1: \documentstyle[prc,aps,floats,psfig]{revtex}
2: \begin{document}
3: \draft
4:
5: \title{Comment on ``Determination of the chiral coupling constants $c_3$ and
6: $c_4$ in new $pp$ and $np$ partial-wave analyses''}
7:
8:
9: \author{D. R. Entem$^{1,2,}$\footnote{Electronic addresses:
10: dentem@uidaho.edu, entem@usal.es}
11: and R. Machleidt$^{1,}$\footnote{Electronic address: machleid@uidaho.edu}}
12:
13: \address{$^1$Department of Physics, University of Idaho, Moscow, ID 83844, USA\\
14: $^2$Nuclear Physics Group, University of Salamanca, E-37008 Salamanca, Spain}
15:
16:
17: \date{\today}
18:
19:
20: \maketitle
21:
22:
23: \begin{abstract}
24: In a recent study [M.C.M. Rentmeester {\it et al.},
25: Phys.\ Rev.\ C {\bf 67}, 044001 (2003)], the Nijmegen group reports on
26: the determination of the chiral low-energy constants (LEC), $c_3$ and $c_4$,
27: involved in the 2$\pi$-exchange part of the $NN$ amplitude at
28: next-to-next-to-leading order (NNLO) of chiral perturbation theory.
29: This analysis does not apply
30: the uniquely-determined and model-independent $NN$ amplitudes
31: at NNLO and uses, instead, amplitudes that are up 90\% smaller.
32: We point out that this flaw produces a large systematic error,
33: rendering the
34: Nijmegen method unsuitable for a reliable determination of the LEC.
35: \end{abstract}
36:
37: \pacs{PACS numbers: 11.80.Et, 12.39.Fe, 13.75.Cs, 21.30.-x}
38:
39:
40: In a recent paper~\cite{RTS03}, the Nijmegen group reports on a study
41: of the chiral two-pion exchange interaction
42: which contains the low-energy constants (LEC) $c_1$, $c_3$, and $c_4$.
43: The long-range part of this interaction is applied in an analysis
44: of the world $pp$ and $np$ data below 500 MeV laboratory energy.
45: The authors state that, based upon this analysis,
46: they were able to determine the LEC
47: $c_3$ and $c_4$ with an accuracy of 2-5\%.
48: This recent Nijmegen analysis~\cite{RTS03} is an update of
49: an earlier one~\cite{Ren99}, and the present note applies
50: to both.
51:
52: It is the purpose of this Comment to point out that
53: the Nijmegen determination of the LEC appears to have
54: basic
55: flaws which can be estimated to cause a large systematic error.
56:
57: The traditional approach for describing nuclear forces
58: has been the meson-exchange model in which mesons of
59: increasing mass are used to generate contributions of
60: decreasing range. This expansion is then truncated
61: at a certain range which is believed to be unimportant
62: for traditional nuclear physics purposes.
63:
64: The new approach to nuclear forces, that was initiated by
65: Weinberg \cite{Wei90} and pioneered
66: by Ord\'o\~nez~\cite{OK92}, Ray, and van Kolck~\cite{ORK94,Kol99},
67: has a foundation and philosophy which are quite different
68: from meson phenomenology. Based upon the chiral symmetry
69: of QCD, an expansion is made in powers of the nucleon momenta.
70: This is known as chiral perturbation theory ($\chi$PT).
71:
72: The crucial point to realize is that---in contrast to
73: meson phenomenology---$\chi$PT is a {\it theory} rather than a
74: model, i.~e., it makes
75: {\it exact} predictions at each order (if the constants of the
76: theory are known).
77:
78: Physical constants are to be determined in a
79: model-independent way, if by all means possible.
80: While model-independent determinations are {\it per se\/} impossible
81: if one is dealing with a model or phenomenology,
82: model-independence is inherent to the framework of $\chi$PT.
83:
84: At next-to-next-to-leading order (NNLO), the theory consists
85: of one-pion-exchange (OPE), two-pion-exchange (TPE)
86: at ${\cal O}(Q^3)$~\cite{KBW97},
87: where $Q$ denotes a momentum or pion
88: mass, and contact terms of order $Q^0$ and $Q^2$. These
89: contributions define {\it uniquely\/} the $NN$ amplitude
90: at NNLO. A problem is that the parameters of the contact
91: terms are not known. However, since these contacts
92: contribute only to $S$ and $P$ wave amplitudes,
93: the $NN$ amplitudes for partial waves with orbital
94: angular momentum $L\geq 2$ do not receive
95: any contact contributions~\cite{EGM98}.
96: Consequently, at NNLO, the $NN$ amplitudes for $D$ and
97: higher partial waves are determined
98: by OPE and TPE at ${\cal O}(Q^3)$
99: in an unique and completely model-independent way~\cite{KBW97,EGM98}.
100: These two contributions depend
101: (apart from the nucleon and pion mass)
102: on five constants:
103: the axial-vector coupling constant, $g_A$,
104: the pion decay constant, $f_\pi$,
105: and three LEC that appear in the dimension-two $\pi N$
106: Lagrangian and are, conventionally, denoted by
107: $c_1$, $c_3$, and $c_4$.
108: The constants $g_A$ and $f_\pi$ are known from other sources
109: (we use $g_A=1.29$ and $f_\pi=92.4$ MeV and the same is used
110: in Ref.~\cite{RTS03}) and for $c_1$ the educated estimate
111: $c_1=-0.76$ GeV$^{-1}$ can be made~\cite{RTS03,Ren99}.
112: Now only two constants are open, namely, $c_3$ and $c_4$.
113: Since the LEC are also involved in the $\pi N$ amplitude,
114: these constants have been determined by calculating
115: the $\pi N$ amplitude up to a certain order of $\chi$PT
116: and adjusting the constants such that
117: the empirical $\pi N$ information is fit~\cite{FMS98,BM00}.
118: Similarly, the
119: $NN$ amplitudes of partial waves with $L\geq 2$
120: can be compared with the corresponding empirical
121: amplitudes to extract $c_3$ and $c_4$.
122:
123: Once all five constants are fixed, then there exists
124: a unique prediction for all $NN$ amplitudes with $L\geq 2$.
125: Using $c_3=-4.78$ GeV$^{-1}$ and $c_4=3.96$ GeV$^{-1}$,
126: these uniquely determined $NN$ amplitudes are shown
127: in Fig.~1 by the solid lines for one $D$ and one
128: $F$ wave case (see Refs.~\cite{KBW97,EM02} for the
129: details of how these amplitudes are calculated
130: perturbatively; conventions as in Ref.~\cite{EM02}).
131:
132: {\it It must be stressed again that we are dealing
133: here with a theory and that a theory makes
134: exact predictions. Thus, the solid line in Fig.~1 is
135: the unique result of $\chi$PT at NNLO for the NN system
136: (using the above constants).
137: Using the same constants, there is no other result
138: at NNLO and every theoretical physicist will independently
139: reproduce this result.}
140:
141: However, when we compare this result with the one the Nijmegen
142: group obtains and uses in their analysis (dashed line in Fig.~1),
143: we notice a discrepancy of up to 90\%.
144: When such amplitudes are applied in an analysis,
145: a large systematic
146: error is produced and the extracted constants
147: will carry such large error.
148:
149: The Nijmegen amplitudes are obtained as a result
150: of representing
151: the chiral OPE and TPE at NNLO
152: in terms of an $r$-space potential that is cut off
153: at $r=1.6$ fm.
154: The motivation for doing this may be a confusion between
155: the old range argument (that applies to meson-model
156: based potentials) and the fundamentals of
157: $\chi$PT which is an expansion in powers of nucleon
158: momenta, and not in terms of ranges.
159: As pointed out before,
160: $\chi$PT has, by principle, no model-dependence,
161: whereas model-dependence is clearly introduced
162: in the analyses of Refs.~\cite{RTS03,Ren99}.
163:
164: Another point of concern is that the Nijmegen group uses
165: $\chi$PT at NNLO up to
166: 500 MeV. It is wellknown that $\chi$PT at NNLO
167: is appropriate only up to 50 MeV in $D$ waves and
168: about 150 MeV in $F$ waves~\cite{KBW97,EM02}.
169:
170: In conclusion,
171: the method of determination of the chiral LEC, $c_3$ and $c_4$,
172: applied in the recent as well as the earlier Nijmegen
173: analyses~\cite{RTS03,Ren99} may be unreliable.
174:
175: This work was supported in part by the U.S. National Science
176: Foundation under Grant No.~PHY-0099444,
177: the Spanish Ministerio de Ciencia y Tecnolog{\'\i}a
178: under Contract No. BFM2001-3563, and the Spanish Junta de Castilla y Le\'on under
179: Contract No. SA-109/01.
180:
181:
182: \begin{figure}
183: \vspace{-5cm}
184: \hspace{2cm}
185: \psfig{figure=fig1.ps,height=16.0cm}
186: \vspace{-6.5cm}
187: \caption{On-shell $NN$ amplitudes
188: up to 500 MeV laboratory energy
189: for the $^1D_2$ and $^3F_4$ states.
190: The solid lines represent the correct amplitudes
191: for OPE plus TPE at NNLO (using the constants given in the text).
192: The dashed lines show the amplitudes used in the Nijmegen
193: analysis for OPE plus TPE at NNLO (using the same constants).
194: Thick lines represent the real parts of the amplitudes
195: and thin lines the imaginary parts. (Note that the thin lines are $\approx 0$
196: on the scale of the figures
197: and hardly distinguishable.)}
198: \end{figure}
199:
200:
201: \begin{references}
202: \bibitem{RTS03} M. C. M. Rentmeester, R. G. E. Timmermans,
203: and J. J. de Swart,
204: Phys.~Rev.~C {\bf 67}, 044001 (2003).
205: \bibitem{Ren99} M. C. M. Rentmeester, R. G. E. Timmermans,
206: J. L. Friar, and J. J. de Swart,
207: Phys. Rev. Lett. {\bf 82}, 4992 (1999).
208: \bibitem{Wei90} S. Weinberg, Phys.\ Lett.\ B {\bf 251}, 288 (1990);
209: Nucl.\ Phys.\ {\bf B363}, 3 (1991).
210: \bibitem{OK92}
211: C. Ord\'o\~nez and U. van Kolck,
212: Phys.\ Lett.\ B {\bf 291}, 459 (1992).
213: \bibitem{ORK94}
214: C. Ord\'o\~nez, L. Ray, and U. van Kolck,
215: Phys.\ Rev.\ Lett.\ {\bf 72}, 1982 (1994);
216: Phys.\ Rev.\ C {\bf 53}, 2086 (1996).
217: \bibitem{Kol99} U. van Kolck, Prog.\ Part.\ Nucl.\ Phys.\ {\bf 43}, 337 (1999).
218: \bibitem{KBW97} N. Kaiser, R. Brockmann, and W. Weise,
219: Nucl.\ Phys.\ {\bf A625}, 758 (1997).
220: \bibitem{EGM98} E. Epelbaum, W. Gl\"ockle, and U.-G. Mei\ss ner,
221: Nucl.\ Phys.\ {\bf A671}, 295 (2000).
222: \bibitem{FMS98} N. Fettes, U.-G. Mei\ss ner, S. Steiniger,
223: Nucl.\ Phys.\ {\bf A640}, 199 (1998).
224: \bibitem{BM00} P. B\"{u}ttiker and U.-G. Mei\ss ner,
225: Nucl.\ Phys.\ {\bf A668}, 97 (2000).
226: \bibitem{EM02} D. R. Entem and R. Machleidt, Phys. Rev. C {\bf 66}, 014002 (2002).
227: \end{references}
228:
229: \end{document}
230: