Physrevd 32 2235

Download as pdf or txt
Download as pdf or txt
You are on page 1of 2

PHYSICAL REVIEW D VOLUME 32, NUMBER 8 15 OCTOBER 1985

Rotating the Skyrme soliton


C6sar Gomez
Departamento de F;sica Tedrica, Universidad de Salamanca, Salamanca, Spain
(Received 29 March 1984; revised manuscript received 7 June 1985)
We study the effect on the Dirac sea of the topologically nontrivial 2a rotation of a soliton. As a conse-
quence of the index theorem, the Wess-Zumino-Witten term is induced by fermions with the correct quan-
tized coefficient.

In this Brief Report we wish to study the peculiar connec- topological arguments7 that 7r2( Ln 3 ( G ) ) = 7r5 ( G ) = Z.
tion found by Witten' between the fermionic number of the In our particular case something more specific can be said.
Skyrme soliton and the Wess-Zumino term T. Indeed, representing the soliton configuration as'
The soliton of a nonlinear u model with the Wess-
Zumino term included has a topological number given by
the 7 r 3 of the flavor group:

~-2s tr(r/aiu-lua,u-iua,u-~)d3x ,
=
V(X) 0
,I;
with V ( X ) in SU(2) the nontrivial sphere in Ln3(G) is al-
(3)

24r2 ways connected with a nontrivial path in SU(2) given by the


effect of the 27r rotation on U.
This topological number can be interpreted as the baryon L~~ now couple quarks to the pion field, ~h~ corre-
number using three different arguments. The simplest one sponding Dirac operator will be given by
is just the extension of the Goldstone-Wilczek technique2 to
the particular case of a soliton background-field configura-
tion. In fact, if we couple fermions to the pion field we get D = [ ~ a ~ €9
a , 1 - p €9 U ( x ) I = ~ a , ( x )
for the induced current j,= [$, y,+] the expectation value3
with U ( x ) the pion field. This operator induces a Fred-
(j,) = ~ ~ ~ *+ . . .~ , ~ (2) ~ holm~ operator
~ a ~ dimension
(the ~ ~ the kernel
of a ~ cokernel
and ~ is~ a ~
which is, when completely added, the topological charge de- finite) if the pion field satisfies the following condition:
fined in (1). Another possibility is to calculate the
anomalous contribution, to the baryon current. of the T I U ( x ) l Z B for I x I Z C , (5)
term.' Finally, a very interesting way to connect (1) with
the baryon number is with the use of a mixed bag-soliton with B and C two positive constants.*
model4 in which the bag baryon number is defined by the Now the path in R 3 ( S U ( 2 ) ) defined by the 2 n rotation
spectral asymmetry induced by the chiral bag boundary con- can be extended to a map from S1 into where we denote
ditions. In this approach one can see that the topological by F the space of Fredholm operators. It is clear that this
loop in 9 is noncontractible. As an application of the
number we lose by introducing a bag hole inside the soliton
configuration is recovered in the form of a spectral asym- Atiyah-Patodi-Singer theorem9 we can compute for a loop of
metry for the Dirac sea inside the bag. operators the spectral flow in terms of the ordinary index.
In general the index can be found by analyzing the homoto-
In order to find the statistics of these solitons a very nice
argument can be used. The 27r adiabatic rotation of a soli- py properties of the symbol of the ~ p e r a t o r . ' ~Using the
general notation given in (4) the symbol u ( D ) will be
ton is a nontrivial path in 7r~~q(SU(2)).The special role
given by
played here by SU(2) corresponds to the fact that the soli-
ton configuration is rotationally invariant. In terms of T it
is possible to prove that the 27r rotation always corresponds
to a nontrivial value of T when this one is computed on any and for our family we will have u ( D ) = a,(x, ?)ta, 2
five-dimensional manifold with the boundary being the 27r where q is a parameter on S1. In our case u ( D ) is then de-
rotation path in configuration space [we remind the reader fined on S 1 x S 3 and the homotopy class is again given by
that 7r5(SU ( N ) ) = Z, N Z 3 assures the independence of r 4 ( S U ( 2 ) ) = Z ( 2 ) , where we are using the identifications
this integral of the particular five-dimensional manifold - -
v 4 ( s 3 ) 7rl(S3,S3) and S3 SU(2). In a more mathemati-
used in the computation of T (see Ref. 5)1.6 cal language what we get is that the spectral flow is given by
The effect on the Dirac sea of the nonzero topological an element of KR (s') which is z (2)."
number (1) of our soliton configuration was what we used It is worth stressing here that the spectral flow associated
in order to give a physical interpretation to this topological with the loop of operators we get by a 2 n rotation of the
number. Now we have some new nontrivial topology, soliton configuration is in principle independent of the ex-
indeed, the 2 n rotation of the soliton. The problem we ad- istence of a Wess-Zumino term in the Lagrangian. In fact,
dress here is the effect on the Dirac sea of this topologically using the topology of our model, we can study its effect on
nontrivial path in the configuration space. fermions coupled to the system. In a more recent
Let us start with some mathematics. The 2 n rotation de- language1' we can refer to this phenomenon of spectral flow
fines a nontrivial two-sphere in n 3 ( G ), the space of maps in terms of "Hilbert twisted bundles." If we quantize the
from S3 into the group G. In fact, we know from general fermions on each background configuration U (x,,t ) we can

2235 @ 1985 The American Physical Society


2236 BRIEF REPORTS 32
-

define a Hilbert bundle on S', parametrized by t, in such a compensates this change in T,ff; in other words, we get the
way that the state corresponding to the Dirac sea 9: gives spin of the Skyrme soliton independently of the WZW term,
us, when we go around S', a Mobius-like picture. as a consequence of the nontrivial topology of a 2~ rotation.
The effective action for the configuration U (x,) we ob- Similar results can be found in Ref. 13.
tain by integrating the fermions [In det@ ( U )1 will change The previous analysis shows how, in a model without the
after a 2~ rotation by rrilql, where IqI, the spectral flow, WZW term, this term can be induced by fermions, since the
will be given by the number of flavors: index theorem is enough to get the correct quantized coeffi-
cient. In particular, the noninvariance of the WZW term
under parity transformations appears here in the same way
as in the model of Ref. 13, as a result of the "anomaly" of
We see that the Wess-Zumino-Witten (WZW) term just reffunder nontrivial 277 rotations.

'E Witten, Nucl. Phys. B223, 433 (1983). 8C. Callias, Commun. Math. Phys. 62, 213 (1978).
2J. Goldstone and F. Wilczek, Phys. Rev. Lett. 47, 986 (1981). 9MvI. F. Atiyah, V. K. Patodi, and I. M. Singer, Proc. Cambridge Phi-
3R. Aviv and A. Zee, Phys. Rev. D 5, 2372 (1972). 10s. Soc. 79, 71 (1976); E. Witten, Phys. Lett. 117B, 324 (1982).
4J. Goldstone and R. L. Jaffe, Phys. Rev. Lett. 51, 1518 (1983). 'OR. Bott and R. Seeley, Commun. Math. Phys. 62, 235 (1978).
5E.Witten, Nucl. Phys. B223, 422 (1983). "1. Singer, Santa Barbara notes, 1982 (unpublished).
a very mathematical language what we are using is the fact that 12M. Berry, Proc. R. Soc. London A392, 45 (1984); B. Simon, Phys.
the family index which agrees with the spectral flow is given by Rev. Lett. 51, 2167 (1983); P. Nelson and L. Alvarez Gaum6,
an element of KR ( s ' ) which
, is Z ( 2 ) . Santa Barbara report No. NSF-ITP-84-149 (unpublished).
'M. Atiyah and J. D. S. Jones, Commun. Math. Phys. 61, 97 I3A. Redlich, Phys. Rev. Lett. 52, 8 (1984); A. Niemi and
(1978). G. Semenoff, /bid. 51, 2077 (1983).

You might also like