1: %---------------------------------------------------------------------------
2: %\documentstyle{article}
3: %\documentstyle[osa,onecolumn]{revtex}
4:
5:
6: \documentstyle[prl,aps]{revtex}
7: %%%%%%%%%%%%%%%%%%%%%%%%%%%%
8: \renewcommand{\thesection}{\Roman{section}}
9:
10: \begin{document}
11: \title{Pion Form Factor and Ambiguities in a Renormalizable Version of the
12: Nambu-Jona-Lasinio Model\thanks{%
13: Supported by CNPq-Brazil}}
14: \author{B. M. Rodrigues, A. L. Mota\thanks{%
15: motaal@ufsj.edu.br},}
16: \address{{\normalsize Departamento de Ci\^{e}ncias Naturais, Universidade Federal de
17: S\~{a}o Jo\~{a}o del Rei}\\
18: {\normalsize Caixa Postal 110, CEP 36.300-000, S\~ao Jo\~ao del Rei, MG,
19: Brazil}}
20: \maketitle
21:
22: \begin{abstract}
23: We analyze the presence of an ambiguity in the pion electromagnetic form
24: factor within a renormalizable version of the Nambu-Jona-Lasinio model. We
25: found out that the ambiguity present on the evaluation of the form factor
26: decouples from its transversal part, confirming previous results obtained
27: ignoring the ambiguity. This result helps us to understand the role played
28: by finite but undetermined quantities in Quantum Field Theories.
29: \end{abstract}
30:
31: %
32: %==========================================================================
33: % Introduction
34: %==========================================================================
35: %
36:
37: \section{Introduction}
38:
39: Recently (\cite{Colladay},\cite{Kostelecky},\cite{Jackiw}) it was brougth to
40: attention the existence of finite but undetermined radiative corrections in
41: Quantum Field Theory. These finite amplitudes are related to differences
42: between divergent quantities of the same degree of divergence, and the
43: result obtained by employing a particular regularization scheme can be
44: different of the result obtained by another one (in particular, this
45: difference appears between gauge invariant and non invariant schemes)\cite
46: {Perez}. This situation becomes more complex in models with parity violating
47: quantities, where dimensional regularization is inappropriate. A classical
48: example is the ABJ anomaly\cite{ABJ}, where one must choose between the
49: transversality of the vector or axial-vector currents. At this point, the
50: ambiguity plays a crucial role in determining which one of the symmetry
51: relations will be violated.
52:
53: Scarpelli et al. \cite{Scarpelli} studied this situation employing a
54: regularization independent procedure that have been called Implicit
55: Regularization (IR) (\cite{Orimar},\cite{Brizola}). By using IR, they had
56: shown that the same ambiguities can occur in several situations, as in QED
57: vacuum polarization tensor. In this case, gauge invariance has to be used to
58: fix the ambiguity value. As proposed in \cite{Jackiw}, the ambiguity was
59: fixed in order to preserve some physical relationship, e.g., symmetry. One
60: of the advantages of the IR procedure is that the amplitudes can be
61: evaluated almost all the time without specifying some particular
62: regularization scheme. Of course, particular attention must be paid to
63: regularization dependent quantities. Another procedure that have been
64: employed based on the same principles is the differential regularization
65: \cite{DifferentialR}, where the amplitudes are evaluated in configuration
66: space. For our purposes here, i.e., evaluate the pion form factor in a
67: regularization independent way in energy-momentum space, IR is more
68: appropriate.
69:
70: Non-renormalizable models do not show the presence of ambiguities, in the
71: sense that the regularization scheme employed is part of the model. This is
72: the case of the Nambu-Jona-Lasinio (NJL) model (\cite{Nambu},\cite{Klevansky}%
73: ), and one can say that Pauli-Villars regularized NJL model is different of
74: the sharp covariant cut-off regularized NJL model. Once defined which
75: regularization scheme will be used on the amplitudes evaluation of the NJL
76: model, also the result of the differences between divergent quantities will
77: be defined. Thus, there are no reasons to leave the ambiguity undetermined
78: in such models. In general, gauge invariant regularization schemes are used
79: to treat the regularized NJL model, but regularization schemes that destroy
80: the quadratic divergence, necessary to do the correct adjustment of the
81: model to the phenomenology, must be avoided. This is the case of the usual
82: gauge invariant Pauli-Villars regularization, and some modifications of the
83: scheme must be done in order to preserve the quadratic divergence and
84: symmetries of the model\cite{Hiller}.
85:
86: But in renormalizable extensions of the NJL model (\cite{NJLRen},\cite{Mota}%
87: ) the regularization scheme is not part of the model, and thus these models
88: can present ambiguities. Due its relative simplicity, the renormalizable
89: extension of the scalar/pseudo-scalar section of the NJL SU(2) model
90: provides a good scenario where one can study the role played by ambiguities
91: in a QFT: (i) it is renormalizable; (ii) it is a fermionic model; (iii) it
92: presents chiral symmetry, in the vanishing fermion mass limit; (iv) it
93: presents parity violating couplings; (v) it presents ambiguities; (vi) there
94: are experimental data available to check its results. Of course there are
95: another models whose fulfill some of the characteristics listed above, such
96: as the chiral Schwinger model, the Gross-Neveu model, and so on.
97:
98: We will adopt here the renormalizable extension of the NJL model presented
99: in \cite{Mota}. This renormalizable non-trivial extension is constructed by
100: using a mean field expansion \cite{Eguchi} that results in a renormalizable
101: but trivial effective Lagrangean \cite{Guralnik}. The first-order effective
102: Lagrangean presents kinetic and mesonic interaction terms radiatively
103: generated, and by augmenting the original effective lagrangean with similar
104: terms, one can avoid the non-renormalizability and triviality of the model
105: \cite{Mota}. This procedure, of course, leads to a model that is different
106: from the original regularized NJL model, but that is related to the latter.
107: In this condition, results are formally identical to the results presented
108: by the NJL model, but without any explicit cut-off dependence (the
109: connection limit can be taken), leaving the model regularization scheme's
110: independent.
111:
112: Both regularized and renormalized versions of the NJL model are appropriated
113: to the study of mesonic properties, and reproduces observables with a
114: reasonable agreement with experimental data. As discussed before, there is
115: not ambiguities in regularized NJL model, since they are fixed by the choice
116: of an specific regularization scheme. But in the renormalizable extension of
117: the NJL model, these ambiguities are present on the calculation of several
118: processes. The results presented in \cite{Mota} do not took in account the
119: influence of these ambiguities, they were fixed by implicitly employing a
120: gauge invariant regularization scheme that fixes them to zero. In this
121: letter, in particular, we will evaluate the pion electromagnetic form factor
122: within the renormalizable extension of the NJL model employing IR and,
123: following the prescription suggested in \cite{Jackiw}, leaving the ambiguity
124: undetermined. We will show that the ambiguity present on the pion form
125: factor cannot be fixed by symmetry relationships (Ward Identity) and does
126: not affect the previsibility of the model.
127:
128: \section{The Pion Charge Form Factor}
129:
130: In the scalar/pseudo-scalar sector of the NJL model, the pion form factor is
131: obtained by evaluating the diagram depicted in figure 1. There is a whole
132: contribution of the vector/axial-vector sector that reproduces the influence
133: of the vector mesons (Vector Dominance Model - VDM) that is missing in the
134: present analysis. In the time-like region the contribution coming from these
135: vector mesons is very important, so we will restrict our analysis to the
136: space-like region. We must also remark that our interest is to study the
137: presence of ambiguities in this process, and thus our results will not be
138: jeopardized by not employing the VDM.
139:
140: \begin{figure}[h]
141: \caption{Feynman diagram for the pion charge form factor in the NJL model.}
142: \label{diagram}
143: \end{figure}
144:
145: The amplitude correspondent to the diagram showed in figure \ref{diagram} is
146: given by
147: \begin{equation}
148: -i\Gamma ^\mu (p,p^{\prime })=N_cN_f\,g_{\pi qq}^2\int \frac{d^4k}{(2\pi )^4}%
149: Tr\{\gamma ^5\frac i{(\not{k}-\not{p})-m}\gamma ^5\frac i{\not{k}-m}\gamma
150: ^\mu \frac i{(\not{k}-\not{q})-m}\}. \label{Amplitude1}
151: \end{equation}
152: with $q=p+p^{\prime }$.
153:
154: Evaluating (\ref{Amplitude1}) by using IR corresponds to isolate, without
155: the use of any explicit regularization, the divergences of the amplitude in
156: terms that are independent of the external momentum. As remarked in \cite
157: {Scarpelli}, we must not make use of symmetrical integration. By proceeding
158: in this way, we obtain (see notes on Appendix)
159: \begin{eqnarray}
160: \Gamma ^\mu &=&2N_Cg_{\pi qq}^2(p^\mu -p^{\prime \mu })\frac 1{4N_Cg_\pi ^2}%
161: +2N_Cg_{\pi qq}^2m_\pi ^2(p^\mu -p^{\prime \mu })\left( Z_0^{\prime }(m_\pi
162: ^2)-I_3(q^2,m_\pi ^2,m_\pi ^2)\right) \label{Amplitude2} \\
163: &&-\,2N_Cg_{\pi qq}^2q_\nu Y^{\mu \nu }, \nonumber
164: \end{eqnarray}
165:
166: where
167: \begin{eqnarray}
168: Z_0(p^2) &=&\frac 1{16\pi ^2}\int_0^1dz\ell n\left[ 1-\frac{p^2}{m^2}%
169: z(1-z)\right] \\
170: &=&-\frac{2i}{(4\pi )^2}\{-1+\frac{\sqrt{p^4-4p^2m^2}}{p^2}\arctan h(\frac{%
171: p^2}{\sqrt{p^4-4p^2m^2}})\} \nonumber
172: \end{eqnarray}
173:
174: \begin{equation}
175: Z_0^{\prime }(p^2)=\frac{dZ_0(p^2)}{dp^2},
176: \end{equation}
177: and
178: \begin{equation}
179: I_3(p,q)=i\int \frac{d^4k}{(2\pi )^4}\frac 1{%
180: (k^2-m^2)[(k-p)^2-m^2][(k-q)^2-m^2]}{.}
181: \end{equation}
182:
183: At this point, we identify the ambiguity in (\ref{Amplitude2}) by the term
184: \begin{equation}
185: Y^{\mu \nu }=2\int \frac{d^4k}{(2\pi )^4}\frac{k^\mu k^\nu }{(k^2-m^2)^3}%
186: -g^{\mu \nu }\int \frac{d^4k}{(2\pi )^4}\frac 1{(k^2-m^2)}.
187: \label{Ambiguity}
188: \end{equation}
189: It is a simple matter to evaluate the quantity in (\ref{Ambiguity}) in
190: different regularization schemes. So, in dimensional regularization or gauge
191: invariant Pauli-Villars one can obtain $Y^{\mu v}=0$, but in covariant sharp
192: cut-off one can obtain $Y^{\mu \nu }=\frac{ig^{\mu \nu }}{2(4\pi )^2}$.
193: Since we are not specifying here any explicit regularization scheme, we will
194: adopt the procedure suggested in \cite{Jackiw} and leave this ambiguity
195: undetermined up to the end of the evaluation. Setting, by using Lorentz
196: invariance,
197: \begin{equation}
198: Y^{\mu \nu }=g^{\mu \nu }\alpha
199: \end{equation}
200: we can write (\ref{Amplitude2}) as
201: \begin{eqnarray}
202: \Gamma ^\mu &=&2N_Cg_{\pi qq}^2(p^\mu -p^{\prime \mu })\frac 1{4N_Cg_\pi ^2}
203: \label{Amplitude3} \\
204: &&+2N_Cg_{\pi qq}^2m_\pi ^2(p^\mu -p^{\prime \mu })\left( Z_0^{\prime
205: }(m_\pi ^2)-I_3(p^2,m_\pi ^2,m_\pi ^2)\right) \nonumber \\
206: &&-\,2N_Cg_{\pi qq}^2q^\mu \alpha . \nonumber
207: \end{eqnarray}
208:
209: The divergent terms present in (\ref{Amplitude3}) were absorbed in the
210: contra-terms of the model, since it is renormalizable, in the same way
211: presented in \cite{Mota}, resulting in renormalized parameters $g_\pi $ and $%
212: g_{\pi qq}$.
213:
214: \section{The Ambiguity}
215:
216: Equation (\ref{Amplitude3}) shows us that the ambiguity belongs to the
217: longitudinal term of the amplitude $\Gamma ^\mu $, the transverse part being
218: free of ambiguities. In order to verify if symmetries (such as gauge
219: invariance) fix the ambiguity, we must verify the Ward identity related to
220: the pion-photon vertex function, that in the NJL model reads
221: \begin{equation}
222: q_\mu \Gamma ^\mu =g_{\pi qq}^2(T^{\pi \pi }(p,q)-T^{\pi \pi }(p))
223: \label{Ward}
224: \end{equation}
225: where
226: \[
227: T^{\pi \pi }(p,q)=\int {\frac{d^4k}{(2\pi )^4}}Tr\{\gamma _5\frac 1{\not{k}-%
228: \not{p}-m}\gamma _5\frac 1{\not{k}-\not{q}-m}\}
229: \]
230: and
231: \[
232: T^{\pi \pi }(p)=\int {\frac{d^4k}{(2\pi )^4}}Tr\{\gamma _5\frac 1{\not{k}-m}%
233: \gamma _5\frac 1{\not{k}-\not{p}-m}\}.
234: \]
235:
236: In the left hand side of eq. (\ref{Ward}), the product of the photon
237: momentum $q_\mu $ by the transverse part of $\Gamma ^\mu $ vanishes. Thus,
238: only the ambiguous part of $\Gamma ^\mu $ survives, leading to
239: \begin{equation}
240: 2N_Cg_{\pi qq}^2q^2\alpha =g_{\pi qq}^2(T^{\pi \pi }(p,q)-T^{\pi \pi }(p))
241: \label{Ward2}
242: \end{equation}
243:
244: The right hand side of eq. (\ref{Ward2}) can be explicitly evaluated, and,
245: also applying IR, the Ward identity (\ref{Ward}) can be exactly verified.
246: Thus, the Ward identity (\ref{Ward}) is satisfied even in the presence of
247: the ambiguity (\ref{Ambiguity}), and does not fix its value. This ensures
248: charge conservation as a feature independent of the presence of the
249: ambiguity.
250:
251: The next question to be answered is how the ambiguity present in the pion
252: electromagnetic vertex function $\Gamma ^\mu $ affects the pion form factor.
253: Let us recall that the amplitude corresponding to figure \ref{diagram} can
254: be obtained by
255: \begin{equation}
256: M=\lim_{p^2,p^{\prime 2}=m_\pi ^2}\varepsilon _{1\mu }\Gamma ^\mu
257: (p,p^{\prime })
258: \end{equation}
259: where $\varepsilon _{1\mu }$ is the polarization vector of the off-shell
260: photon with momentum $q=p+p^{\prime }$. Since $\varepsilon _{1\mu }$ is
261: transverse to the photon momentum, the longitudinal part of $\Gamma ^\mu $
262: will not be present in the amplitude. Thus, this amplitude and, in
263: consequence, the pion charge form factor, will be free from the ambiguity
264: present in $\Gamma ^\mu (p,p^{\prime })$.
265:
266: \section{Numerical Results}
267:
268: Although the ambiguity does not affect the pion form factor, we present here
269: the numerical results obtained for the pion form factor within the
270: renormalizable extension of the NJL model. The results and the parameters
271: set presented here are the same as in \cite{Mota}, we reproduce these
272: results for reasons of completeness. We apply implicit regularization, and
273: fit the following sets of parameters: ($m=350MeV,\,\,g_\pi =3.752$ and $\mu
274: _\pi =141MeV$) and ($m=210MeV,\,\,g_\pi =2.25$ and $\mu _\pi =141MeV$). The
275: parameters are adjusted to reproduce $f_\pi =93.3MeV$ and $m_\pi =139MeV$.
276: The comparison between the results of the renormalizable extension of the
277: NJL model and experimental data are shown in figure \ref{chargeff}. The
278: electromagnetic radius of the pion can be obtained by
279: \begin{equation}
280: <r_\pi ^2>=-6\frac{dF_\pi (q^2)}{dq^2}|_{q^2=0}
281: \end{equation}
282: and, for the sets of parameters above we obtain, respectively
283: \begin{equation}
284: <r_\pi ^2>=0.58fm;\,\,\,<r_\pi ^2>=0.6fm
285: \end{equation}
286: that are to be compared with the experimental result \cite{Amendolia}
287: \begin{equation}
288: <r_\pi ^2>_{\exp }=0.678\pm 0.012fm
289: \end{equation}
290: We observe that a better fit is obtained with the lower constituent quark
291: mass, $m=210MeV$. In fact, if the constituent quark mass is lowered, the fit
292: with the experimental data is improved. But, as stated before, the model
293: studied here does not include the vector/axial-vector meson sector, and a
294: complete agreement with experiment is not to be expected.
295:
296: \begin{figure}[h]
297: \caption{Pion charge form factor in the space--like region. The results
298: obtained from the renormalizable extension of the NJL model are presented
299: for $m=210MeV$ (dotted line) and $m=350MeV$ (solid line) and are compared to
300: experimental data \protect\cite{Amendolia}. }
301: \label{chargeff}
302: \end{figure}
303:
304: \section{Conclusion}
305:
306: In summary, we obtained the pion charge form factor within a renormalizable
307: extension of the SU(2) NJL model with scalar/pseudo-scalar couplings. We had
308: shown that, in this model, the pion-photon vertex function contains an
309: ambiguous term, i.e., a finite but regularization dependent term. We verify
310: that, even in the presence of the ambiguity, the Ward identity related to
311: this vertex function is satisfied, so it does not fix the ambiguity value.
312: Also, in the amplitude related to this process, and consequently in the pion
313: charge form factor, the ambiguity is not present, since it is transverse to
314: the photon polarization vector, and thus decouples from the physical content
315: of the model.
316:
317: This provides an example of one of the roles played by ambiguities in
318: Quantum Field Theory: in previous works, it was shown that ambiguities can
319: be fixed either by symmetry relationships (as in QED\cite{Scarpelli}) or by
320: phenomenology (as in the neutral pion electromagnetic decay or in proton
321: decay\cite{Jackiw}). Here, we found out that there is a third situation,
322: where the ambiguity cannot be fixed by symmetry, but completely decouples
323: from the physical content of the model, remaining independent from the
324: regularization scheme employed.
325:
326: \section{Acknowledgments}
327:
328: This work was supported by CNPq-Brazil.
329:
330: \section{Appendix}
331:
332: In order to show explicitly how we isolated the ambiguity in the pion charge
333: form factor, we will proceed, in this appendix, the computation of the
334: ambiguous part of (\ref{Amplitude1}). After taking the traces on color,
335: flavor and Dirac spaces, we obtain the following two ambiguous integrals
336: \begin{equation}
337: \xi ^{2\mu }=\int \frac{d^4k}{(2\pi )^4}\frac{k^2k^\mu }{%
338: (k^2-m^2)[(k-p)^2-m^2][(k-q)^2-m^2]}, \eqnum{A1}
339: \end{equation}
340: and
341: \begin{equation}
342: p_\mu \xi ^{\mu \nu }=p_\mu \int \frac{d^4k}{(2\pi )^4}\frac{k^\mu k^\nu }{%
343: (k^2-m^2)[(k-p)^2-m^2][(k-q)^2-m^2]}. \eqnum{A2}
344: \end{equation}
345:
346: By adding and subtracting a $m^2k^\mu $ term on the numerator of (A1), we
347: obtain
348: \begin{eqnarray}
349: \xi ^{2\mu } &=&\int \frac{d^4k}{(2\pi )^4}\frac{k^\mu }{%
350: [(k-p)^2-m^2][(k-q)^2-m^2]} \eqnum{A3} \\
351: &&+m^2\int \frac{d^4k}{(2\pi )^4}\frac{k^\mu }{%
352: (k^2-m^2)[(k-p)^2-m^2][(k-q)^2-m^2]} \nonumber \\
353: &=&I_t^\mu +m^2\xi ^\mu \nonumber
354: \end{eqnarray}
355:
356: The last integral in (A3) is finite, and can be evaluated directly, yielding
357: \begin{equation}
358: \xi ^\mu =(p^\mu -p^{\prime \mu })(Z_0^{\prime }(m_\pi ^2)-I_3(q^2,m_\pi
359: ^2,m_\pi ^2)) \eqnum{A4}
360: \end{equation}
361:
362: Introducing one Feynman parameter on $I_t^\mu $, we obtain
363: \begin{equation}
364: I_t^\mu =\int_0^1dx\int \frac{d^4k}{(2\pi )^4}\frac{k^\mu }{[(k-t)^2-M^2]^2}
365: \eqnum{A5}
366: \end{equation}
367:
368: where
369: \begin{equation}
370: M^2=m^2-(p-q)^2x(1-x), \eqnum{A6}
371: \end{equation}
372: and
373: \begin{equation}
374: t^\mu =p^\mu x+q^\mu (1-x). \eqnum{A7}
375: \end{equation}
376:
377: By adding and subtracting a $t^\mu $ term in the numerator of (A5), we
378: obtain
379: \begin{equation}
380: I_t^\mu =\int_0^1dx\left\{ \int \frac{d^4k}{(2\pi )^4}\frac{k^\mu -t^\mu }{%
381: [(k-t)^2-M^2]^2}+t^\mu \int \frac{d^4k}{(2\pi )^4}\frac 1{[(k-t)^2-M^2]^2}%
382: \right\} \eqnum{A8}
383: \end{equation}
384:
385: The last integral in (A8) is logarithmically divergent, an we can safely
386: proceed a shift in the integration momentum $k-t\rightarrow k^{\prime }$.
387: More careful must be taken in the evaluation of the first integral in (A8),
388: since it is linearly divergent. To properly isolate the ambiguity term
389: present in this integral, we introduce the translation operator $e^{-t^\mu
390: \frac \partial {\partial k^\mu }}$ and rewrite (A8) as
391: \begin{equation}
392: I_t^\mu =\int_0^1dx\left\{ \int \frac{d^4k}{(2\pi )^4}e^{-t^\mu \frac
393: \partial {\partial k^\mu }}\left\{ \frac{k^\mu }{(k^2-M^2)^2}\right\} +t^\mu
394: \int \frac{d^4k}{(2\pi )^4}\frac 1{[k^2-M^2]^2}\right\} . \eqnum{A9}
395: \end{equation}
396:
397: Expanding the translation operator, we find
398: \begin{eqnarray}
399: \int \frac{d^4k}{(2\pi )^4}e^{-t^\mu \frac \partial {\partial k^\mu }%
400: }\left\{ \frac{k^\mu }{(k^2-M^2)^2}\right\} &=&\int \frac{d^4k}{(2\pi )^4}%
401: \frac{k^\mu }{(k^2-M^2)^2}-t^\nu \int \frac{d^4k}{(2\pi )^4}\frac \partial {%
402: \partial k^\nu }\left\{ \frac{k^\mu }{(k^2-M^2)^2}\right\} \eqnum{A10} \\
403: &&+\frac{t^\nu t^\rho }2\int \frac{d^4k}{(2\pi )^4}\frac{\partial ^2}{%
404: \partial k^\nu \partial k^\rho }\left\{ \frac{k^\mu }{(k^2-M^2)^2}\right\}
405: +... \nonumber
406: \end{eqnarray}
407:
408: The integrals with odd terms in its integrands in (A10) vanish, and all the
409: terms with derivatives of order greater than 2 will also vanish. We obtain
410: \begin{equation}
411: I_t^\mu =\int_0^1dx\left\{ \int \frac{d^4k}{(2\pi )^4}\frac{-2g^{\mu \nu }}{%
412: (k^2-M^2)^2}+\int \frac{d^4k}{(2\pi )^4}\frac{8k^\mu k^\nu }{(k^2-M^2)^3}%
413: \right\} +t^\mu \int \frac{d^4k}{(2\pi )^4}\frac 1{[k^2-M^2]^2}. \eqnum{A11}
414: \end{equation}
415:
416: We identify the terms in brackets in (A11) as the ambiguous part of $\xi
417: ^{2\mu }$. Nevertheless, there is still a dependence on the integrand of
418: (A11) on $p$, $q$ and $x$, and one can argue if there is some finite
419: contribution coming from (A11). We remark that this term is ambiguous, and
420: the sum of an ambiguity with any finite non-ambiguous term will remain
421: ambiguous. Although this fact, one can find, after successively applying the
422: following relationship
423: \begin{equation}
424: \frac 1{k^2-M^2}=\frac 1{k^2-m^2}+\frac{M^2-m^2}{(k^2-m^2)(k^2-M^2)}
425: \eqnum{A12}
426: \end{equation}
427: and some straightforward calculations, that
428: \begin{eqnarray}
429: &&\int_0^1dx\left\{ \int \frac{d^4k}{(2\pi )^4}\frac{-2g^{\mu \nu }}{%
430: (k^2-M^2)^2}+\int \frac{d^4k}{(2\pi )^4}\frac{8k^\mu k^\nu }{(k^2-M^2)^3}%
431: \right\} \eqnum{A13} \\
432: &=&\int \frac{d^4k}{(2\pi )^4}\frac{-2g^{\mu \nu }}{(k^2-m^2)^2}+\int \frac{%
433: d^4k}{(2\pi )^4}\frac{8k^\mu k^\nu }{(k^2-m^2)^3}=4Y^{\mu \nu }. \nonumber
434: \end{eqnarray}
435:
436: The evaluation of $\xi ^{\mu \nu }$ follows almost the same steps we follow
437: in the evaluation of $\xi ^{2\mu }$. Since it is contracted with the
438: momentum $p$, we made use of
439: \begin{equation}
440: k.p=\frac 12[(k^2-m^2)+p^2+((k-p)^2-m^2)] \eqnum{A10}
441: \end{equation}
442: and obtain
443: \begin{equation}
444: p_\mu \xi ^{\mu \nu }=\frac 12\{p^2\xi ^\mu +I_t^\mu -I_q^\mu \},
445: \eqnum{A11}
446: \end{equation}
447: where $\xi ^\mu $ is given by (A4), $I_t^\mu $ is given by (A13) and
448: \begin{equation}
449: I_q^\mu =\int \frac{d^4k}{(2\pi )^4}\frac{k^\mu }{(k^2-m^2)[(k-q)^2-m^2]},
450: \eqnum{A12}
451: \end{equation}
452: can be computed in the same way we proceeded on the computation of $I_t^\mu $%
453: .
454:
455: %
456: %==========================================================================
457: % REFERENCES
458: %==========================================================================
459: %
460:
461: \begin{references}
462: \bibitem{Colladay} {\small {D. Colladay and V. A. Kostelec\'{y}, Phys. Rev.
463: D 55, 6760 (1997).}}
464:
465: \bibitem{Kostelecky} {\small {R. Jackiw and V. A. Kostelec\'{y}, Phys. Rev.
466: Lett. 82, 3572 (1999).}}
467:
468: \bibitem{Jackiw} {\small {R. Jackiw, }}{\it When Radiative Corrections are
469: Finite but Undetermined, }{\small {MIT-CPT 2835, hep-th/9903044.}}
470:
471: \bibitem{Perez} {\small {M. P\'{e}rez-Victoria, Phys. Rev. Lett. 83, 2518
472: (1999).}}
473:
474: \bibitem{ABJ} {\small {J. S. Bell, R. Jackiw, Nuovo Cimento A 60, 47 (1969).%
475: }}
476:
477: \bibitem{Scarpelli} {\small {A. P. Ba\^{e}ta Scarpelli, M. Sampaio, M. C.
478: Nemes, Phys. Rev. D 63, 046004 (2001).}}
479:
480: \bibitem{Orimar} {\small {O. A. Battistel, A. L. Mota and M. C. Nemes, Mod.
481: Phys. Lett A 13, 1597 (1998).}}
482:
483: \bibitem{Brizola} {\small {A. Brizola, O. A. Battistel, M. Sampaio and M.
484: C. Nemes, Mod. Phys. Lett. A, 1509 (1999).}}
485:
486: \bibitem{DifferentialR} {\small {D. Freedman, K. Johnson, J. I. Latorre,
487: Nuc. Phys. B 371, 353 (1992).}}
488:
489: \bibitem{Nambu} {\small {Y. Nambu, G\ . Jona-Lasinio, Phys. Rev. 122, 345
490: (1961).}}
491:
492: \bibitem{Klevansky} {\small {S. P. Klevansky, {\it Rev. Mod. Phys.}, {\bf 64%
493: }, 3 (1992).}}
494:
495: \bibitem{Hiller} {\small {A. A. Osipov, B. Hiller and A. H. Blin, Phys.
496: Lett. B 475, 324 (2000).}}
497:
498: \bibitem{NJLRen} {\small {K. Langfeld, C. Kettner, and H. Reinhardt, Nuc.
499: Phys. A 608 (1996) 331.}}
500:
501: \bibitem{Mota} {\small {A.L. Mota, M.C. Nemes, H.Walliser,B.Hiller, Nuc.
502: Phys. A 652, 73 (1999).}}
503:
504: \bibitem{Eguchi} {\small {T. Eguchi, Phys. Rev. D 17, 611 (1978).}}
505:
506: \bibitem{Guralnik} {\small {G. S. Guralnik and K. Tamvakis, Nucl. Phys. B
507: 148, 283 (1979).}}
508:
509: \bibitem{Amendolia} {\small {S. R. Amendolia et al., Phys. Lett B 178, 435
510: (1986).}}
511: \end{references}
512:
513: \end{document}
514: