1: \documentclass{revtex4}
2: \usepackage {amssymb}
3: \usepackage {amsmath}
4: \usepackage {epsfig}
5: \usepackage{graphicx}
6: \usepackage {hyperref}
7: \def\noi{\noindent}
8: \def\be {\begin{equation}}
9: \def\ee {\end{equation}}
10: \def\beq{\begin{eqnarray}}
11: \def\eeq{\end{eqnarray}}
12: \def\nonu{\nonumber}
13: \setlength{\parindent}{0.pt}
14:
15: \begin{document}
16:
17: \markboth{B.C. L\"{u}tf\"{u}o\={g}lu and F. Ta\c{s}k\i{i}n}
18: {Renormalization Group Analysis of a G\" ursey Model Inspired
19: Field Theory II}
20:
21:
22:
23: \title{Renormalization Group Analysis of a G\" ursey Model Inspired
24: Field Theory II}%
25:
26: \author{B.C. L\"{u}tf\"{u}o\={g}lu$^1$}%
27: \email{bcan@itu.edu.tr}%
28:
29: \author{F. Ta\c{s}k\i n$^{1,2}$ }%
30: \email{taskinf@itu.edu.tr}%
31: \affiliation{$^1$ Department of Physics, Istanbul Technical
32: University, Istanbul, Turkey \\$^2$ Department of Physics, Erciyes
33: University, Kayseri, Turkey.}
34: %\affiliation{}%
35:
36: \date{\today}%
37:
38:
39: \begin{abstract}
40: Recently a model, which is equivalent to the scalar form of G\"
41: ursey model, is shown to be a nontrivial field theoretical model
42: when it is gauged with a $SU(N)$ field. In this paper we study
43: another model that is equivalent to the vector form of the
44: G\"{u}rsey model. We get a trivial theory when it is coupled with
45: a scalar field. This result changes drastically when it is coupled
46: with an additional $SU(N)$ field. We find a nontrivial field
47: theoretical model under certain conditions.
48: \end{abstract}%
49:
50: \maketitle%
51: %\keywords {Gauged constrained models, RG equations, nontrivial
52: %models}
53: %\ccode{PACS 11.10.-z, 11.10.Hi, 12.60.Rc,}
54:
55: \section{Introduction}
56:
57: Historically, there has always been a continuing interest in
58: building nontrivial field theoretical models. A while ago it was
59: shown that perturbative expansions are not adequate in deciding
60: whether a model is nontrivial or not. Baker et al. showed that the
61: $\phi^{4}$ theory, although perturbatively nontrivial, went to a
62: free theory as the cutoff was lifted in four dimensions
63: \cite{ba_ki_79,ba_ki_81}. Continuing research is going on this
64: subject \cite{kl_06}. Alternative methods become popular.
65: Renormalization group (RG) methods are the most commonly used one.
66: They were first introduced by Wilson et al. \cite{wi_ko_74}.
67: Another method is using exact RG algorithm which were proposed by
68: Polchinski \cite{po_84}. Recent studies gave important insights on
69: both methods \cite{so_07_40_5733,so_07_40_9675,ig_it_so_07}.
70:
71: Another endeavor is building a model of nature using only
72: fermions. Here all the observed bosons are constructed as
73: composites of these ingredient spinors. In solid state physics,
74: electrons come together to form bosonic particles \cite{mi_93,
75: ba_co_sc_57}. Historically, the first work on models with only
76: spinors goes back to the work of Heisenberg \cite{he_54}. Two
77: years later G\"{u}rsey proposed his model as a substitute for the
78: Heisenberg model \cite{gu_56}. This G\"{u}rsey's spinor model is
79: important since it is conformally invariant classically and has
80: classical solutions \cite{ko_56} which may be interpreted as
81: instantons and merons \cite{ak_82}, similar to the solutions of
82: pure Yang-Mills theories in four dimensions \cite{fu_sh_90}. This
83: original model can be generalized to include vector, pseudovector
84: and pseudoscalar interactions.
85:
86: We have worked on different forms of the G\" ursey model
87: \cite{ho_lu_06,ho_ta_07,ho_lu_ta_07} using the earlier works
88: \cite{ak_ar_du_ho_ka_pa_82-34,ak_ar_du_ho_ka_pa_82-41,ak_ar_ho_pa_83,ar_ho_83,ar_ho_ka_85}
89: as a starting point. In those references it was claimed that a
90: polynomial lagrangian could be written equivalently to G\" ursey's
91: non-polynomial lagrangian. Recently it is shown that they are
92: equivalent only in a naive sense \cite{ho_lu_06,ho_ta_07}. In
93: \cite{ho_lu_06}, using perturbative methods, we showed that only
94: composite particles took part in physical processes whereas the
95: constituent fields did not interact with each other. Recently in
96: \cite {ho_lu_ta_07}, we showed that, when this model is coupled to
97: a constituent $U(1)$ gauge field, we were mimicking a gauge
98: Higgs-Yukawa (gHY) system, which had the known problems of the
99: Landau pole, with all of its connotations of triviality. There,
100: our motivation was the famous Nambu-Jona-Lasinio model
101: \cite{na_jo_61}, which was written only in terms of spinor fields.
102: This model was shown to be trivial \cite{ko_ko_94,zi_89}. Recent
103: attempts to gauge this model to obtain a nontrivial theory are
104: given in references
105: \cite{ha_ki_ku_na_94,ao_mo_su_te_to_99,ao_mo_su_te_to_00,ku_te_99,ko_ta_ya_93}.
106:
107: The essential point in our analysis is the factor of $\epsilon$ in
108: the composite propagator \cite{ho_lu_06,ho_ta_07}. This main
109: difference makes many of the diagrams convergent when the cutoff
110: is removed. Consequently, we find that we can construct a
111: nontrivial model from the scalar G\" ursey model when a
112: non-Abelian gauge field is coupled to the fermions
113: \cite{ho_lu_07}. In this paper we will investigate the vector form
114: of the G\" ursey model. Here we will closely follow the line of
115: discussion followed in the references
116: \cite{ha_ki_ku_na_94,ho_lu_07}.
117:
118: This article is organized as follows. In the next section we
119: describe the vector form of the G\" ursey like model. There we
120: derive the composite vector field propagator. In section 3, we
121: couple a constituent scalar field to our model and discuss the new
122: results. Then we solve the renormalization group equations (RGE's)
123: and find a Landau pole in the solution. In section 4, we introduce
124: another field, a non-Abelian gauge field to the model. In the
125: subsections we write the new RGE's and derive the solutions by
126: using some RG invariants. We discuss some limiting cases of the
127: coupling constant solutions before giving the criteria's of the
128: nontriviality condition in section 5. Then we find the fixed point
129: solutions. In the following subsections we analyze the solutions
130: of the coupled equations and find their asymptotic behaviors. The
131: final section is devoted to conclusions.
132:
133: \section{The Model}
134:
135: The vector form of the pure spinor G\" ursey model
136: \cite{ak_ar_ho_pa_83} is given as
137:
138: \beq %
139: L=\overline{\psi}\left(i\partial\!\!\!/ -ig\partial
140: \!\!\!/g^{-1}-m\right)\psi+\alpha \left[(\overline{\psi}
141: \gamma_{\mu}\psi)(\overline{\psi}\gamma^{\mu}\psi)\right]^{2/3}.
142: \label{gursey lagrangian}
143: \eeq %
144:
145: Here only the spinor fields have kinetic part. The $g$ field is a
146: pure gauge term to restore the local gauge symmetry, when the
147: spinor field is transformed. This non-polynomial Lagrangian has
148: been converted to an equivalent polynomial form by introducing
149: auxiliary fields $\lambda_{\mu}$ and $G_{\mu}$ in
150: \cite{ak_ar_ho_pa_83}. The constrained Lagrangian in the
151: polynomial form is given as
152:
153: \beq %
154: L_{c}=\overline{\psi}\left[i\partial \!\!\!/
155: -ig\partial \!\!\!/g^{-1}+e(G\!\!\!/
156: +\lambda\!\!\!/)-m\right]\psi
157: -e^4\lambda_{\mu}G^{\mu}G^{2}
158: +\mbox{ghost terms}. \label{s yenilagran} %
159: \eeq %
160:
161: Recently it was shown that this equivalence should be taken only
162: "naively" \cite{ho_ta_07}. This expression contains two constraint
163: equations, obtained from writing the Euler-Lagrange equations for
164: the auxiliary fields. Hence it should be quantized by using the
165: constraint analysis $\grave{a}$ \textit{la} Dirac \cite{di_64}.
166: This calculation is performed using the path integral method. We
167: find out that one can write the effective Lagrangian as
168: %
169: \beq %
170: L_{eff}=\overline{\psi}\left[i\partial \!\!\!/ -ig\partial
171: \!\!\!/ g^{-1}+e\left( G\!\!\!/+\lambda \!\!\!/
172: \right)-m\right]\psi -e^4\lambda_{\mu}G^{\mu}G^{2}+
173: \overline{w}^{\mu}(g_{\mu\nu}G^2 + 2G_{\mu}G_{\nu})w^{\nu}.
174: \label{etkinlagran} %
175: \eeq
176: %
177: Here $\overline{w}^\mu$ and $w^{\nu}$ are the ghost fields. With a
178: suitable redefinition of the fields the effective action can be
179: given as
180: %
181: \beq %
182: S_{eff}=Tr\ln\left(i\partial \!\!\!/+eJ\!\!\!/+m\right)%\nonumber \\
183: + \int dx^4 \left[e^4\left(J_{\mu}J^{\mu}J_{\lambda}J^{\lambda}\right)
184: +\mbox{other terms} \right], \label{eff_action}%
185: \eeq %
186: %
187: where $ J_{\mu}=-ig\partial_{\mu} g^{-1}+ G_{\mu}+ \lambda_{\mu}$.
188: The second derivative of the effective action with respect to the
189: $J_{\mu}$ field gives us the induced inverse propagator as
190: %
191: \beq %
192: \left . \frac{\partial^2 S_{eff}}{\partial J_{\mu}\partial
193: J_{\nu}}\right |_{J_{\mu}=0}= -\frac{g^2}{3\pi^2}\left(q_{\mu}q_{\nu}-g_{\mu\nu}q^2\right)
194: \left[\frac{1}{\epsilon}+\mbox{finite part}\right]. %
195: \eeq %
196: %
197: Here dimensional regularization is used for the momentum integral
198: and $\epsilon = 4-n$. All the other fields not shown in this
199: expression, including ghost fields arising from the constrained
200: equations, decouple from the model. The only remaining fields are
201: the spinors and the $J_{\mu}$ field. This procedure is explicitly
202: carried out in \cite{ho_ta_07}. In the Feynman gauge the
203: propagator of the composite vector field can be written as
204: $\epsilon\frac{g^{\mu\nu}}{p^2}$ where the spinor propagator is
205: the usual Dirac propagator in the lowest order.
206:
207: Although the original Lagrangian does not have a kinetic term for
208: the vector field, one loop corrections generate this term and make
209: this composite field as a dynamical entity like it is done in
210: \cite{ho_lu_06}, where the composite vector field is replaced by
211: composite scalar field. In the literature there are also other
212: similar models with differential operators in the interaction
213: Lagrangian \cite{am_ba_da_ve_81}.
214:
215:
216: In reference \cite{ho_ta_07}, the contributions to the fermion
217: propagator at higher orders were investigated by studying the
218: Dyson-Schwinger equations for the two point function. We found
219: that there is a phase which has no additions to the existing
220: fermion mass. %
221: %
222: %\pagebreak
223: %
224: \section{Coupling with A Scalar Field}
225: %
226: We may add a constituent complex scalar field to the model and
227: investigate the consequences of this addition. Our motivation is
228: the work of Bardeen et al. \cite {ba_le_lo_86, le_lo_ba_86}. When
229: they added a vector field to the Nambu-Jona-Lasinio model, a
230: complementary procedure to our work, they got interesting results.
231: Since we already have a composite vector field, we can couple a
232: massless scalar field which has its kinetic term, a self
233: interacting term with coupling constant $a$ and an interaction
234: term with new coupling constant $y$ in the Lagrangian. Then the
235: effective Lagrangian becomes
236: %
237: \beq %
238: L_{eff}=\overline{\psi}\left[i\partial \!\!\!/ -ig\partial
239: \!\!\!/ g^{-1}+e\left( G\!\!\!/+\lambda \!\!\!/
240: \right)-m\right]\psi -e^4\lambda_{\mu}G^{\mu}G^{2}+
241: \overline{w}^{\mu}(g_{\mu\nu}G^2 + 2G_{\mu}G_{\nu})w^{\nu}+
242: \frac{\partial_{\mu}\phi\partial^{\mu}\phi}{2}-
243: \frac{a}{4}\phi^{4}-y\overline{\psi}\phi\psi.
244: \label{etkinlagran+skalar1} %
245: \eeq %
246: %
247: Since the $G_{\mu}$, $\lambda_{\mu}$ and ghost fields decouple,
248: this Lagrangian reduces to the effective expression given below.
249: %
250: \beq %
251: L_{eff}=\overline{\psi}\left(i\partial \!\!\!/ +eJ\!\!\!/-
252: y\phi-m\right)\psi -e^4J^{4}+ \frac{\partial_{\mu}\phi
253: \partial^{\mu}\phi}{2}- \frac{a}{4}\phi^{4}.
254: \label{etkinlagran+skalar2} %
255: \eeq %
256: %
257: If our fermion field had a color index $i$ where $i=1...N$, we
258: could perform an 1/N expansion to justify the use of only ladder
259: diagrams for higher orders for the scattering processes. Although
260: in our model the spinor has only one color, we still consider only
261: ladder diagrams anticipating that one can construct a variation of
262: the model with N colors. In the following subsection we summarize
263: the changes in our results for this new model.
264: %
265: \subsection{New Results and Higher Orders}
266: %
267: In the model described in reference \cite{ho_ta_07}, it is shown
268: that only composites can scatter from each other with a finite
269: expression, due to the presence of $\epsilon$ in the composite
270: vector propagator. There is also a tree-diagram process where the
271: spinor scatters from a composite particle, a Compton-like
272: scattering, with a finite cross-section. This diagram can be
273: written in the other channel, which can be interpreted as spinor
274: production out of vector particles. Note that in the original
275: model the four spinor kernel was of order $\epsilon $. The lowest
276: order diagram, vanishes due to the presence of the composite
277: vector propagator. In higher orders this expression can be written
278: in the quenched ladder approximation \cite{mi_93}, where the
279: kernel is separated into a vector propagator with two spinor legs
280: joining the proper kernel. If the proper kernel is of order
281: $\epsilon$, the loop involving two spinors and a vector propagator
282: can be at most finite that makes the whole diagram in first order
283: in $\epsilon$. This fact shows that there is no nontrivial
284: spinor-spinor scattering in the original model.
285:
286: These results changes drastically with scalar field coupling. Two
287: fermion scattering is now possible due to the presence of the
288: scalar field instead of vector field channel. In lowest order this
289: process goes through the tree diagram given in Figure
290: \ref{fig789}.a. At the next higher order the box diagram with two
291: spinors and two scalar particles, Figure \ref{fig789}.b, is
292: finite from dimensional analysis. If the scalar particles are used
293: as intermediaries, the spinor production from scattering of
294: composite vector particles becomes possible as shown in Figure
295: \ref{fig789}.c where the dotted, straight and wiggly lines
296: represent scalar, spinor and composite vector particles,
297: respectively.
298:
299: %
300: \begin{figure}[h]
301: \begin{center}
302: $\begin{array}{c@{\hspace{1cm}}c@{\hspace{5mm}}c}
303: \multicolumn{1}{l}{\mbox{\bf }}&
304: \multicolumn{1}{l}{\mbox{\bf }}&
305: \multicolumn{1}{l}{\mbox{\bf }}\\
306: [-0.53cm]
307: \epsfxsize=20mm \epsffile{51-1}&
308: \epsfxsize=18mm \epsffile{52-0}&
309: \epsfxsize=22mm \epsffile{41-1} \\
310: [0.4cm]
311: \mbox{\bf (a)} &
312: \mbox{\bf (b)} &
313: \mbox{\bf (c)}
314: \end{array}$
315: \end{center}
316: \caption{(a) Two spinor scattering through the scalar particle
317: channel, (b) Higher order diagram for two spinor scattering,
318: \hspace{8mm}(c) Spinor production from scattering of composite
319: vectors.} \label{fig789}
320: \end{figure}
321: %
322: \subsection{Renormalization Group Equations and Solutions}
323: %
324: In reference \cite{ho_ta_07}, it is widely discussed that the
325: $<{\overline{\psi}}\psi J_\mu>$ vertex and the spinor box diagram
326: give finite results. The higher diagrams do not change this
327: result, since each momentum integration is accompanied by an
328: $\epsilon$ term in the composite vector propagator. Therefore,
329: there is no need for infinite coupling constant renormalization.
330:
331:
332: In the new model where a massless scalar field is added, all the
333: three coupling constants are renormalized. One can write the first
334: order RGE's for these coupling constants, similar to the analysis
335: in \cite{ha_ki_ku_na_94}. We take $\mu_0$ as a reference scale at
336: low energies, $t=ln (\mu/\mu_0)$, where $\mu$ is the
337: renormalization point.
338: %
339: \beq
340: 16\pi^2\frac{d}{dt}y(t) &=& A y ^3(t),\label{y3} \\
341: 16\pi^2\frac{d}{dt}e(t) &=& B e(t) y^2(t),\label{eg2}\\
342: 16\pi^2\frac{d}{dt}a(t) &=& C a^2(t)-D y^4(t) \label{g4}.
343: \eeq
344: %
345: Here $A$, $B$, $C$ and $D$ are positive numerical constants. We
346: find out that Yukawa and $<{\overline{\psi}}\psi J_\mu>$ vertices
347: have only scalar correction. The composite vector correction to
348: these vertices are finite due to the $\epsilon$ in the propagator.
349: Therefore, our equations differ from those in reference
350: \cite{ho_lu_07,ha_ki_ku_na_94}. These processes are illustrated in
351: diagrams shown in Figure \ref{fig456}.
352: %
353: \begin{figure}[h]
354: \begin{center}
355: $\begin{array}{c@{\hspace{1cm}}c@{\hspace{5mm}}c@{\hspace{5mm}}c}
356: \multicolumn{1}{l}{\mbox{\bf }}&
357: \multicolumn{1}{l}{\mbox{\bf }}&
358: \multicolumn{1}{l}{\mbox{\bf }}&
359: \multicolumn{1}{l}{\mbox{\bf }}\\
360: [-0.53cm]
361: \epsfxsize=13mm \epsffile{01-1} &
362: \epsfxsize=13mm \epsffile{01-2} &
363: \epsfxsize=20mm \epsfysize=8mm \epsffile{01-3} &
364: \epsfxsize=16mm \epsffile{01-4} \\
365: [3mm]
366: \mbox{\bf (a)} &
367: \mbox{\bf (b)} &
368: \mbox{\bf (c)} &
369: \mbox{\bf (d)}
370: \end{array}$
371: \end{center}
372: \caption{The three coupling constant corrections in one loop. }
373: \label{fig456}
374: \end{figure}
375:
376: The RGE's have the immediate solutions
377: %
378: \beq
379: y^2(t)&=&\frac{y_0^2}{Z(t)}, \\
380: e(t)&=&e_0Z(t)^{-B/2A}, \\
381: a(t)&=& \frac{A\pm\sqrt{A^2+CD}}{C}\frac{y_0^2}{Z(t)},
382: \eeq
383: %
384: where $Z(t)=1-\frac{Ay_0^2}{8\pi^2}t$.
385:
386: The main problem of models with U(1) coupling, namely the Landau
387: pole, is expected to make our new model a trivial one. We expect
388: that coupling to a non-Abelian gauge theory will remedy this
389: defect by new contributions to the RGE's. Thus, obtaining a
390: nontrivial model will be possible. Coupling to a non-Abelian gauge
391: field will also give us more degrees of freedom in studying the
392: behavior of the beta function. This may allow us to find the
393: critical number of gauge and fermion fields to obtain a zero of
394: this function at nontrivial values of the coupling constants of
395: the model.
396: %
397: \section{Coupling with a Non-Abelian Field}
398: %
399: In this section we consider our model with $SU(N_{C})$ gauge field
400: interaction, where the spinors have $N_{f}$ different flavors.
401: Although we study in the leading order of $\frac{1}{N_{C}}$
402: expansion, where all the planar diagrams contribute to the RGE's,
403: we are interested in the high-energy asymptotic region where the
404: gauge coupling is perturbatively small;
405: $\frac{g^{2}N_{C}}{4\pi}\ll 1$. However, the number of fermions is
406: in the same order as $N_{C}$. Only $n_{f}$ fermions have a
407: degenerate large Yukawa coupling. We start with the effective
408: Lagrangian
409: %
410: \beq %
411: L_{eff}=\sum_{i=1}^{N_{f}}\overline{\psi}_{i}\left(iD \!\!\!\!/ +eJ\!\!\!/
412: -m\right)\psi_{i}-e^4J^{4}+ \frac{\partial_{\mu}\phi
413: \partial^{\mu}\phi}{2}- \frac{a}{4}\phi^{4}-\sum_{i=1}^{n_{f}}
414: \overline{\psi}_{i}y\phi\psi_{i}-\frac{1}{4} Tr [F_{\mu \nu} F^{\mu \nu}]
415: + L_{\mbox{ghost}}+ L_{\mbox{g.f.}}.
416: \label{etkinlagran+skalar+gauge} %
417: \eeq %
418: %
419: The gauge field belongs to the adjoint representation of the color
420: group $SU(N_C) $ where $D_{\mu}$ is the color covariant
421: derivative. $y$, $a$, $e$ and $g$ are the Yukawa, quartic scalar,
422: composite vector and gauge coupling constants, respectively.
423:
424: There are two kind of ghost fields in the model. The first one,
425: which comes from the composite constraints, decouples from our
426: model \cite{ak_ar_ho_pa_83,ak_ar_du_ho_ka_pa_82-41}. The second
427: one, coming from the gauge condition on the vector field, do not
428: decouple and contribute to the RGE's in the usual way.
429: %
430: \subsection{Renormalization Group Equations and Solutions}
431: %
432: In this subsection we will analysis the RGE's in the leading order
433: of the approximation given above. In the one loop approximation
434: the RGE's are
435: %
436: \beq
437: 16\pi^2\frac{d}{dt}g(t) &=& -A g^3(t),\label{5g3} \\
438: 16\pi^2\frac{d}{dt}y(t) &=& B y^3(t)-Cy(t)g^2(t),\label{5y3}\\
439: 16\pi^2\frac{d}{dt}e(t) &=& D e(t)y^2(t)-Ee(t)g^2(t),\label{5eg2}\\
440: 16\pi^2\frac{d}{dt}a(t) &=& F a(t)y^2(t)-Gy^4(t)
441: \label{5L4}.
442: \eeq
443: %
444: Here $A$, $B$, $C$, $D$, $E$, $F$ and $G$ are positive constants.
445:
446: In the RGE's we see that the diagrams, where the composite vector
447: field takes part, are down by order of $\epsilon$. Therefore we do
448: not have contributions proportional to $e^2(t)$, $e^3(t)$,
449: $y(t)e^2(t)$ and $g(t)e^2(t)$. Also we neglect the scalar loop
450: contribution to the gauge coupling $g(t)$, similar to the work of
451: \cite{ha_ki_ku_na_94}.
452:
453: The solutions of the first RG equation (\ref{5g3}) can be given as %
454: %
455: \beq%
456: g^{2}(t)=g_{0}^{2}\Bigg(1+\frac{A\alpha_0}{2\pi}t\Bigg)^{-1},
457: \label{g_nin_cozumu}%
458: \eeq%
459: %
460: where $\alpha_0=\frac{g_{0}^2}{4\pi}$. We define %
461: %
462: \beq
463: \eta(t)\equiv\frac{\alpha(t)}{\alpha_{0}}\equiv\frac{g^2(t)}{g_0^2},
464: \eeq %
465: %
466: where $g_0=g(t=0)$ which is the initial value at the reference
467: scale $\mu_0$. For the solution of the second RG equation
468: (\ref{5y3}), we can propose a RG invariant $H(t)$ as
469: %
470: \beq
471: H(t)=-\eta^{-1+C/A}(t)\left[1-\frac{C-A}{B}\frac{g^2(t)}{y^2(t)}\right].
472: \eeq %
473: Since $H(t)$ is a constant, we call it $H_0$. Then, the solution
474: of the Yukawa coupling constant can be written as
475: %
476: \beq
477: y^2(t)=\frac{C-A}{B}g^2(t)\left[1+H_{0}\eta^{1-\frac{C}{A}}(t)\right]^{-1}.
478: \label{y_nin_cozumu}
479: \eeq %
480: The solution of the third RG equation
481: (\ref{5eg2}) can be defined by another RG invariant $P(t)$ if and
482: only if the constants $B$ equals to $D$ and $C$ equals to $E$.
483: Then the invariant becomes
484: %
485: \beq
486: P(t)=-\eta^{-1+C/A}(t)\left[1-\frac{B}{C-A}\frac{y^2(t)}{g^2(t)}\right]
487: \frac{e^2(t)}{y^2(t)}\frac{g^2(t)}{y^2(t)}.
488: \eeq %
489: The solution of the composite vector coupling $e(t)$ can be
490: written as
491: %
492: \beq
493: e^2(t)=-\frac{P_{0}}{H_{0}}\left(\frac{C-A}{B}\right)^{2}g^2(t)\left[1+H_{0}
494: \eta^{1-\frac{C}{A}}(t)\right]^{-1}. \label{e_nin_cozumu}
495: \eeq %
496: %
497: where $P_{0}$ denotes the value of the invariant $P(t)$. The
498: solution of the last RG equation (\ref{5L4}) can be defined by
499: another RG invariant $K(t)$, given as
500: %
501: \beq %
502: K(t)=-\eta^{-1+\frac{2C}{A}}(t)\left[
503: 1-\frac{2C-A}{2B}\frac{a(t)}{y^2(t)}\frac{g^2(t)}{y^2(t)}\right].
504: \eeq %
505: We can rewrite the solution with a value of the invariant $K(t)$
506: as $K_{0}$
507: %
508: \beq %
509: a(t)=\frac{2(C-A)^{2}}{(2C-A)B}g^{2}(t)
510: \frac{1+K_{0}\eta^{1-\frac{2C}{A}}(t)}{\left[1+H_{0}\eta^{1-\frac{C}{A}}(t)\right]^{2}}
511: \label{a_nin_cozumu}.%
512: \eeq
513: %
514: Here we notice that the RG constants $H_{0}$, $P_{0}$ and $K_{0}$
515: play important roles on the behavior of the solutions of the
516: coupling equations (\ref{g_nin_cozumu}), (\ref{y_nin_cozumu}),
517: (\ref{e_nin_cozumu}), (\ref{a_nin_cozumu}). Similar works have
518: been studied in \cite{ha_ki_ku_na_94,ho_lu_07}. The values of the
519: constants are given in these equations as %
520: \beq %
521: A=\frac{11N_C-4T(R)N_f}{3}, \hspace{5mm} B=D=\frac{G}{4}=2n_fN_C,
522: \hspace{5mm} C=E=6C_2(R), \hspace{5mm} F=G. %
523: \eeq %
524: Here $C_2(R)$ is a second Casimir,
525: $C_2(R)=\frac{(N_{C}^{2}-1)}{2N_{C}}$, $R$ is the fundamental
526: representation with $T(R)=\frac{1}{2}$.
527:
528: Before entering the analysis of the fixed point, we briefly
529: investigate the results of some limits.
530: %
531: \subsubsection{The limiting case A$\rightarrow$$+0$ for finite $t$}
532: In this case the coupling constants solutions can be written as
533: %
534: \beq g^2(t)&=&g^2_0, \\
535: y^2(t)&=&\frac{8\pi^{2}}{B}\frac{\alpha}{\alpha_{c}}
536: \left[1+H_{0}exp\left(\frac{\alpha}{\alpha_{c}}t\right)\right]^{-1},\\
537: e^2(t)&=&-\frac{16\pi^{3}}{B^{2}}\frac{P_{0}}{H_{0}}\frac{\alpha}{\alpha_{c}^{2}}
538: \left[1+H_{0}exp\left(\frac{\alpha}{\alpha_{c}}t\right)\right]^{-1},\\
539: a(t)&=&\frac{8\pi^2}{B}\frac{\alpha}{\alpha_{c}}
540: \frac{\left[1+K_{0}exp\left(\frac{2\alpha}{\alpha_{c}}t\right)\right]}
541: {\left[1+H_{0}exp\left(\frac{\alpha}{\alpha_{c}}t\right)\right]^{2}}.
542: \eeq %
543: Here $\alpha_{0}=\alpha$ and $\frac{C}{2\pi}=\frac{1}{\alpha_c}$.
544: %
545: \subsubsection{The limiting case A$\rightarrow$$C$ for finite $t$}
546: In this limit case the solutions of the couplings
547: (\ref{y_nin_cozumu}), (\ref{e_nin_cozumu}) and
548: (\ref{a_nin_cozumu}) seem to vanish. If we suggest new RG
549: invariant $H_{1}$, instead of $H_{0}$, as
550: $H_{0}=-1+\frac{C-A}{A}H_{1}$ , we find that two of the coupling
551: solutions do not vanish, whereas composite vector coupling goes to
552: zero.
553: These behaviors are given below %
554: %
555: \beq
556: y^{2}(t)&=&\frac{A}{B}g^{2}(t)\left[H_{1}+\ln\eta(t)\right]^{-1}, \\
557: e^{2}(t)&=&P_{0}\left(\frac{C-A}{B}\right)\left(
558: \frac{A}{B}\right)g^{2}(t)\left[H_{1}+\ln\eta(t)\right]^{-1}, \\
559: a(t)&=& \frac{2A}{B}g^{2}(t) \frac{1+K_{0}\eta^{-1}(t)}{
560: \left[H_{1}+\ln\eta(t)\right]^{2}}.
561: \eeq
562: %
563: It is amusing to see that the added interactions nullify the
564: original vector-spinor coupling.
565: \subsubsection{The limiting case A$\rightarrow$$2C$ for finite $t$}
566: %
567: In this limit case only the quartic coupling constant solution
568: (\ref{a_nin_cozumu}) behaves critically. Similarly we can redefine
569: RG invariant $K_{1}$ instead of $K_{0}$ as
570: $K_{0}=-1+\frac{2C-A}{A}K_{1}$, then the quartic coupling solution
571: takes the form
572: %
573: \beq %
574: a(t)=\frac{C}{B}g^{2}(t)\frac{K_{1}+\ln\eta(t)}{
575: \left[1+H_{0}\eta^{1/2}(t)\right]^{2}}.
576: \eeq %
577: %
578: This limit is not allowed because it does not give asymptotic
579: freedom.
580:
581: In the next section we will mention which criteria are needed to
582: define a nontrivial theory.
583: %
584: \section{Nontriviality of the system} \label{nontriviality}
585: %
586: To have a nontrivial theory all the running coupling constants
587: should not diverge at any finite energy, which means the absence
588: of Landau poles of the system. For a consistent theory these
589: solutions should not vanish identically and must have real and
590: positive values. These conditions make the model unitary and
591: satisfy the vacuum stability criterion. Note that if we decouple
592: the scalar and composite vector field from the system, we have a
593: nontrivial theory, similar to QCD. Therefore, $e(t)\equiv
594: g(t)\equiv a(t)\equiv 0$ solution will not be named as the
595: nontriviality of our composite model. The mass parameter can be
596: renormalized in the $\overline{MS}$ scheme and the mass can be
597: chosen as zero.
598:
599: Remember that we are restricted by neglecting the scalar loop
600: contributions to the gauge coupling where the composite vector
601: contributions are not neglected but down due to the presence of
602: $\epsilon$ in its propagator. If the Yukawa and/or quartic scalar
603: couplings become so large and break the $1/N_{C}$ expansion then
604: the behavior of the gauge coupling might be affected.
605:
606: These restriction conditions are the same as the ones in the gHY
607: system which was discussed widely in \cite{ha_ki_ku_na_94}. A
608: while ago, one of us, B.C.L., with a collaborator, studied the
609: scalar form of the G\"ursey model in this fashion \cite{ho_lu_07}.
610: In that model, there is a composite scalar field with a propagator
611: completely different from a constituent scalar field used in
612: reference \cite{ha_ki_ku_na_94}. There, we showed that a
613: restriction is not needed between the scalar and the gauge field
614: coupling since the contribution of the scalar field to the gauge
615: field is down by the factor of $\epsilon$ in the scalar
616: propagator. In this work, the vector form of the G\"ursey Model,
617: we have constituent scalar field and composite vector field which
618: is missing in gHY system. This composite field adds a new RGE to
619: the system but does not contribute to the former ones in gHY
620: system with a totally different reason.
621:
622: After these remarks we will discuss the nontriviality conditions
623: of our model in the following subsections.
624: %
625: \subsection{Fixed Point Solution}
626: The RGE's can be rewritten as %
627: \beq %
628: 8\pi^{2}\frac{d}{dt}\left[\frac{y^{2}(t)}{g^{2}(t)}\right]&=&
629: Bg^{2}(t)\left[\frac{y^{2}(t)}{g^{2}(t)}\right]
630: \left[\frac{y^{2}(t)}{g^{2}(t)}-\frac{C-A}{B}\right], \\
631: 8\pi^{2}\frac{d}{dt}\left[\frac{e^{2}(t)}{y^{2}(t)}\frac{g^{2}(t)}
632: {y^{2}(t)}\right]&=&(C-A)g^{2}(t)\left[\frac{e^{2}(t)}{y^{2}(t)}\right]
633: \left[\frac{g^{2}(t)}{y^{2}(t)}-\frac{B}{C-A}\right], \\
634: 8\pi^{2}\frac{d}{dt}\left[\frac{a(t)}{y^{2}(t)}\frac{g^{2}(t)}
635: {y^{2}(t)}\right]&=&(2C-A)g^{2}(t)
636: \left[\frac{a(t)}{y^{2}(t)}\frac{g^{2}(t)}{y^{2}(t)}
637: -\frac{2B}{2C-A}\right].
638: \eeq
639: %
640: The fixed point solutions can be given as %
641: %
642: \beq %
643: \frac{y^{2}(t)}{g^{2}(t)}&=&\frac{C-A}{B}, \\
644: \frac{e^{2}(t)}{y^{2}(t)}&=&\mbox{Arbitrary constant}, \\
645: \frac{a(t)}{y^{2}(t)}&=&\frac{2(C-A)^{2}}{B(2C-A)}.
646: \eeq%
647: %
648: These are also the solutions of the equations
649: (\ref{y_nin_cozumu}), (\ref{e_nin_cozumu}) and
650: (\ref{a_nin_cozumu}) where the RG invariants are
651: $P_{0}=H_{0}=K_{0}=0$ as $P_{0}=\zeta H_{0}$. Here $\zeta$ is a
652: constant.
653: %
654: It is clear that the behavior of all the coupling constants are
655: determined by the gauge coupling which means the Kubo, Sibold and
656: Zimmermann's "coupling constant reduction" \cite{ku_si_zi_89}.
657: This corresponds to the Pendleton-Ross fixed point
658: \cite{pe_ro_81} in the context of the RGE. Remark that only the
659: case, $C>A$, prevents the violation of the unitarity and keeps the
660: stability of the vacuum. This gives rise to nontriviality of the
661: model when the RG invariants are set to zero. In the following
662: subsections we will analysis the coupling constant solutions only
663: in this case with non zero RG invariants.
664: %
665: \subsection{Yukawa Coupling}
666: %
667: The Yukawa coupling solution is given in equation
668: (\ref{y_nin_cozumu}). It is obvious that the sign of the RG
669: invariant, $H_{0}$, plays an important role in the behavior of the
670: solution where B is positive. The ultraviolet (UV) limit of
671: $\eta(t)$ is needed before continuing the analysis in $C>A$ case.
672: %
673: \beq %
674: \eta^{1-\frac{C}{A}}(t \rightarrow\infty)\rightarrow \ %
675: \begin{array}{ll}
676: +\infty. \\
677: \end{array}%
678: \eeq %
679: The UV behavior of Yukawa coupling with a non zero RG invariant
680: $H_{0}$ is
681: %
682: \beq %
683: y^2(t\rightarrow\infty)\rightarrow\ \left\{%
684: \begin{array}{lrl}
685: +0, \hspace{5mm} &\hbox{$0<$}&\hbox{$H_{0}<\infty$;} \\
686: \hbox{Landau Pole}, &\hspace{5mm} \hbox{$-1<$}&\hbox{$H_{0}<0$;} \\
687: -0, \hspace{5mm} &\hbox{$-\infty<$}&\hbox{$H_{0}\leq-1$.} \\
688: \end{array}%
689: \right. %
690: \eeq %
691: %
692: For $-1<H_{0}<0$ case, there exists a finite value of $t$ before it goes to infinity %
693: \beq %
694: 1+\frac{A\alpha_{0}}{2\pi}t=\left(\frac{-1}{H_{0}}\right)^{A/(C-A)}.
695: \eeq %
696: %
697: In this $t$ value Yukawa coupling diverges and changes its sign.
698: These asymptotic behaviors show that the theory is nontrivial if
699: and only if the RG invariant $H_{0}$ is positive.
700:
701: The RG flows in the $(g^{2}(t),y^{2}(t))$ plane are shown in
702: Figure \ref{y_kare_g_kare}. The upper bound of the figure denotes
703: the "Landau Pole".
704: %
705: \begin{figure}[htb!]
706: \epsfxsize=85mm %
707: \epsffile{y_kare_g_kare.eps} %
708: \caption{Plot of $g^2(t)$ vs. $y^2(t)$ for different values of
709: $H_0$. The arrows denote the flow directions toward the UV
710: region.} \label{y_kare_g_kare}%
711: \end{figure}
712: %
713: %\pagebreak
714: %
715: \subsection{Composite Vector Field Coupling}
716: %
717: The composite vector coupling solution is given in equation
718: (\ref{e_nin_cozumu}). In this case not only the sign of $H_{0}$
719: but also the sign of $P_{0}$ is crucial for nontriviality. Since
720: $H_{0}$ is positive, $P_{0}$ must be negative. The composite
721: vector field coupling behaves similarly to the Yukawa coupling up
722: to a constant multiplier. In figure \ref{e_kare_y_kare} we plot
723: $e^2(t)$ vs. $y^2(t)$ where $P_{0}<0$, $H_{0}>0$. Both coupling
724: constants approach the origin as $t$ goes to infinity. Thus, our
725: model fulfills the condition required by the asymptotic freedom
726: criterion.
727: %
728: \begin{figure}[htb!]
729: \epsfxsize=85mm %
730: \epsffile{e_kare_y_kare.eps} %
731: \caption{Plot of $y^2(t)$ vs. $e^2(t)$ for the values of
732: $H_{0}>0$ and $P_{0}<0$.}
733: \label{e_kare_y_kare}%
734: \end{figure}
735: %
736: %\pagebreak
737: %
738: \subsection{Quartic Scalar Field Coupling}
739: %
740: Finally quartic scalar coupling solution given in equation
741: (\ref{a_nin_cozumu}) can be analyzed. We have already restricted
742: ourselves with $C>A$, $H_{0}>0$ and $P_{0}<0$ for nontriviality.
743: In the limit where $t\gg 1$ , the $\eta$ terms in the last
744: fraction of equation (\ref{a_nin_cozumu}) become dominant
745: therefore $1$ can be neglected. Hence we can express the solution
746: as
747: %
748: \beq
749: a(t)\approx \frac{2(C-A)^{2}}{(2C-A)B}g_{0}^{2}\eta(t) \frac{K_{0}\eta^{1-2C/A}(t)}{
750: \left[H_{0}\eta^{1-C/A}(t)\right]^{2}},
751: \eeq %
752: %
753: which is equal to
754: %
755: \beq
756: a(t\rightarrow\infty)=\frac{2(C-A)^{2}}{(2C-A)B}g_{0}^{2} \frac{K_{0}}{H_{0}^{2}}.
757: \eeq %
758: %
759: This asymptotic behavior shows that to have a nontrivial model the
760: RG invariant $K_{0}$ should be equal to zero. The other
761: possibilities for a non zero solution for $K_{0}$ is been widely
762: discussed in the reference \cite{ha_ki_ku_na_94}. In Figure
763: \ref{a_y_kare}, we plot the RG flows in $(a(t),y^2(t))$ plane for
764: different values of $H_0$ higher than zero while the gauge
765: coupling $\alpha(t=0)$ is fixed to one. The origin is the limit
766: where $t$ goes to infinity, there both
767: coupling constants approach zero when $K_{0}=0$. %
768: %
769: \begin{figure}[htb!]
770: \epsfxsize=85mm %
771: \epsffile{a_y_kare.eps} %
772: \caption{Plot of $a(t)$ vs. $y^2(t)$ for the values of
773: $H_{0}>0$ and $K_{0}=0$.}
774: \label{a_y_kare}%
775: \end{figure}
776: %
777: %\pagebreak
778: %
779: \section{Conclusion}
780: %
781: A while ago, one of us, F.T., with a collaborator, showed that the
782: scattering of composite vector particles gives nontrivial results
783: while the constituent spinors do not. In that work
784: \cite{ho_ta_07}, a polynomial Lagrangian model inspired by the
785: vector form of G\" ursey model was used. Here we couple a
786: constituent massless scalar field to our previous model. We find
787: out that many of the features, related to the creating and
788: scattering of the spinor particles of the original model, are not
789: true anymore. In the one loop approximation we find the RGE's
790: whose solutions have all the problems associated with the Landau
791: pole, like the case in reference \cite{ho_lu_ta_07}. To remedy
792: this defect we couple a $SU(N_{C})$ non-Abelian gauge field to the
793: new model. We solve the new RGE's and conclude that if the
794: conditions $C>A$, $H_{0}>0$, $P_{0}\leq 0$ and $K_{0}=0$ are
795: satisfied, the model gives a result which can be interpreted as a
796: nontrivial field theoretical model. We find fixed point solutions
797: where the coupling constants are not equal to zero. In section
798: \ref{nontriviality} we plot the UV region behavior of the coupling
799: constants. There, they all go to zero asymptotically which means
800: asymptotic freedom, which is another feature of a nontrivial
801: model.
802:
803: Our calculation shows that one can construct nontrivial field
804: theory starting from constrained Lagrangians.
805:
806: \vspace{5mm}\textbf{Acknowledgement}: We thank to Mahmut Horta\c
807: csu for discussions and both scientific and technical assistance
808: while preparing this manuscript. We also thank Nazmi Postac\i
809: o\={g}lu for technical discussions. This work is supported by the
810: ITU BAP project no: 31595. This work is also supported by TUBITAK,
811: the Scientific and Technological Council of Turkey.
812:
813:
814: \begin{thebibliography}{99}
815:
816: \bibitem{ba_ki_79} G.A. Baker and J.M. Kincaid, {\it Phys. Rev. Lett.} {\bf 42,} 1431 (1979).
817:
818: \bibitem{ba_ki_81} G.A. Baker and J.M. Kincaid, {\it J. Stat. Phys.} {\bf 24,} 469 (1981).
819:
820: \bibitem {kl_06} J. R. Klauder, {\it Ann. Phys. (N.Y.)} {\bf 322,} 2569 (2007). hep-th/0609091.
821:
822: \bibitem{wi_ko_74} K.G. Wilson and J.B. Kogut, {\it Phys. Rep.} {\bf12,} 75 (1974).
823:
824: \bibitem{po_84} J. Polchinski, {\it Nucl. Phys. B} {\bf 231,} 269 (1984).
825:
826: \bibitem{so_07_40_5733} H. Sonoda, {\it J. Phys. A:Math. Theor.} {\bf 40,} 5733 (2007), hep-th/0612294.
827:
828: \bibitem{so_07_40_9675} H. Sonoda, {\it J. Phys. A:Math. Theor.} {\bf 40,} 9675 (2007), hep-th/0703167.
829:
830: \bibitem{ig_it_so_07} Y. Igarashi, K. Itoh and H. Sonoda, {\it Prog. Theor. Phys.} {\bf 118,} 121 (2007), hep-th/0704.2349.
831:
832: \bibitem{ba_co_sc_57} J. Bardeen, L.N. Cooper and J.R. Schrieffer, {\it Phys. Rev.} {\bf 108}, 1175 (1957).
833:
834: \bibitem{mi_93} V.A. Miransky, {\it{ Dynamical Symmetry Breaking in Quantum Field Theories}} (World Scientific, Singapore,
835: 1993).
836:
837: \bibitem{he_54} W. Heisenberg. {\it Z. Naturforschung A} {\bf 9,}
838: 292 (1954).
839:
840: \bibitem{gu_56} F. G\" ursey, {\it Nuovo Cimento} {\bf 3,} 988 (1956).
841:
842: \bibitem{ko_56} F. Kortel, {\it Nuovo Cimento} {\bf 4,} 210 (1956).
843:
844: \bibitem{ak_82} K.G. Akdeniz, {\it Lett. Nuovo Cimento} {\bf 33,} 40 (1982).
845:
846: \bibitem{fu_sh_90} W.I. Fushchich and W.M. Shtelen, {\it
847: J. Phys. A: Math. Gen.} {\bf 23,} L517 (1990).
848:
849: \bibitem{ho_lu_06} M. Horta\c{c}su and. B.C. L\"{u}tf\"{u}o\u{g}lu,
850: {\it Mod. Phys. Lett. A} {\bf 21,} 653 (2006), hep-th/0506024.
851:
852: \bibitem{ho_ta_07} M. Horta\c{c}su and F. Ta\c{s}k\i n, {\it Int. J. Mod. Phys. A} {\bf 22,}
853: 83 (2007), hep-th/0605217.
854:
855: \bibitem{ho_lu_ta_07} M. Horta\c csu, B.C. L\"utf\"uo\u{g}lu and F. Ta\c sk\i n, {\it Mod. Phys. Lett. A} {\bf 22,} 2521 (2007), hep-th/0611116.
856:
857: \bibitem{ak_ar_du_ho_ka_pa_82-34} K.G. Akdeniz, M. Ar\i k, M. Durgut, M. Horta\c{c}su, S. Kaptano\={g}lu, N.K. Pak, {\it Phys. Lett. B} {\bf 116,} 34 (1982).
858:
859: \bibitem{ak_ar_du_ho_ka_pa_82-41} K.G. Akdeniz, M. Ar\i k, M. Durgut, M. Horta\c{c}su, S. Kaptano\={g}lu, N.K. Pak, {\it Phys. Lett. B} {\bf 116,} 41 (1982).
860:
861: \bibitem{ak_ar_ho_pa_83} K.G. Akdeniz, M. Ar\i k, M. Horta\c csu, N.K. Pak, {\it Phys. Lett. B} {\bf 124,} 79 (1983).
862:
863: \bibitem{ar_ho_83} M. Ar\i k and M. Horta\c csu, {\it J. of Phys. G} {\bf 9,} L119 (1983).
864:
865: \bibitem{ar_ho_ka_85} M. Ar\i k, M. Horta\c csu and J. Kalayc\i, {\it J. of
866: Phys. G} {\bf 11}, 1 (1985).
867:
868: \bibitem{na_jo_61} Y. Nambu and G. Jona-Lasinio, {\it Phys. Rev.} {\bf 122,} 345 (1961).
869:
870: \bibitem{ko_ko_94} A. Kocic and J.B. Kogut, {\it Nucl. Phys. B} {\bf 422,} 593 (1994).
871:
872: \bibitem {zi_89} J. Zinn-Justin, {\it Quantum Field Theory and Critical Phenomena} (Clarendon, Oxford, 1989).
873:
874: \bibitem{ha_ki_ku_na_94} M. Harada, Y. Kikukawa, T. Kugo and H. Nakano, {\it Prog. Theor. Phys.} {\bf 92,} 1161 (1994).
875:
876: \bibitem{ao_mo_su_te_to_99} K. Aoki, K. Morikawa, Jun Sumi, H. Terao and M. Tomoyose, {\it Prog. Theor. Phys.} {\bf 102,} 1151 (1999), hep-th/9908042.
877:
878: \bibitem{ao_mo_su_te_to_00} K.I. Aoki, K. Morikawa, J.I. Sumi, H. Terao and M. Tomoyose, {\it Phys. Rev. D} {\bf 61,} 045008 (2000), hep-th/9908043.
879:
880: \bibitem{ku_te_99} K. Kubota and H. Terao , {\it Prog. Theor. Phys.} {\bf 102,} 1163 (1999), hep-th/9908062.
881:
882: \bibitem{ko_ta_ya_93} K. Kondo, M. Tanabashi and K. Yamasaki, {\it Prog. Theor. Phys.} {\bf 89,} 1249 (1993), hep-ph/9212208.
883:
884: \bibitem{ho_lu_07} M. Horta\c{c}su and. B.C. L\"{u}tf\"{u}o\u{g}lu,
885: {\it Phy. Rev. D} {\bf 76,} 025013 (2007), hep-th/0703103.
886:
887: \bibitem{di_64} P.A.M. Dirac, {\it Lectures on Quantum Mechanics},
888: (Belfer Graduate School of Science, Yeshiva University, New York,
889: 1964).
890:
891: \bibitem{am_ba_da_ve_81} D. Amati, R. Barbieri, A.C. Davis and G. Veneziano, {\it Phys. Lett. B} {\bf 102,} 408
892: (1981).
893:
894: \bibitem{ba_le_lo_86} W.A. Bardeen, C.N. Leung and S.T. Love, {\it Phys. Rev. Lett.} {\bf 56,}
895: 1230 (1986).
896:
897: \bibitem{le_lo_ba_86} C.N. Leung, S.T. Love and W.A. Bardeen, {\it
898: Nucl. Phys. B} {\bf273,} 649 (1986).
899:
900: \bibitem{ku_si_zi_89} J. Kubo, K. Sibold and W. Zimmermann, {\it Phys. Lett. B} {\bf 220,} 191 (1989).
901:
902: \bibitem{pe_ro_81} B. Pendleton and G.G. Ross, {\it Phys. Lett. B} {\bf98,} 291 (1981).
903:
904: \end{thebibliography}
905: \end{document}
906: