Drift effect on vacuum birefringence in a strong electric and magnetic field*
Huang Yong-Sheng†, Wang Nai-Yan, Tang Xiu-Zhang
High Power KrF Excimer Laser Laboratory, China Institute of Atomic Energy, Beijing 102413, China

Corresponding author. E-mail: huangyongs@gmail.com

*Project supported by the National Basic Research Program of China (Grant No. 2011CB808104) and the National Natural Science Foundation of China (Grant No. 11105233).

Abstract

As an important QED effect to detect the vacuum polarization, birefringence in the presence of a strong electric and magnetic field, E0 B0, E0 cB0, is considered. The directional dependence of birefringence is obtained. In two special cases: E0 = 0 and E0 = cB0, our results are consistent with the previous ones. The refractive index of the probe wave propagating in the E0 × B0 direction decreases with E0/ cB0, while that in the − E0 × B0 direction increases with E0/ cB0. The physics of the direction dependence of birefringence maybe the E0 × B0 drift velocity of the virtual electrons and positrons.

Keyword: 42.25.Bs; 42.50.–p; 03.70.+k; 42.50.Pq; vacuum birefringence; critical Schwinger field; refractive index; E × B drift
1. Introduction

In the strong field, , B0 ≲ 4.3 × 109  T, which is close to the critical Schwinger field of electron– positron pair production, [14] the vacuum exhibits properties of nonlinear optical medium, [57] and leads to photon splitting, [8] dichroism, [9] birefringence, [1014] and vacuum bragg scattering.[15] The results of PVLAS collaboration[13, 14] are so close to the polarization rotation in a vacuum with a rotating transverse magnetic field, predicted by Adler, [10] Heyl and Hernquist.[11] Biswas and Melnikov[12] pointed out that the rotation of a magnetic field does not impact the birefringence. Adler also calculated the refractive indices of the probe wave in a rotation magnetic field and probed the dependence of the refractive indices on the frequency of the magnetic field. However, if there is a strong electric field, E0, perpendicular to the strong magnetic field, B0, the propagating characteristics of the wave in the E0 × B0 direction and in the − E0 × B0 direction are quite different from each other. For E0 = cB0, several researchers[1620] obtained the refractive index for the “ head-on collision" probe wave (Eqs.  (16) and (17) in Ref.  [19] and Eq.  (15) in Ref.  [20]). Their results also showed that there is no interaction between the probe wave incident in the E0 × B0 direction, i.e., the refractive index is unit.

In general, in the field (E, B) of a tightly focused low-frequency laser, | E| < c| B| . If the laser frequency is smaller than 1 × 1020  rad/s, the scaling length of the field is much larger than the Compton wavelength of electron. In that case, the field can be equivalent to the homogenous field and the effective Lagrangian 𝔏 eff[5] is available. Therefore, birefringence in a crossed field (E0B0 and E0cB0) should be reconsidered. In order to check birefringence in a more general regime, we consider the interaction between a high-frequency electromagnetic wave and a strong electromagnetic field: , , and E0cB0, where and are the unit vectors of the x and y directions, respectively.

2. Basic assumptions and theoretical model

In this paper, we show that the refractive index of a forward wave with the wave vector kE0 × B0 is different from that of a counter propagating wave in a vacuum with a strong homogeneous static electric and magnetic field, solving the Maxwell equations corrected by the effective Lagrangian[5]

where E and B is the total electric and magnetic field respectively, using

and

which are valid for | E| , c| B| ≪ Ecrit ≈ 1018  V/m and the frequency of the field ω ω e ≈ 1020  rad/s. The quantum parameter, , where ε 0 is the permittivity of vacuum, , me is the electron static mass, and α is the fine-structure constant.

First, the external field is assumed to be and , where and are the unit vectors of the x and y directions respectively, E0cB0 ≲ 1018  V/m, which is the critical Schwinger field of electron– positron pair production in a vacuum.

Let E = E0 + Eq and B = B0 + Bq be the total electromagnetic field, and Eq and Bq are the nonclassical parts which are expressed as

Let | E1| , | E2| , | cB1| , | cB2| ≪ | E0| , and | cB0| , where c is the light velocity. The wave vector . Using the effective Lagrangian 𝔏 eff, [5] the QED corrected electric displacement field Dq = Dε 0α 0E0 and magnetic field strength are given by

where α 0 = 1 + 2ξ 0, c0 = E0/B0, β 0 = c0/c, and . In fact, c0 is just the drift velocity of the virtual electrons and positrons in the field.

With the wave assumptions that E1, 2, B1, 2 ∝ exp[i(kzω t)], the linearization of the corrected source-free Maxwell equations

gives

3. Dispersion relationship and the refractive indices

With Eq.  (8), the dispersion relationship for parallel polarization is obtained after some straightforward algebra

With Eq.  (7), the dispersion relationship for perpendicular polarization is given by

Let n = ck/ω be the refractive index. With Eqs.  (9) and (10), the refractive index of a probe wave propagating in the positive z direction, i.e., in the E0 × B0 direction for parallel polarization and perpendicular polarization, is respectively given by

and

Similarly, the refractive index of a probe wave propagating in the negative z direction, i.e., in the − E0 × B0 direction for parallel polarization and perpendicular polarization, is respectively given by

and

With Eqs.  (11) and (14), the difference of the refractive indices for the parallel polarization and the perpendicular polarization in the E0 × B0 direction is obtained as

In the − E0 × B0 direction, it is obtained as

Equation  (15) shows that the difference of the refractive index for the parallel polarization wave and the perpendicular polarization wave is proportional to ξ . Here the wave propagates in the positive z direction, i.e., E0 × B0. Equation (16) shows the refractive index for the parallel polarization wave and the perpendicular wave propagating in the negative z direction, i.e., − E0 × B0. The results predict the difference of the refractive indices is proportional to ξ . Figure  1 shows the drift effect on vacuum birefringence in a homogenous strong crossed field. The birefringence is amplified by the drift effect for the counter-propagating probe wave and is reduced for the probe wave propagating in the direction of E0 × B0.

Fig.  1. Drift effect on vacuum birefringence in a homogenous strong crossed field.

4. Comparison and discussion

To compare our results with that of the literatures, two special cases are considered: (a) E0 = 0, i.e., vacuum birefringence in the presence of a homogenous magnetic field; (b) E0 = cB0 in a crossed field.

For E0 = 0, c0 = β 0 = 0, 0 = − 1, and the refractive indices are

for parallel polarization, and

and the difference of the refractive indices is

All the results are the same as those of Refs.  [10]– [13]. Therefore, in a homogenous magnetic field, there is no drift velocity of virtual particles and there is no direction dependence of the refraction.

Another well-known case is E0 = cB0, β 0 = 1, and c0 = c. Therefore, the refractive indices in the E0 × B0 direction are trivial and n+ , par = n+ , per = 1. There is no interaction between the probe wave and the crossed field. It is also consistent with Eqs.  (16) and (17) in Ref.  [19] and Eq.  (15) in Ref.  [20] in the E0 × B0 direction.

However, for the counter-propagating probe wave, the refractive index is given by

for parallel polarization, and

which is also consistent with Eqs.  (16) and (17) in Ref.  [19] and Eq.  (15) in Ref.  [20] in the − E0 × B0 direction.

The refractive indices of the probe wave in the E0 × B0 direction and in the − E0 × B0 direction are different from each other. The difference of the refractive indices for parallel polarization and perpendicular polarization are also different in the two directions. Define

which tends to be zero as E0 → 0.

The physics of the direction dependence of the refractive index maybe the E0 × B0 drift of the virtual electrons and positrons in the strong field. The virtual electrons and positrons have the same E0 × B0 drift velocity, . Therefore, it induces the direction dependence of refraction. Equation  (22) shows that the difference is proportional to the drift velocity.

In the experiment to measure the birefringence, the light is reflected and propagates through the field forward and backward repeatedly for N times. Set the longitudinal length of the field as L, the elipticity value[14] is given by

which is times that in a strong magnetic field.

5. Conclusion

In conclusion, vacuum birefringence is different in the E0 × B0 direction and in the − E0 × B0 direction. The E0 × B0 drift of the virtual electrons and positrons may be the key point. The influence of the drift effect on the ellipticity in the experiments has been given explicitly.

Acknowledgment

We are grateful to Felix Karbstein for the interesting and enlightening discussion.

Reference
1 Abdukerim N, Li Z L and Xie B S 2013 Phys. Lett. B 726 820 DOI:10.1016/j.physletb.2013.09.014 [Cited within:1] [JCR: 0.479]
2 Tang S, Xie B S, Lu D, Wang H Y, Fu L B and Liu J 2013 Phys. Rev. A 88 012106 DOI:10.1103/PhysRevA.88.012106 [Cited within:1] [JCR: 3.042]
3 Su Q, Su W, Z Q, Jiang M, Lu X, Sheng Z M and Grobe R 2012 Phys. Rev. Lett. 109 253202 DOI:10.1103/PhysRevLett.109.253202 [Cited within:1] [JCR: 7.943]
4 Xie B S, Mohamedsedik M and Dulat S 2012 Chin. Phys. Lett. 29 021102 DOI:10.1088/0256-307X/28/2/021102 [Cited within:1] [JCR: 0.811] [CJCR: 0.4541]
5 Heisenberg W and Euler H 1936 Z. Physik 98 714 DOI:10.1007/BF01343663 [Cited within:4] [JCR: 1.407]
6 Bialynicka-Birula Z and Bialynicki-Birula I 1970 Phys. Rev. D 2 2341 http://www.ncbi.nlm.nih.gov/pubmed/9957939 [Cited within:1] [JCR: 5.156]
7 Boillat G 1970 J. Math. Phys. 11 941 DOI:10.1063/1.1665231 [Cited within:1] [JCR: 1.296]
8 Adler S L and Schubert C 1996 Phys. Rev. Lett. 77 1695 DOI:10.1103/PhysRevLett.77.1695 [Cited within:1] [JCR: 7.943]
9 Zavattini E, Zavattini G, Ruoso G, Polacco E, Milotti E, Karuza M, Gastaldi U, Di Domenico G, Della Valle F, Cimino R, Carusotto S, Cantatore G and Bregant M 2006 Phys. Rev. Lett. 96 110406 DOI:10.1103/PhysRevLett.96.110406 [Cited within:1] [JCR: 7.943]
10 Adler S L 2007 J. Phys. A: Math. Theor. 40F143 DOI:10.1088/1751-8113/40/5/F01 [Cited within:3] [JCR: 1.766]
11 Heyl J S and Hernquist L 1997 J. Phys. A: Math. Gen. 30 6485 DOI:10.1088/0305-4470/30/18/022 [Cited within:1] [JCR: 1.577]
12 Biswas S and Melnikov K 2007 Phys. Rev. D 75 053003 DOI:10.1103/PhysRevD.75.053003 [Cited within:1] [JCR: 5.156]
13 Zavattini E, Zavattini G, Ruoso G, Polacco E, Milotti E, Karuza M, Gastaldi U, Di Domenico G, Della Valle F, Cimino R, Carusotto S, Cantatore G and Bregant M 2007 Nucl. Phys. B Proc. Suppl. 164 264 DOI:10.1016/j.nuclphysbps.2006.11.096 [Cited within:2]
14 Della Valle F, Gastaldi U, Messineo G, Milotti E, Pengo R, Piemontese L, Ruoso G and Zavattini G 2013 New J. Phys. 15 053026 DOI:10.1088/1367-2630/15/5/053026 [Cited within:3] [JCR: 4.063]
15 Kryuchkyan G Y and Hatsagortsyan K Z 2011 Phys. Rev. Lett. 107 053604 DOI:10.1103/PhysRevLett.107.053604 [Cited within:1] [JCR: 7.943]
16 Narozhnyi N B 1969 JETP 28 371 [Cited within:1] [JCR: 1.524]
17 Baier R and Breitenlohner P 1967 Acta Phys. Austriaca 25 212 [Cited within:1]
18 Baier R and Breitenlohner P 1967 Nuovo Cim. B 47 117 DOI:10.1007/BF02712312 [Cited within:1] [JCR: 0.225]
19 Heinzl T, Liesfeld B, Amthorb K U, Schwoererb H, Sauerbreyc R and Wipfd A 2006 Opt. Commun. 267 318 DOI:10.1016/j.optcom.2006.06.053 [Cited within:3] [JCR: 1.438]
20 Brezin E and Itzykson C 1971 Phys. Rev. D 3 618 DOI:10.1103/PhysRevD.3.618 [Cited within:4] [JCR: 5.156]