9512.net

甜梦文库

甜梦文库

当前位置：首页 >> >> # Muon anomalous magnetic moment in a Calculable Model with one Extra Dimension

Muon anomalous magnetic moment in a Calculable Model with one Extra Dimension

arXiv:hep-ph/0111288v2 17 Dec 2001

Giacomo Cacciapagliaa, 1 , Marco Cirellia, 2 , Giampaolo Cristadorob, 3 Scuola Normale Superiore and INFN, Piazza dei Cavalieri 7, I-56126 Pisa, Italy ‘Enrico Fermi’ Dep. of Physics, University of Pisa, via Buonarroti 2, I-56126 Pisa, Italy

a

b

Abstract In the framework of a recently proposed extension of the Standard Model, with N = 1 SuperSymmetry in 5 dimensions, compacti?ed on R1 /Z2 × Z′ 2 , we compute the muon anomalous magnetic 2 moment at one loop to order (MW R) , where R is the compacti?cation radius. We ?nd the corrections to be small with respect to the SM pure weak contribution and not capable of explaining the present discrepancy between theory and experiment for any sensible value of R.

PACS: 11.10.Kk, 12.60.Jv, 13.40.Em

1 2

cacciapa@cibs.sns.it mcirelli@cibs.sns.it 3 cristado@cibs.sns.it

1

Introduction

Models with extra (space) dimensions often fail to produce quantitative predictions for physical observables and, as a consequence, can hardly be ruled out or con?rmed by present energy experiments. On the other hand, in [1] a model was proposed where calculability is achieved for a number of physical quantities. In fact, in spite of the non renormalizability of 5 dimensional Yang-Mills theory, they are ?nite and cut-o? independent, owing to the underlying supersymmetric structure. The insensitivity to ultraviolet physics of such observables is one of the main goals of the model. The branching ratio of B → Xs γ [2] and the Higgs production via gluon fusion [3] have already been computed, while in the present paper we deal with the muon anomalous magnetic moment a? . The measurement of a? is one of the most stringent tests for “new physics” scenarios, thanks to its current impressive precision. Moreover, the recent results of E821 experiment at BNL [4] have highlighted a discrepancy with the present Standard Model prediction and have therefore made even more interesting the comparison with reliable predictions made by new models. In particular, attempts have been made in order to study the e?ect of extra space dimensions [5]. Let us brie?y review the present status of the SM prediction [6]. The QED contribution has been computed up to order α5 and it gives the core of the experimental value: aQED = (11 658 470.0 ± ? 0.3) · 10?10 . The purely weak terms are also known with a small uncertainty, up to two loops: aweak = ? ? 10 (15.1 ± 0.4) · 10 . On the contrary, hadronic loop corrections provide relevant contributions but also carry the greatest theoretical uncertainty [7]. In fact, since it is quite hard to deal with low energy ?10 QCD, experimental inputs are also needed. Light-by-light scattering yields albl ? = ?(8.6 ± 3.2) · 10 hvp and hadronic vacuum polarization a? = (684.9 ± 6.4) · 10?10 . Summing up all these contributions and uncertainties, the SM prediction is: aSM = (11 659 162.0 ± 7.5) · 10?10 ? On the other hand, the present world average for the experimental value (including BNL results) is [6]:

?10 aexp ? = (11 659 202.3 ± 15.1) · 10

This leads to a discrepancy a bit larger than 2σ :

SM aexp = (40.3 ± 16.9) · 10?10 ? ? a?

(1)

Actually, this result is not well established. On the theoretical side, recent results [8] [9] claim that the light-by-light contribution has opposite sign, thus reducing the discrepancy to about 1σ . At the same time, the error on the experimental value should be improved by further results at BNL. In this work we compute the corrections ?a? to the anomalous magnetic moment due to the new ?elds in the theory, up to order (MW R)2 where R is the compacti?cation radius of the 5th dimension.

1

2

The model

In this section we brie?y summarize the main features of the model [1] we are working with. It is a 5D theory with N=1 supersymmetry, compacti?ed on R1 /Z2 × Z′ 2 , and the gauge group is the Standard Model one: SU (3) × SU (2) × U (1). The ?elds content is made up of the vector hypermultiplets and a matter hypermultiplet for each SM matter ?eld Q, U , D , L, E , H , all living in the bulk. From a 4D point of view, a matter hypermultiplet splits into a pair of chiral supermultiplets with conjugate quantum numbers (property that we indicate with a c ), while a vector hypermultiplet splits into a vector and a chiral supermultiplet. Under the orbifolding, global supersymmetry is completely broken a ` la Scherk-Schwarz, leaving the SM ?elds as the only zero modes. Anyhow, restricted local supersymmetric transformations still hold. The most general Lagrangian, according to this set of symmetries (gauge group, orbifold parities and local supersymmetry), is:

L(x, y ) = L5 + δ(y ) L4 + δ(y ? πR/2) L′ 4

(2)

where L5 is N = 1 supersymmetric in 5D whereas L4 and L′ 4 are 4D lagrangians invariant under di?erent N = 1 supersymmetries valid on the ?xed points of the two parities. To be consistent with ref. [1], we will write all hypermultiplets in terms of supermultiplets of the 4-dimensional N=1 supersymmetry located on the brane y = 0. So, the gauge vector and matter hypermultiplets are: 1 Wρ , λ ⊕ λ′ , Σ = √ (σ + iW5 ) 2

c ψX , φX ⊕ ψX , φc X

Under the two parities all the ?elds have de?nite transformation properties, given in Table 1 together with the corresponding eigenfunctions and the spectrum of every tower of Kaluza-Klein states. A? , ψM , φH λ, φM , ψH

c? ? ψΣ , φc M , ψH c? ? φΣ , ψM , φc H

(+, +) (+, ?) (?, +) (?, ?)

2n cos y R 2n ? 1 y cos R 2n ? 1 y sin R 2n sin y R

2n , n = 0, 1, 2, 3... R 2n ? 1 , n = 1, 2, 3... R 2n ? 1 , n = 1, 2, 3... R 2n , n = 1, 2, 3... R

Table 1: Gauge, Matter and Higgs ?elds content of the theory with their orbifolding properties In [10] it is shown that in the model under consideration the hypercharge current has an anomaly, which is localized on the two branes. A relevant feature is that, from a 4 dimensional point of view, the integrated anomaly vanishes. Gauge invariance can however be recovered with a suitable modi?cation of the theory. The overall consistency of the model modi?ed by the addition of a Chern-Simons term, for instance, is currently under examination. Such modi?cations, however, should not a?ect the calculation 2

presented in this paper, since anomalous contributions only enter at higher order. The terms in the 5D action, that are relevant for the following calculation, are: 1 ′ c c ρ σ Dρ ψ X + ? TrFM N F M N ? Tr iλσ ρ Dρ λ ? iλ′ σ ρ Dρ λ + iψX σ ρ Dρ ψ X + iψX 4 √ ? c? c ? M c +(DM φX )? (D M φX ) + (DM φc X ) (D φX ) + 2g(5) φX λψX ? ψX λφX + h.c. + √ ′ c c ′ + 2g(5) ψX λ φX + φc X λ ψX + ψX ΣψX + h.c. + π c? c? c c +δ(y ? R)λ? ψL ψE φ? + ψL ψH 0 φE + ψE ψH 0 φL + H0 2 c? c? c c ?(ψν ψE φ? + ψν ψH + φE + ψE ψH + φν ) H+

S =

d5 x

(3)

where the fermionic ?elds are Weyl spinors, the gauge ?elds WM , λ, λ′ and σ contains the group generators, e.g. λ ≡ λa T a (the generators being normalized according to Tr(T a T b ) = δab ) and the index X can run over E , L, H . The gauge coupling g is intended to be the UY (1) coupling g′ or the SUL (2) coupling g according to the gauge ?elds involved. From now on, for the four gauge hypermultiplets we use the notation:

QED Zs Ws

′ ′

Aρ , γ ?, γ ?c , Σγ Zρ , z ?, z ?c , Σz

± , Σ± ±, w ?± , w ?c Wρ w

(we remind that, also in this case, the subscribed c only means that the second spinors have conjugate quantum numbers).

2.1

Mass Eigenstates

The brane interactions introduce mass mixing among the states of muon and smuon KK towers: since it will be useful for our calculation to work with the mass eigenstates, we proceed to determine eigenvalues and eigenvectors from the very beginning (see e.g. the discussion in [2]). For the fermions, the mass eigenvalues turn out to be: 2n ? , m± (n = 1, 2, ...) n =m± R R c± ± corresponding to mass eigenvectors whose Weyl spinor components we denote by (?0 , ?c 0 ) and (?n , ?n ). Thus m is the mass of the ordinary muon. For the scalars: m0 = m ≡ (2n ? 1) ±m R c± corresponding to eigenvectors Φ± n and Φn . m ?± n = (n = 1, 2, ...)

3

The interaction eigenstates1 are expressed in terms of the mass eigenstates (at order ?) by ψL,0 = ?0 + ?

n

1 + ? ? ?? n 2n n ? 2(2n) ?+ l + 1? ?+ l ? 2(2n) ?? l +?

n ? κnl ?+ n ? ?n

(4) (5) √ 2 ? ?0 ? 2l (6)

ψL,l =

c ψE,l

1 √ 2 1 √ 2

1+

=

? 1? 2(2n)

? ? 1+ 2(2n)

?? l

??

κln ?+ n

n

+

?? n

φL,l = φc E,l =

1 √ 2 1 √ 2

1+

? 2(2n ? 1) ? 2(2n ? 1)

Φ+ l + 1?

? 2(2n ? 1) ? 2(2n ? 1)

Φ? l +?

n

? Jln Φ+ n ? Φn ? Jnl Φ+ n + Φn n

(7) (8)

? 1?

Φ+ l + 1+

Φ? l +?

2(2l?1) 2(2l) with κln = (2l)2 ?(2n)2 and Jln = (2l?1)2 ?(2n?1)2 , and zero for n = l. Expressions which are identical to c ) if (? , ?± ) are replaced by (?c , ?c± ); for (φc , φ ) it is su?cient (4), (5), (6) hold for (ψE,0 , ψE,n , ψL,n E 0 n n 0 L c± in equations (7), (8). to replace Φ± with Φ n n

2.2

Couplings to the mass eigenstates

Starting from eq. (3), one has to integrate out the 5th dimension y and rewrite the interaction vertices in terms of the mass eigenstates (4)-(8). Taking into account momentum conservation along the 5th direction, p5 , the bulk interactions give rise to several trilinear vertices. The muon masses and Higgs vertices, on the other hand, being located on the brane, violate p5 -conservation and are proportional to ?. As we are interested in O(R2 ) contributions, only graphs where the internal lines carry the same index n, in terms of the mass eigenstates, have to be taken into account. This simpli?es the calculation of the relevant vertices. As an explicit example, the interaction terms between KK photons and the various components in the muon towers, in the mass eigenstate basis, are: 1 ? ? eψ L,0 σ ρ ψL,n Aρ 1? ?+ n = e?0 σ ρ √ n + 1+ 2(2n) 2(2n) 2 ? 1 e c c ? ρ √ ψ E,2n σ ρ ψE,n Aρ ? σ ρ √ ?+ n = ?e n ? ?n An 2(2n) 0 2 2

ρ ?? n An

In this particular case, summing up the two contributions, the couplings of order ? vanish, so that the net result is:

Note that in this paper the brane lagrangians L4 and L′ 4 are reversed respect to ref [1]. For more details, see appendix A in [3].

1

4

1 ? ρ e?0 σ ρ √ ?+ n + ?n An 2

(9)

√ This allows to recognize e/ 2 as the correct coupling with a KK photon. On the other hand, the interactions of the zero mode gauge bosons are still diagonal in n and have √ a universal coupling, e.g. e = e(5) / 2πR for the photon. Note that this relation holds for all gauge √ couplings, so in the following g = g(5) / 2πR is the 4D gauge coupling, while s = sin θW and c = cos θW are functions of the usual SM Weinberg angle.

3

Calculation of a?

The e?ective Lagrangian term which ?xes the notation for the anomalous magnetic moment a? = (g? ? 2)/2 is:

In the Feynman–’t Hooft gauge (ξ = 1, see below), one can identify the ?ve general types of diagrams in ?g. 1, each giving a contribution ?a? whose explicit expressions are listed in app. A.

? ? ? ¤ ?

L= ie a? ?σρσ F ρσ ? 2m

(10)

×

×

?

?

×

?

??

???

?

×

?

??

?

Figure 1: General patterns of graphs contributing to the (g ? 2). As in the following we will distinguish between diagrams with chirality ?ip on the external or internal fermionic line, in eq. (51)-(54) it is easy to trace back the external cross contribution and the internal 5

cross one (the latter proportional to mf /m). On the other hand, in graph V the chirality ?ip is provided by the only scalar interaction with the fermion.

3.1

QED

In this section we consider all the diagrams involving QED hypermultiplet ?elds. The one loop Standard Model contribution [11] simply comes from graph IV with a massless photon in the loop, with g+ = g? = e: α 2π

aγ ? = 3.1.1 KK vectors

(11)

The interaction lagrangian for KK photons has been displayed in (9) and the graphs are of type IV. As a general rule, dealing with the diagrams involving KK ?elds, it is useful to separate the external and internal cross contributions. This is motivated by the fact that in the latter case it is necessary to use couplings and loop masses at order ? while in the former O(1) is enough. The following sketched diagrams are drawn in terms of Weyl fermions and the emitted zero-mode photon is implicitly attached to any internal charged line. The states in the loop are KK modes of the ?eld indicated, always carrying the same index n for the reasons explained above. The resulting ?a? , for a given internal KK state, are:

?

?

and the trivial integration over the Feynman parameter x is left explicit in order to discuss supersymmetric cancellations (see below). 3.1.2 Scalars

? ?

?

?

?

?

?2 e2 16π 2 (2n)2

dx 4x(1 ? x)2 ? 8x(1 ? x)

(12)

?

?

?

e2 ?2 16π 2 (2n)2

dx 8x(1 ? x) ? 16x(1 ? x)2

(13)

The gauge scalar ?eld A5 mixes mode by mode with the massive vector Aρ . Including also the mass term for the gauge scalar σγ , the mass lagrangian, for each mode n, is:

6

Lmass = ??ρ A

ρ

2n A5 R

1 ? 2

2n R

2 2 σγ

1 ? 2ξ

2n ?ρ A ? ξ A5 R

ρ

2

(14)

where the Rξ gauge ?xing term is there to cancel the mixing terms. It is convenient to work in the ξ = 1 gauge: we obtain a photon propagator with no longitudinal modes and two scalars σγ and A5 (with equal mass) that can be combined in the complex scalar Σγ . The interaction terms are then:

so that at the relevant order in ?, only an external cross graph is possible. Equivalently, in terms of the real scalar ?elds it happens that the internal cross diagram with an A5 ?eld is cancelled out by the same diagram with a σγ . The external cross diagram is of type II so that the resulting ?a? is

?

3.1.3

Fermions

The gauge fermions γ ? and γ ?c are mixed by the kinetic term along the 5th direction. They are involved in the following interaction terms √ 1 ?n √ ? 2e?0 γ 2 √ 1 ?c,n √ + 2e?c 0γ 2 ? ?+ Φ?? n + Φn 2n ? 1 ? + Φ+ ? Φ? n 2n ? 1 n

?

?

√

c c 2e ψE Σγ ψE + ψL Σγ ψL + h.c.

(15)

?

?

?2 e2 16π 2 (2n)2

dx 4x(1 ? x)2

(16)

??

?n and ?0 with γ ?c,n . and in similar expressions (replacing Φ with Φc ) for the interactions of ?c 0 with γ From graph II, one gets ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

?

?

? ?

¨?

?

?+ + Φ?? n ? Φn ? Φ+ n + Φn

(17)

?

¨
?

?

??

e2 ?2 16π 2 (2n ? 1)2

dx 4x(1 ? x)2 × 2

(18)

?

7

?

?

3.1.4

Total QED contribution

First of all, in order to test the computation, we consider the supersymmetric limit in which both fermions and bosons have the same parities and therefore the same bulk mass. In this case, we ?nd that the total contributions of the KK modes vanishes, as expected. Of course, the appearance of zero modes for the matter and gauge partners also cancels the SM result. Then, we collect all the QED contributions and we obtain:

2 e2 2π ( mR ) 16π 2 18

? ?

¨?

?
?

?

¨
?

? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

?

e2 ?2 16π 2 (2n ? 1)2

dx 8x(1 ? x)2 × 2

(19)

?

?

?

= ?aQED ?

(20)

3.2

Neutral Weak sector

In this section we consider all the diagrams involving Z hypermultiplet ?elds and neutral Higgs. We de?ne 1 c 1 ? + s2 2

gL

=

g

gR = ?

s2 g c

The one loop SM result [12] comes from graph IV with a zero mode Z boson and couplings g? = gL , g+ = ?gR : aZ ? = g2 m2 4s4 ? 2s2 ? 1 2 48π 2 MZ c2 (21)

The discussion for the KK modes mimics the QED case, thus in the following we only point out the di?erences. 3.2.1 KK vectors

The interaction lagrangian consists in 8

1 1 ? ? ? c? + + + ρ ρ ?n Zn ?n + jL j + ?c + ?c ?0 σ ρ √ jL n + jR ?n Zn + h.c. 0 σρ √ 2 2 R where:

(22)

The two ?a? then amount to

?

3.2.2

? ?

?

? gL + gR 2(2n) ? ? = gL ? jL gL + gR 2(2n)

+ = gL + jL

? gR + gL 2(2n) ? ? gR + gL = ?gR + jR 2(2n)

+ = ?gR ? jR

(23)

?

?

?

?2 1 2 16π (2n)2

2 2 dx(gL + gR ) 2x(1 ? x)2 ? 4x(1 ? x)

(24)

?

?

?

??

?2 1 16π 2 (2n)2

dx

1 gL gR + (gL + gR )2 8x(1 ? x) 2

? gL gR 16x(1 ? x)2

(25)

Scalars

With respect to the QED case (14), the lagrangian for scalar ?elds is a bit more complicated, involving also mixing terms with ?H 0 , the imaginary part of φH 0 . With the gauge ?xing piece already included, the lagrangian for each mode n is:

2 2 2 σZ 2

Lmass = ??ρ Z ρ

2n Z5 ? MZ ?H 0 R

?

1 2n ? 0 ? MZ Z5 2 R H ?

1 1 2 2 ?H 0 ? ? MZ 4 2

2n R

1 ?ρ Z ρ ? ξ 2ξ

2n Z5 ? MZ ?H 0 R

(26)

Vector-scalar mixings are again cancelled and, for ξ = 1, there are no o?-diagonal terms ?H 0 Z5 , so the three real scalars σZ , Z5 , ?H 0 are the mass eigenstates. The ?rst two combine in the complex ?eld ΣZ in terms of which, at the relevant order in ?, only the external cross graph is possible:

9

?

3.2.3

Fermions

c , for each mode n, is: The mixing matrix for the four neutralinos z ?, z ?c , ψH 0 , ψH 0

?

?

?

?

?2 1 16π 2 (2n)2

2 2 dx(gL + gR ) 2x(1 ? x)2

(27)

??

Lmass = (ψH 0 , z ?)

2n?1 R

MZ

2n?1 R

?MZ

c ψH 0 z ?c

(28)

They are involved in the following interaction terms: ? 1 + ?+ (Φ?? + Φ? gL (Φ?? 2?0 z ?n √ n ) + n ? Φn ) + (gL ? gR ) 2(2n ? 1) n 2 √ 1 + 2 ?c ?n √ gL ? gR and Φ ? Φc + 0z 2 √ c c 1 ? + ?n √ + 2 ?0 z (Φ? ? Φ+ gR (Φ? n) + n + Φn ) ? (gR ? gL ) 2(2n ? 1) n 2 √ c 1 √ gL ? gR and Φ ? Φc + ?n + 2 ?0 z 2 1 ? + c 1 c? c+ c ?0 √ (Φ? +2 n + Φn ) + ?0 √ (Φn + Φn ) ψH,n v 2 2 √

(29)

The external cross graphs are readily computed: ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?

?

?

and it is possible to write three internal cross diagrams, using also the o?-diagonal propagator in eq. (28):

? ?

¨?

?

?

¨
?

?

??

1 ?2 16π 2 (2n ? 1)2

2 2 dx(gL + gR ) 2x(1 ? x)2 × 2

(30)

?

10

?

?

?

3.2.4

Again, as in case of QED, one can check that in the supersymmetric limit, discussed above, the contribution of the KK modes vanishes. Moreover, in the limit gL = ?gR = e, the QED results are recovered. Then, we sum up all the results of the neutral weak sector and obtain: g2 1 (mR)2 2 2 16π c 1 s2 2s4 ? + 4 3 3 π2 12

? ? ?

¨?

?
?

?

? ?

1 ?2 16π 2 (2n ? 1)2

2 2 dx (gL ? gR )2x(1 ? x) + gL gR 8x(1 ? x)2 (31)

¨
?

?

? ?

?

?

1 ?2 16π 2 (2n ? 1)2

2 2 dx (gR ? gL )2x(1 ? x) + gL gR 8x(1 ? x)2 (32)

¨
?

?

??

?

? ?

?2 1 16π 2 (2n ? 1)2

g dx (gL + gR ) 4x(1 ? x) c

(33)

Total neutral weak contribution

?aZ ? =

(34)

3.3

Charged Weak sector

In this section we consider all the diagrams involving W ± hypermultiplets ?elds and charged Higgs. The SM result [13] comes from graph III, where the loop contains a zero mode W boson and a neutrino √ and g+ = g/ 2, g? = 0: 5g2 m2 2 96π 2 MW

aW ? = 3.3.1 KK vectors

(35)

The interaction terms involving the KK vectors are quite simple g g ? ρ ρ c √ Wn ?0 σρ ψ ν,n ? √ Wn ?0 σρ ψ ν,n 2 n 2 2 2 and the external and internal cross diagrams are readily computed using graph III : 11 (36)

?

?

There also are diagrams of type V. The only nonvanishing one has a Higgs scalar, since the contributions of W5 and σW cancel out:

3.3.2

Scalars

The gauge ?xing procedure works as in the Z ’s case: the interaction lagrangian terms only allow for an external cross contribution:

c + c ? g ψν Σ ψL + ψL Σ ψν + h.c.

3.3.3

Fermions

? , ψ + , ψc The charginos w ?? , w ?c H H + mix for each n according to:

? ? ? ?

??

?

? ?

?2 g2 16π 2 (2n)2

dx (1 ? x)2 (x ? 3/2)

(37)

??

?

?

?2 g2 16π 2 (2n)2

3 dx (1 ? x)2 2

(38)

?

?·

??

?

?

g2 ?2 16π 2 (2n)2

dx (1 ? x)2

(39)

?

?· ?

?

??

?2 g2 2 16π (2n)2

dx x(1 ? x)2

(40)

Lmass =

? ?c ψH + , w

? 2MW

2n?1 √R

√

2MW 2n?1 R

c ψH + w ??

?

2n ? 1 + + w ?c w ? R

(41)

12

They enter the following interaction terms ? φc? w ? ? ?c + 2(2n ? 1) ν,n n 0 ? + c ? φν,n w ?c,n ?0 gφc ?c,n ?0 + g ν,n w 2(2n ? 1) 2m c c ? ψ + φc? ? v 0 H ,n ν,n The external cross diagrams are of the usual form ? ? ? ? ?· ?
? ? ? ? ? ? ? ? ? ? ? g2 ?2 ? dx x(1 ? x)2 × 2 2 (2n ? 1)2 ? 16 π ? ? ? ?
? ?
? ? ? ? ? ? ? ? ? ?

+ gφ? ?n ?0 ? g ν,n w

+ , (w ? )? — w ? )? — and there are three internal cross diagrams with propagators (w ? + )? — w ?c ?c ?? and (w ?c φc H+ :

?

?

?

3.3.4

Total charged weak contribution

Again in this case, we correctly ?nd that in the supersymmetric limit the sum of the KK states contributions vanishes. 13

? ? ? ? ?

· ?

?· ??

(42)

(43)

?

?

?2 g2 16π 2 (2n ? 1)2

dx (1 ? x)2

(44)

? ?

?

? ?

?2 g2 16π 2 (2n ? 1)2

dx (1 ? x)2

(45)

? ?

?·

?

? ?

g2 ?2 16π 2 (2n ? 1)2

dx 4(1 ? x)2

(46)

We then gather all the contributions of the charged weak sector into the result: ?aW ? =?

2 g2 27 π ( mR ) 16π 2 6 12

(47)

3.4

Result to leading order in (MW R)2

Collecting all partial results (20), (34) and (47) we have: ?aKK =? ? g2 m2 11 ? 18s2 (MW R)2 2 192 MW 12c2

(48)

Numerically, for 1/R = 370 ± 70 GeV , it is

.06 .6 weak ?10 ?aKK = 0.07 +0 = ?(1.1 +0 , ? ?0.02 · a? ?0.3 ) · 10

which means that, for any sensible value of R (see [14]), ?aKK is well inside the uncertainties that ? a?ect the SM prediction.

4

MW R corrections

In the previous section, we have neglected the ElectroWeak Symmetry Breaking in the masses of the KK modes for the gauge ?elds, since this was enough to get eq. (48). However, in this way we are neglecting contributions that in principle can be quite relevant, since the dimensionless expansion parameters MW R or MZ R can be as large as 1/3. In this chapter we estimate these contributions. They are recovered if the full gauge ?elds masses are kept in the denominators of eq. (51)–(55), so the corrections to our previous results can be expressed in a simple way. Of course, these corrections only apply to the pure weak sector. To begin with, let us focus on the neutral weak sector. First of all, what we have to do is to study the exact spectrum: de?ning the dimensionless parameter ζ = MZ R, for every n there is a vector boson Z? whose full mass squared is (2n)2 + ζ 2 , in unit of 1/R2 . The scalars ?elds masses are determined from eq. (26): the nth KK mode of Σz has mass squared (2n)2 + ζ 2 , while, for the real ?eld ?H 0 , it 3 has mass squared (2n)2 + ζ 2 . For the neutralinos, diagonalization of eq. (28) shows that there are 2 two degenerate states with mass squared (2n ? 1)2 + ζ 2 . The rest of the computation proceeds as in 1 the previous section. To get the modi?ed results, it is su?cient to substitute the sum in eqs. (2n)2 (24), (25), (27), (30)–(33) with: 1 N2 or N 2 + ζ2 (N 2 + ζ 2 )2 where N = 2n for bosons and N = 2n ? 1 for fermions. The ?rst is the one to be used in all external cross diagrams, while for internal cross ones the formulae are a bit more involved. To be speci?c, in the case of the vectors, eq. (25) becomes: 14

?

1 2 ? 16π 2

dx

(2n)2 1 1 2 ? g g 16 x (1 ? x ) gL gR + (gL + gR )2 8x(1 ? x) L R 2 (2n)2 + ζ 2 ((2n)2 + ζ 2 )2

(49)

With a bit of algebra, all contributions can be recast in a common form and the net result is: 1 m2 2 2 2 (gL + gR ) f1 (ζ ) + gL gR f2 (ζ ) 2 (MW R) 16π 2 MW

?aZ ? =

(50)

where the functions f1 and f2 are de?ned in app. B. This replaces eq. (34), with all corrections to any order in (MW R)2 included. Numerically, for 1/R = 370 ± 70 GeV , the result is only about 3% smaller than eq. (34). In the case of the charged weak sector, the calculation is very similar, so we expect the correction to be not so far from a few percent.

5

Conclusion

We have performed the calculation at one loop of the corrections to muon anomalous magnetic moment coming from the presence of one extra dimension in the model in ref. [1], to ?rst order in (mR)2 and in (MW R)2 . We have also shown that the complete calculation in (MW R)2 yields small deviations, so to our purposes the simple analytic form given in eq. (48) is a good approximation. The corrections are relatively small, for any sensible value of R, at the 10% level of the pure weak contribution in the SM, and well inside the uncertainties of the hadronic contribution.

Appendix A

In this appendix we list the contributions to a? from the ?ve graphs2 in ?gure 1:

1 2 h2 mf 1 + + h? 2x(1 ? x)2 + h+ h? 2x(1 ? x) 2 2 2 m xmf + (1 ? x)m2 s ? x(1 ? x)m

?aI ? =

Qs 2 m 16π 2

dx

0

(51)

?aII ? =?

Qf 2 m 16π 2

1

dx

0

2 h2 mf + + h? 2x(1 ? x)2 + h+ h? 2(1 ? x)2 2 m

xm2 s

2

1 2 + (1 ? x)m2 f ? x(1 ? x)m

(52)

Similar results can also be found in refs. [15].

15

?aIII ? =

Qv 2 m 16π 2

1

dx

0

2 + g2 g+ mf ? 2(1 ? x)2 (2x ? 3) + g+ g? 6(1 ? x)2 2 m

xm2 f

1 2 + (1 ? x)m2 v ? x(1 ? x)m

(53)

?aIV ? =?

Qf 2 m 16π 2

1

dx

0

2 + g2 g+ mf ? g+ g? 8x(1 ? x) 4x(1 ? x)2 ? 8x(1 ? x) + 2 m 1 2 2 xmv + (1 ? x)m2 f ? x(1 ? x)m 1 0

(54)

?aV ? =

Qv mτ 16π 2

dx(g+ h+ + g? h? )(1 ? x)2

xm2 f

1 2 + (1 ? x)m2 v ? x(1 ? x)m

(55)

where Qs,f,v are the electric charges of the particles emitting the zero-mode photon and ms,f,v are the masses in the loop. Note that in eq. (55) we assume ms = mv . The couplings are de?ned by: vρ f γ ρ [g+ P + + g? P ? ] ? sf [h+ P + + h? P ? ] ?

? Aρ eτ s+ Wρ

+ W ?A e [gνρ (k ? p+ )σ + gρσ (p+ ? p? )ν + gσν (p? ? k)ρ ] Wρ σ ν

where P ± are projectors on the two Weyl components of the Dirac ?elds.

Appendix B

In this appendix we list the functions to be inserted in eq. (50):

ζ 1 π π dy 2(ζ 4 ? y 4 ) πy 2 tanh y ? πy 2 coth y + 2y = 8 ζ 0 2 2 1 4 + 6 8 48π 5 ζ 5 Li2 ?e?πζ ? Li2 e?πζ = + 336π 4 ζ 4 Li3 ?e?πζ ? Li3 e?πζ 2 3ζ 3π ζ +1440π 3 ζ 3 Li4 ?e?πζ ? Li4 e?πζ + 4320π 2 ζ 2 Li5 ?e?πζ ? Li5 e?πζ +

f1 =

+8640πζ Li6 ?e?πζ ? Li6 e?πζ ?42π 4 ζ 4 ζ (3) + 17145 ζ (7)

+ 8640 Li7 ?e?πζ ? Li7 e?πζ

+

16

f2 =

ζ π π 1 dy 8(y 2 ? ζ 2 )(3y 2 ? 2ζ 2 ) πy 2 tanh y ? πy 2 coth y + 2y = 8 ζ 0 2 2 4 4 + 136π 4 ζ 4 Li3 ?e?πζ ? Li3 e?πζ ? 6 8 8π 5 ζ 5 Li2 ?e?πζ ? Li2 e?πζ = 2 ζ π ζ + 3840π 2 ζ 2 Li5 ?e?πζ ? Li5 e?πζ + +960π 3 ζ 3 Li4 ?e?πζ ? Li4 e?πζ

+

+8640πζ Li6 ?e?πζ ? Li6 e?πζ

+ 8640 Li7 ?e?πζ ? Li7 e?πζ

+

+28π 4 ζ 4 ζ (3) ? 930π 2 ζ 2 ζ (5) + 17145 ζ (7) where ζ (n) is the Riemann zeta function, not to be confused here with our variable ζ = MZ R. In the integrals, the hyperbolic tangent is produced by the sum over the fermionic KK states, while the cotangent and the “2y ” pieces by the bosonic states.

Acknowledgements

We would like to thank Riccardo Barbieri and Riccardo Rattazzi for useful discussions and suggestions. We are also grateful to Gilberto Colangelo for informations about the present status of SM (g ? 2)? . This work was supported by the EC under the RTN contract HPRN-CT-2000-00148.

17

References

[1] R.Barbieri, L.J.Hall, Y.Nomura, Phys.Rev.D63:105007,2001, arXiv:hep-ph/0011311; [2] R.Barbieri, G.Cacciapaglia, A.Romito, arXiv:hep-ph/0107148; [3] G.Cacciapaglia, M.Cirelli, G.Cristadoro, arXiv:hep-ph:0111287; [4] H.N.Brown et al. (Muon g-2 Collaboration), Phys.Rev.Lett. 86:2227-2231,2001, arXiv:hepex/0102017; A.Czarnecki, W.J.Marciano, Phys.Rev. D64:013014,2001, arXiv:hep-ph/0102122; [5] P.Nath, M.Yamaguchi, Phys.Rev. D60:116006,1999, arXiv:hep-ph/9903298; R.Casadio, A.Gruppuso, G.Venturi, Phys. Lett. B495:378, 200, arXiv:hep-th/0010065; Seong Chan Park, H.S.Song, Phys. Lett. B506:99,2001, arXiv:hep-ph/0103072; K.Agashe, N.G.Deshpande, G.-H.Wu, Phys.Lett. B511:85,2001, arXiv:hep-ph/0103235; T.Appelquist, B.A.Dobrescu, Phys.Lett B516:85,2001, arXiv:hep-ph/0106140; [6] J.Prades, invited talk at “Kaon 2001”, 6-13 June 2001, Pisa, Italy, arXiv:hep-ph/0108192; [7] for a recent review: W.J.Marciano, B.L.Roberts, arXiv:hep-ph/0105056; [8] M.Knecht, A.Ny?eler, arXiv:hep-ph/0111058; M.Knecht, A.Ny?eler, M.Perrottet, E.de Rafael, arXiv:hep-ph/0111059; [9] I.Blokland, A.Czarnecki and K.Melnikov, arXiv:hep-ph/0112117; [10] C.A.Scrucca, M.Serone, L.Silvestrini, F.Zwirner, arXiv:hep-th/0110073; [11] J.Schwinger, Phys.Rev. 73:416-417,1948; Phys.Rev. 76:790-817:1949; [12] G.Altarelli, N.Cabibbo, L.Maiani, Phys.Lett. B40:415,1972; [13] S.J.Brodsky, J.D.Sullivan, Phys.Rev. 156:1644,1967; [14] R.Barbieri, L.J.Hall, Y.Nomura, arXiv:hep-ph/0110102; [15] I.Vendramin, Nuovo Cim. A101:731,1989; K.R.Lynch, arXiv:hep-ph/0108080 and arXiv:hep-ph/0108081 .

18

赞助商链接

- Improved Determination of alpha_QED(M_Z^2) and the Anomalous Magnetic Moment of the Muon
- The Muon Anomalous Magnetic Moment and the Standard Model
- Anomalous magnetic moment of muon in 3-3-1 models
- The anomalous magnetic moment of the muon in the D-brane realization of the Standard Model
- Muon anomalous magnetic dipole moment in the $mu$-problem solvable extra U(1) models

更多相关文章：
更多相关标签：