tilde{D}_{M}, which are all related to the Jordan frame metric g~MN\tilde{g}_{MN}. With the metric (8), we obtain the non-vanishing components of the spin connection ωM\omega_{M},

ω~μ=12A~γμγ5.\displaystyle\tilde{\omega}_{\mu}=\frac{1}{2}\tilde{A}^{\prime}\gamma_{\mu}\gamma_{5}. (43)

The variation of (41) with respect to Ψ¯\bar{\Psi} yields the five-dimensional Dirac equation,

Γ~MD~MΨ=ηf(ϕ)Ψ,\displaystyle\tilde{\Gamma}^{M}\tilde{D}_{M}\Psi=\eta f(\phi)\Psi, (44)

Substituting Eqs. (43) into the above equation, one can obtain the explicit form of the Dirac equation in the metric (8)

[γμμ+γ5(z+2A~)ηeA~f(ϕ)]Ψ=0,\displaystyle\left[\gamma^{\mu}\partial_{\mu}+\gamma^{5}\left(\partial_{z}+2\tilde{A}^{\prime}\right)-\eta\;\text{e}^{\tilde{A}}f(\phi)\right]\Psi=0, (45)

Then we employ the chiral decomposition

Ψ(x,z)=e2A~nψLn(x)αLn(z)+nψRn(x)αRn(z),\Psi(x,z)=\text{e}^{-2\tilde{A}}\sum_{n}\psi_{Ln}(x)\alpha_{Ln}(z)+\sum_{n}\psi_{Rn}(x)\alpha_{Rn}(z), (46)

where ψLn(x)=γ5ψLn(x)\psi_{Ln}(x)=-\gamma^{5}\psi_{Ln}(x) and ψRn(x)=γ5ψRn(x)\psi_{Rn}(x)=\gamma^{5}\psi_{Rn}(x) are left-handed and right-handed components of four-dimensional Dirac fields respectively. By substituting this decomposition into the five-dimensional Dirac equation (44), the four-dimensional Dirac parts ψLn(x)\psi_{Ln}(x) and ψRn(x)\psi_{Rn}(x) satisfy the four-dimensional Dirac equations

γμμψLn(x)=mnψRn(x),\displaystyle\gamma^{\mu}\partial_{\mu}\psi_{Ln}(x)=m_{n}\psi_{Rn}(x), (47)
γμμψRn(x)=mnψLn(x).\displaystyle\gamma^{\mu}\partial_{\mu}\psi_{Rn}(x)=m_{n}\psi_{Ln}(x). (48)

and the five-dimensional Dirac parts αL(z)\alpha_{L}(z) and αR(z)\alpha_{R}(z) satisfy the coupled equations

[z+ηeA~f(ϕ)]αLn(z)\displaystyle\left[\partial_{z}+\eta\;\text{e}^{\tilde{A}}f(\phi)\right]\alpha_{Ln}(z) =\displaystyle= mnαRn(z),\displaystyle~{}~{}m_{n}\alpha_{Rn}(z), (49a)
[zηeA~f(ϕ)]αRn(z)\displaystyle\left[\partial_{z}-\eta\;\text{e}^{\tilde{A}}f(\phi)\right]\alpha_{Rn}(z) =\displaystyle= mnαLn(z).\displaystyle-m_{n}\alpha_{Ln}(z). (49b)

After reassembling the two coupled equations, one ultimately achieves the Schrödinger-like equations

[z2+VL,R(z)]αLn,Rn\displaystyle\left[-\partial^{2}_{z}+V_{L,R}(z)\right]\alpha_{Ln,Rn} =\displaystyle= mn2αLn,Rn,\displaystyle m_{n}^{2}\alpha_{Ln,Rn}, (50)

with the effective potentials given by

VL,R(z)\displaystyle V_{L,R}(z) =\displaystyle= [eA~ηf(ϕ)]2z[eA~ηf(ϕ)].\displaystyle\left[\text{e}^{\tilde{A}}\eta f(\phi)\right]^{2}\mp\partial_{z}\left[\text{e}^{\tilde{A}}\eta f(\phi)\right]. (51)

Further, to obtain the effective action of a massless and a series of massive four-dimensional fermions on the brane, the following orthonormal conditions have to be imposed,

αLmαLn\displaystyle\int\alpha_{Lm}\alpha_{Ln} =\displaystyle= δmn,\displaystyle\delta_{mn}, (52)
αRmαRn\displaystyle\int\alpha_{Rm}\alpha_{Rn} =\displaystyle= δmn,\displaystyle\delta_{mn}, (53)
αLmαRn\displaystyle\int\alpha_{Lm}\alpha_{Rn} =\displaystyle= 0.\displaystyle 0. (54)

The chiral zero modes can be solved by applying the factorizing method, given by

αL0,R0=NL0,R0eηeA~f(ϕ)𝑑z=eηeAb(ϕ)f(ϕ)𝑑z,\displaystyle\alpha_{L0,R0}=N_{L0,R0}\text{e}^{\mp\eta\int\text{e}^{\tilde{A}}f(\phi)dz}=\text{e}^{\mp\eta\int\text{e}^{A}b(\phi)f(\phi)dz}, (55)

with NL0,R0N_{L0,R0} the normalization constants. Since we need at least the massless fermion to be localized on the brane, the normalization condition for the zero modes must be satisfied, i.e.,

+e2ηeAb(ϕ)f(ϕ)𝑑z𝑑z<,\displaystyle\int^{+\infty}_{-\infty}\text{e}^{\mp 2\eta\int\text{e}^{A}b(\phi)f(\phi)dz}dz<\infty, (56)

where the plus or minus sign is corresponding to the zero mode of left or right-handed four-dimensional fermion. Obviously, only one of the zero modes can be localized on the brane depending on the value of coupling constant η\eta.

It is well known that the standard model is a chiral theory, where the left and right-handed fermions transform differently under the electroweak gauge group. To generate the chiral fermions on the brane, one can include a left-handed SU(2)LSU(2)_{L} doublet ΨL\Psi_{L} and a right-handed singlet ΨR\Psi_{R} in the five-dimensional bulk, then the corresponding chiral zero modes are picked up by properly setting the coupling constants ηL/R\eta_{L/R} PhysRevD.62.084025 . Here, we take the localization of left-handed four-dimensional fermion as an example, i.e., +e2ηLeAb(ϕ)f(ϕ)𝑑z𝑑z<\int^{+\infty}_{-\infty}\text{e}^{-2\eta_{L}\int\text{e}^{A}b(\phi)f(\phi)dz}dz<\infty. Note that the brane is embedded in a five-dimentional asymptotic AdS spacetime, we have eA(±)=±1kz\text{e}^{A(\pm\infty)}=\pm\frac{1}{kz}, and the kink configuration of the scalar field yields ϕ(±)=±3π2±v0\phi(\pm\infty)=\pm\frac{\sqrt{3}\pi}{2}\equiv\pm v_{0}. Therefore,

e2ηLeAb(ϕ)f(ϕ)𝑑zz2ηLb(v0)f(v0)kasz+,\displaystyle\text{e}^{-2\eta_{L}\int\text{e}^{A}b(\phi)f(\phi)dz}\rightarrow z^{-\frac{2\eta_{L}b(v_{0})f(v_{0})}{k}}~{}~{}\text{as}~{}~{}z\rightarrow+\infty, (57)
e2ηLeAb(ϕ)f(ϕ)𝑑z(z)2ηLb(v0)f(v0)kasz.\displaystyle\text{e}^{-2\eta_{L}\int\text{e}^{A}b(\phi)f(\phi)dz}\rightarrow(-z)^{\frac{2\eta_{L}b(-v_{0})f(-v_{0})}{k}}~{}~{}\text{as}~{}~{}z\rightarrow-\infty. (58)

Then the normalization condition (56) is reduced to that: f(ϕ)f(\phi) must be an odd function of ϕ\phi, b(v0)f(v0)b(v_{0})f(v_{0}) is a non-vanishing constant, and ηL>kb(v0)f(v0)\eta_{L}>\frac{k}{b(v_{0})f(v_{0})}. As two examples, we can assume that f(ϕ)=ϕcosp(ϕ3)f(\phi)=\frac{\phi}{\cos^{p}(\frac{\phi}{\sqrt{3}})} or f(ϕ)=sin(ϕ)cosp(ϕ3)f(\phi)=\frac{\sin(\phi)}{\cos^{p}(\frac{\phi}{\sqrt{3}})}, then the localization of left-handed zero mode is realized for ηL>kv0\eta_{L}>\frac{k}{v_{0}} or ηL>ksinv0\eta_{L}>\frac{k}{\sin v_{0}}, respectively.

It is noted that, although the fermion field couples to the chameleon scalar field ϕ\phi in several places in the action (41), the Yukawa interaction is still essential to localizing the fermion zero mode.

5 Conclusion

In this work, we investigated the localization of qq-form fields and fermion field on the thick brane construct by a chameleon scalar field. We had demonstrated that the localization problem of the vector field can be solved by choosing an appropriate conformal factor b(ϕ)b(\phi) in this model. The conditions for localization of various matter fields were obtained. For the qq-form fields, the zero modes of the scalar and vector fields can be localized, in the condition that b[ϕ()]=zrb[\phi(\infty)]=z^{r} with r<0r<0, while the zero mode of KR field can not. Furthermore, the quasi-localization of qq-form fields was also considered. It was found that the relative probability, mass, and life-time of the resonances increase with the parameter pp. The parity of first resonances of scalar and vector fields is odd, and the one of the KR field is even. In order to be consistent with both the experimental constraints of the photon mass and the deviation from gravitational inverse-square law, the parameters were constrained as 0<p70<p\lesssim 7 and k103eVk\gtrsim 10^{-3}\text{eV}.

For the fermion field, it couples to the chameleon scalar field not only through the Yukawa interaction, but we found that the term of Yukawa interaction is still necessary in order to localize the fermion zero modes. The condition for localizing the left-handed fermion zero mode was found to be, f(ϕ)f(\phi) an odd function of ϕ\phi, b(v0)f(v0)b(v_{0})f(v_{0}) a non-vanishing constant, and η>kb(v0)f(v0)\eta>\frac{k}{b(v_{0})f(v_{0})}.

Acknowledgement

Y. Zhong acknowledges the support of the Natural Science Foundation of Hunan Province, China (Grant No. 2022JJ40033), the Fundamental Research Funds for the Central Universities (Grants No. 531118010195) and the National Natural Science Foundation of China (No. 12275076). He also thanks the generous hospitality offered by the Center of Theoretical Physics at Lanzhou University where part of this work was completed. K. Yang acknowledges the support of the National Natural Science Foundation of China under Grant No. 12005174.

References