03.07.2013 Views

Proceedings of International Conference on Physics in ... - KEK

Proceedings of International Conference on Physics in ... - KEK

Proceedings of International Conference on Physics in ... - KEK

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

the dispersi<strong>on</strong> relati<strong>on</strong> p0 = Ep − µ for particles). Equati<strong>on</strong><br />

(8) implies that eA1γ1 = eA1γ0γ 5 and thus −eA1 can<br />

be identified as the axial (or chiral) chemical potential µ5.<br />

Therefore, we can c<strong>on</strong>clude;<br />

J 1 V = 1<br />

π<br />

∫<br />

dx µ5, (17)<br />

which correctly recovers the (3+1)-dimensi<strong>on</strong>al chiral<br />

magnetic current (4) <strong>on</strong>ce we multiply this by the Landau<br />

level density, eB/(2π). That is,<br />

jV = µ5<br />

π<br />

−→ jV = |eB|<br />

2π<br />

(<strong>in</strong> (1+1) dimensi<strong>on</strong>s)<br />

· µ5<br />

π<br />

(<strong>in</strong> (3+1) dimensi<strong>on</strong>s), (18)<br />

which co<strong>in</strong>cides with Eq. (4).<br />

Here, it is clear that the l<strong>on</strong>gitud<strong>in</strong>al gauge field A 1 ,<br />

which is the Chern-Sim<strong>on</strong>s number <strong>in</strong> (1+1) dimensi<strong>on</strong>s,<br />

plays the role <str<strong>on</strong>g>of</str<strong>on</strong>g> the chiral chemical potential µ5 <strong>in</strong> (3+1)<br />

dimensi<strong>on</strong>s. We note, however, that there is an important<br />

difference; usually µ5 is <strong>in</strong>troduced by hand as a c<strong>on</strong>stant,<br />

but <strong>in</strong> (1+1) dimensi<strong>on</strong>s A 1 must have t-dependence to allow<br />

for n<strong>on</strong>zero QW . We can th<strong>in</strong>k <str<strong>on</strong>g>of</str<strong>on</strong>g> a c<strong>on</strong>crete “<strong>in</strong>stant<strong>on</strong>”<br />

c<strong>on</strong>figurati<strong>on</strong> <strong>in</strong> (1+1) dimensi<strong>on</strong>s simply as<br />

A 1 (t, x) = 2πQW t<br />

= −Et, (19)<br />

eL T<br />

where we limit ourselves to the spatially homogeneous case<br />

and denote the spatial and temporal extents as L and T ,<br />

respectively, and then we have<br />

j 1 V (t) = J 1 V (t) eE<br />

= t. (20)<br />

L π<br />

From this, aga<strong>in</strong>, if multiplied by the Landau-level degeneracy<br />

we can correctly recover the current generati<strong>on</strong> rate<br />

given by Eq. (6), i.e.<br />

d(ejV )<br />

dt = e2E (<strong>in</strong> (1+1) dimensi<strong>on</strong>s)<br />

π<br />

−→ d(ejV ) |eB|<br />

=<br />

dt 2π · e2E (<strong>in</strong> (3+1) dimensi<strong>on</strong>s), (21)<br />

π<br />

which co<strong>in</strong>cides with Eq. (6).<br />

In the same way we can get a f<strong>in</strong>ite axial-vector current<br />

at f<strong>in</strong>ite quark chemical potential µ. To see the anomalous<br />

nature the important fact is that the relati<strong>on</strong> between the<br />

density and the chemical potential is given by the quantum<br />

anomaly <strong>in</strong> (1+1) dimensi<strong>on</strong>s, i.e.<br />

n = − eA0<br />

, (22)<br />

π<br />

which results from the anomaly. One can derive this expressi<strong>on</strong><br />

directly from n = ⟨ψ † (x)ψ(x)⟩ by <strong>in</strong>sert<strong>in</strong>g<br />

the gauge field as lim y 0 →x 0 ψ † (y) exp[−ie ∫ dtA 0 ]ψ(x).<br />

From this we can immediately reach,<br />

J 1 ∫<br />

5 = dx n = 1<br />

∫<br />

dx µ, (23)<br />

π<br />

which represents the chiral separati<strong>on</strong> effect. This is aga<strong>in</strong><br />

the anomaly relati<strong>on</strong> exactly same as that <strong>in</strong> (3+1) dimensi<strong>on</strong>s<br />

<strong>on</strong>ce multiplied by the Landau level density eB/2π.<br />

SCHWINGER MODEL<br />

So far the arguments and the result<strong>in</strong>g expressi<strong>on</strong>s are<br />

quite general. From now <strong>on</strong> we shall go <strong>in</strong>to the dynamical<br />

properties calculat<strong>in</strong>g microscopic quantities <strong>in</strong> a solvable<br />

(1+1)-dimensi<strong>on</strong>al model, i.e. the massless Schw<strong>in</strong>ger<br />

model. The easiest way to accomplish a calculati<strong>on</strong> <strong>in</strong> the<br />

Schw<strong>in</strong>ger model is to use mapp<strong>in</strong>g <strong>on</strong>to a free bos<strong>on</strong>ic<br />

theory. In our case, however, the bos<strong>on</strong>izati<strong>on</strong> rule is a<br />

bit more complicated than usual because we deal with not<br />

<strong>on</strong>ly fermi<strong>on</strong>ic fields (such as the chiral c<strong>on</strong>densate) but<br />

also gauge fields (such as the electric field). So, the Lagrangian<br />

density <str<strong>on</strong>g>of</str<strong>on</strong>g> the corresp<strong>on</strong>d<strong>in</strong>g theory should be<br />

L = 1<br />

2 (∂µ θ)(∂µθ) − mγ(∂ µ θ)(∂µϕ) − 1<br />

2 (∂µ ϕ)∂ 2 (∂µϕ)<br />

(24)<br />

with the bos<strong>on</strong> mass,<br />

m 2 γ = e2<br />

. (25)<br />

π<br />

If the ϕ-field is <strong>in</strong>tegrated out, we get a theory <strong>on</strong>ly <strong>in</strong> terms<br />

<str<strong>on</strong>g>of</str<strong>on</strong>g> the θ-field that is free and has a mass by mγ. Such a<br />

scalar theory is usually used with the bos<strong>on</strong>izati<strong>on</strong> rule;<br />

j µ<br />

V = ¯ ψγ µ ψ = 1<br />

√ π ϵ µν ∂νθ, (26)<br />

j µ<br />

5 = ¯ ψγ µ γ 5 ψ = − 1<br />

√ π ∂ µ θ, (27)<br />

¯ψψ = −c mγ : cos(2 √ πθ) : (28)<br />

with the normal order<strong>in</strong>g : :. Now we remark that ϕ <strong>in</strong><br />

the Lagrangian density (24) comes from the gauge field,<br />

A µ = −ϵ µν∂νϕ (where ϕ <strong>in</strong>cludes an <strong>in</strong>stant<strong>on</strong>-like c<strong>on</strong>figurati<strong>on</strong><br />

∼ 1<br />

2Et2 which does not satisfy the periodic<br />

boundary c<strong>on</strong>diti<strong>on</strong> <strong>in</strong> the t-directi<strong>on</strong>). Then the electric<br />

field takes a form E = ∂2ϕ. Once we <strong>in</strong>tegrate the θ-field<br />

out from the theory, after the Gaussian <strong>in</strong>tegrati<strong>on</strong> <strong>in</strong> the<br />

functi<strong>on</strong>al formalism, Eq. (27) is replaced by<br />

j µ<br />

5<br />

= − 1<br />

√ π ∂ µ θ → − mγ<br />

√π ∂ µ ϕ = − e<br />

π ∂µ ϕ. (29)<br />

The anomaly relati<strong>on</strong> is then derived as<br />

∂µj µ<br />

5<br />

= − e<br />

π ∂2 ϕ = − e<br />

π E = −2qW , (30)<br />

which is fully c<strong>on</strong>sistent with the anomaly relati<strong>on</strong> (11).<br />

In the same manner we can express the vector current <strong>in</strong><br />

terms <str<strong>on</strong>g>of</str<strong>on</strong>g> ϕ, and then we f<strong>in</strong>d,<br />

j µ<br />

V<br />

e<br />

=<br />

π ϵµν µν ∂ν<br />

∂νϕ = 2ϵ<br />

∂2 qW . (31)<br />

It is easy to c<strong>on</strong>firm that this result is fully c<strong>on</strong>sistent with<br />

the previous relati<strong>on</strong>. That is, after the spatial <strong>in</strong>tegrati<strong>on</strong><br />

for the µ = 1 comp<strong>on</strong>ent (or ϕ and qW ), the spatial derivative<br />

∂1 drops and the right-hand side simplifies as −2/∂0,<br />

that is just a t-<strong>in</strong>tegrati<strong>on</strong>. Therefore the right-hand side f<strong>in</strong>ally<br />

becomes −2QW together with the spatial <strong>in</strong>tegrati<strong>on</strong>,<br />

and hence we obta<strong>in</strong> J 1 V = −2QW .

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!