hep-ph0310070/ls2.tex
1: \documentclass[prd,aps,superscriptaddress]{revtex4}
2: %\documentclass[twocolumn,prd,aps]{revtex4}
3: 
4: \usepackage{graphicx}% Include figure files
5: \usepackage{dcolumn}% Align table columns on decimal point
6: \usepackage{bm}% bold math
7: 
8: \def\A{{\bf A}}
9: \def\B{{\bf B}}
10: \def\x{{\bf x}}
11: \def\r{{\bf r}}
12: \def\y{{\bf y}}
13: \def\k{{\bf k}}
14: \def\s{{\bf s}}
15: \def\l{{\bf l}}
16: \def\q{{\bf q}}
17: \def\z{{\bf z}}
18: \def\D{{\bf D}}
19: \def\P{{\bf P}}
20: \def\p{{\bf p}}
21: \def\E{{\bf E}}
22: \newcommand{\bdel} {{\mbox{\boldmath $\nabla$}}}
23: 
24: 
25: \begin{document}
26: \voffset = 0.3 true in
27: \topmargin = -1.0 true in % a mac teTex requirement...
28: 
29: \title{A Coulomb Gauge Model of Mesons}
30: 
31: \author{Norbert Ligterink\footnote{Current address:University of Twente, Control Engineering
32: PO Box 217, 7500 AE Enschede, The Netherlands}
33: }
34: \affiliation{
35: Department of Physics and Astronomy, University of Pittsburgh,
36: Pittsburgh PA 15260}
37: 
38: \author{Eric S. Swanson}
39: \affiliation{
40: Department of Physics and Astronomy, University of Pittsburgh,
41: Pittsburgh PA 15260}
42: \affiliation{
43: Jefferson Lab, 12000 Jefferson Ave,
44: Newport News, VA 23606}
45: 
46: 
47: \vskip .5 true cm
48: \begin{abstract}
49: A model of mesons which is based on the QCD Hamiltonian in Coulomb
50: gauge is presented. The model relies on a novel quasiparticle basis to
51: improve the reliability of the Fock space expansion. It is also 
52: relativistic, yields chiral pions, and is tightly constrained by QCD 
53: (quark masses are the only parameters). 
54: Applications to hidden flavor mesons yield
55: results which are comparable to phenomenological constituent quark 
56: models while revealing the limitations of such models.
57: \end{abstract}
58: 
59: 
60: %\pacs{}
61: \maketitle
62: 
63: \section{Introduction}
64: 
65: The successes of the quark model of the 1960's led directly to 
66: the development of QCD in the early 1970's.  A central feature of
67: the early quark model was the use of constituent quarks as
68: the relevant degrees of freedom of matter fields. Although the 
69: advent of QCD changed the details -- the ``light" constituent
70: quarks of Copley, Karl, and Obryk have become standard and one gluon
71: exchange is typically employed to describe short 
72: range dynamics -- the concept of constituent quarks has remained productive 
73: and pervasive.
74: 
75: 
76: QCD also indicates where the quark model may fail. The canonical 
77: nonrelativistic quark model relies on a potential description of
78: quark dynamics and therefore neglects many-body effects in QCD. 
79: Related to this is the question of the reliability of nonrelativistic
80: approximations, the importance of hadronic decays, and the chiral 
81: nature of the pion.  The latter two phenomena depend 
82: on the behavior of nonperturbative glue and as such are crucial to
83: the development of robust models of QCD and to understanding soft gluodynamics.
84: Certainly,  one expects that gluodynamics will make its presence felt with
85: increasing insistence as experiments 
86: probe higher excitations in the spectrum. Similarly the chiral nature of
87: the pion cannot be understood in a fixed particle number formalism.  This
88: additional complexity is the reason so few models attempt to derive the
89: chiral properties of the pion.  This is an unfortunate situation since
90: the pion is central to much of hadronic and nuclear physics.
91: 
92: To make progress one must either resort to numerical experiments or 
93: construct models which are closer to QCD. Here we present one such model
94: which is based on the QCD Hamiltonian in Coulomb gauge. The Hamiltonian
95: approach is appropriate for an examination of the bound state problem
96: because the familiar machinery of quantum  mechanics may be employed and
97: because all degrees of freedom are physical in Coulomb gauge. Furthermore,
98: an explicit time-independent potential exists which permits the construction
99: of bound states in a fixed Fock sector.
100: The model consists of a truncation of QCD to a set of diagrams which 
101: capture the infrared dynamics of the theory. The efficiency of the
102: truncation is enhanced through the use of quasiparticle degrees of
103: freedom, as will be explained subsequently. Finally, the random phase
104: approximation (RPA) is used to obtain mesons. This many-body truncation is 
105: sufficiently powerful to generate Goldstone bosons and has the
106: advantage of being a relativistic truncation of QCD\cite{RS}.
107: 
108: Because the Hamiltonian is derived from a local density it is covariant,
109: although the use of Coulomb gauge hides this. The truncations
110: which we will employ do not ruin this property.
111: We remark that covariance requires a combination of boost
112: and gauge transformations in noncovariant gauges
113: and therefore some care must be taken in the
114: computation of quantities such as form factors or heavy meson
115: decay rates. Here we focus on static meson properties 
116: in the rest frame where these issues do not arise. Finally, we
117: note that maintaining relativistic invariance in schemes which
118: extend the RPA may be difficult because the interaction is no longer instantaneous
119: at higher order. Thus 
120: different terms must be summed to yield covariant
121: results and amplitudes may arise which do not have simple wavefunction, 
122: or RPA, analogues such as amplitudes with a mixture of forward and backward 
123: moving particles.
124: 
125: 
126: 
127: 
128: 
129: 
130: Using a single framework to generate chiral
131: symmetry breaking and the meson spectrum consistently
132: has been attempted before. LeYaouanc {\it et al.}
133: \cite{LeYaouanc:1984dr} solved a simple gap equation with a
134: quadratic interaction and then used the RPA approximation to
135: obtain chiral pions. Although the interaction is unrealistic
136: it allowed the important simplification of turning integral
137: equations into simple differential equations.
138: Llanes-Estrada and Cotanch \cite{LC} also studied
139: low-lying states with a linear potential while ignoring
140: state mixing.
141: Neither paper considered the effects of the one gluon exchange potential or
142: renormalization. 
143: 
144: An extensive literature on relativistic quark
145: models (which do not consider chiral symmetry breaking) exists.
146: For example, a preliminary study of Fock sector mixing in a
147: relativistic quark model was performed by Koniuk and Zhang\cite{ZK}.
148: Detailed examinations of meson and baryon properties have been carried
149: out by the Bonn group\cite{bonn} in a Salpeter equation framework with
150: a model confinement potential and Dirac structure. Similarly, extensive
151: computations of pion and kaon properties have been carried out in 
152: a covariant Euclidean space Dyson-Schwinger formalism\cite{MR}.
153: 
154: Finally, the gluonic portion of the formalism presented here has been 
155: used to derive the quenched 
156: positive charge conjugation glueball spectrum\cite{ss8}. The results are
157: in very good agreement with lattice computations, indicating that the 
158: method has some promise.
159: 
160: 
161: \section{Model Definition: the Quark Vacuum and Chiral Symmetry Breaking}
162: 
163: Generating the meson spectrum proceeds in three steps: (1) a quasiparticle
164: basis for the gluonic sector of QCD is obtained with standard many-body 
165: techniques, (2) this procedure yields an instantaneous interaction which 
166: is used to construct a quasiparticle (constituent) basis in the quark 
167: sector, (3) bound state properties are obtained in the the random phase 
168: approximation.
169: The first step contains an important
170: complication: the quasiparticle interaction of QCD depends on the
171: quasiparticles themselves (in a way made clear below) and hence must
172: be solved along with the gap equation. This allows the possibility of
173: deriving the constituent quark interaction if one can obtain the
174: functional form of the interaction.  We note that this is similar to
175: solving coupled Dyson-Schwinger equations for, say, the gluon propagator
176: and vertices.
177: 
178: \subsection{QCD in Coulomb Gauge}
179: 
180: 
181: 
182: The Coulomb gauge QCD Hamiltonian may be written as $H_{QCD} = H_0 + H_{int}$
183: with\cite{CL,schwinger, ss7}
184: 
185: 
186: \begin{equation}
187: H_0 = \int \psi^\dagger \left( - i \bm{\alpha} \cdot \bdel + \beta m\right)
188: \psi + {1 \over 2}\int d\x  \left( \bm{\Pi}^2 - \A \bdel^2 \A \right)
189: + {1\over 2}\int d\x d\y\, \rho^a(\x) K^{(0)}(\x-\y) \rho^a(\y) \ \ .
190: \label{h0}
191: \end{equation}
192: The interaction term contains the familiar transverse gluon color charge 
193: interaction and all of the higher order terms 
194: due to the non-Abelian nature of QCD:
195: \begin{eqnarray}
196: H_{int} &= {1\over 2} \int d\x \left[ \B^2 + \A \bdel^2 \A \right]  -
197: g \int \psi^\dagger \bm{\sigma}\cdot \A \psi + V_A + V_B +\nonumber \\
198: & +  {1\over 2}\int d\x d\y\, \rho^a(\x) \left[ K^{ab}(\x-\y;\A) - \delta^{ab}K^{(0)}(\x-\y)
199: \right] \rho^b(\y)
200: \label{hint}
201: \end{eqnarray}
202: The density $\rho^a$ entering these equations is the full color charge due 
203: to quarks and gluons
204: 
205: \begin{equation}
206: \rho^a({\bf x}) =
207:  f^{abc} {\bf A}^b({\bf x}) \cdot \bm{\Pi}^c({\bf x}) +
208:  \psi^{\dag}(\x){\lambda^a\over 2}\psi(\x).
209: \label{rho}
210: \end{equation}
211: The instantaneous non-Abelian Coulomb interaction in Eq. \ref{hint} 
212: is given by 
213: 
214: \begin{equation}
215: K^{ab}({\bf x},{\bf y};\A) \equiv \langle{\bf x},a|
216:  { g \over { \bdel\cdot {\bf D}}}(-\bdel^2)
217:  { g \over { \bdel\cdot{\bf D}}}|{\bf y},b\rangle,
218: \label{coulk}
219: \end{equation}
220: where $\D$ is the covariant derivative in adjoint representation
221: \begin{equation}
222: \D^{ab} = \delta^{ab}\bdel - g  f^{abc} \A^c.
223: \end{equation}
224: The electric and magnetic fields are defined by
225: 
226: \begin{equation}
227: \Pi^a \equiv -\E_{tr}^a = \dot\A^a + g (1 - \bdel^{-2} \bdel
228: \bdel\cdot) f^{abc} A^{0b} \A^c
229: \end{equation}
230: and
231: \begin{equation}
232: \B^a = \bdel \times \A^a + {1\over 2} g f^{abc} \A^b \times \A^c.
233: \end{equation}
234: 
235: 
236: 
237: The interaction in $H_0$ is defined as the vacuum expectation value
238: of the Coulomb interaction
239: 
240: \begin{equation}
241: \delta ^{ab} K^{(0)}(\x-\y) = \langle \Psi_0 \vert K^{ab}(\x,\y;\A)\vert \Psi_0\rangle.
242: \label{k0}
243: \end{equation}
244: The vacuum state will be defined shortly.
245: 
246: 
247: Finally, the imposition of Coulomb gauge restricts the theory to a curved gauge
248: manifold with a metric given by $\langle \Phi| \Psi \rangle = \int {\cal D}\A {\cal J}[\A] \Phi^*(\A)\Psi(\A)$.  The factor ${\cal J}$ is the Faddeev-Popov 
249: determinant given by 
250: 
251: \begin{equation}
252: {\cal J} = {\rm det}(\bdel \cdot \D).
253: \end{equation}
254: 
255: The Faddeev-Popov determinant may be removed from the metric by rescaling the
256: Hamiltonian
257: 
258: $$ 
259: H \to {\cal J}^{1/2} H {\cal J}^{-1/2}
260: $$
261: which is hermitian with respect to $( \Phi|\Psi ) = \int {\cal D\A}
262: \Phi^*(\A) \Psi(\A)$.  We note that  the nontrivial metric induces 
263: two new terms, denoted 
264: $V_A$ and $V_B$, which correspond to the anomalous terms of
265: Schwinger\cite{schwinger,CL}.
266: 
267: 
268: It has been speculated that confinement is related to the well-known
269: Gribov ambiguity\cite{gribov,Z}. The Gribov ambiguity arises because the existence
270: of topologically inequivalent solutions to the gauge condition $\bdel \cdot \A = 0$ implies that the gauge is incompletely specified. 
271: Gribov proposed that the ambiguities with Coulomb gauge may be
272: resolved by considering fields with $\det(\bdel\cdot\A) \geq 0$ which 
273: comprise the `Gribov region' (GR). We note that the appearance of the inverse of
274: the Faddeev-Popov operator in the instantaneous Coulomb potential implies
275: that gauge configurations near the boundary of the Gribov region create a
276: strong infrared enhancement in the interaction. It is possible that this
277: enhancement is the origin of confinement in Coulomb gauge\cite{Z}. 
278: 
279: 
280: Subsequent research has shown that the Gribov region actually contains gauge
281: copies and hence gauge fields must lie in a subset of the Gribov region
282: called the fundamental modular region (FMR).
283: The FMR contains no redundant field
284: configurations and may be defined, for example, as the
285: set of global minima over gauge transformations of the functional 
286: \begin{equation}
287: F_\A [g] = {\rm Tr}\int d^3x (\A^g)^2
288: \end{equation}
289: where $\A^g = g\A g^\dagger - g\bdel g^\dagger$. It is now known that the
290: FMR contains the GR except at certain points where the regions
291: coincide, the FMR is convex, and the GR contains the origin\cite{vanbaal}. 
292: Furthermore, Zwanziger has recently argued\cite{Z3} that observables have their
293: support over the intersection of the FMR and the GR, thereby resurrecting
294: the infrared enhancement argument of the previous paragraph. Finally,
295: topological aspects of QCD  may be introduced to the formalism via the 
296: imposition of wavefunctional boundary
297: conditions on the boundary of the FMR\cite{vanbaal}.
298: 
299: 
300: 
301: 
302: \subsection{Gluon Gap Equation and the Instantaneous Interaction}
303: 
304: It is clear that the interaction of quarks is strongly influenced by
305: the instantaneous Coulomb interaction of Eq. \ref{coulk}.  It is known,
306: moreover, that the instantaneous interaction is renormalization group 
307: invariant\cite{Z}.
308: This fact permits a physical interpretation of the instantaneous
309: potential which is a central aspect of our formalism. Furthermore, Zwanziger
310: has shown that the
311: Coulomb interaction provides an upper bound to the Wilson loop potential
312: and has postulated that this bound is, in fact, saturated\cite{Z1}. We
313: remark that this saturation is crucial to the formalism presented here.
314: This postulate may be checked with lattice computations, unfortunately the
315: results are mixed \cite{ZC,G}. If the evidence to the
316: contrary is confirmed, this approach must be abandoned.
317: 
318: 
319: The nonpolynomial functional dependence of the Coulomb interaction on
320: the vector potential complicates computations in Coulomb gauge.
321: We proceed by separating the interaction 
322: into two parts denoted by $K^{(0)}$ (see Eq. \ref{k0}) and $K-K^{(0)}$.
323: Since $K^{(0)}$ is a vacuum expectation value it contains all diagrams
324: in which gluons are attached to the operator $\rho K \rho$.  Thus the
325: remainder  necessarily contains gluons which propagate. Since gluons
326: are quasiparticles  with a dynamical mass on the order of 1 GeV (the 
327: hybrid-meson mass gap), matrix elements of $K-K^{(0)}$ are suppressed
328: in typical hadronic observables, considerably simplifying computations.
329: 
330: 
331: In spite of this, it is impossible to obtain a closed form expression for the
332: the vacuum expectation value of $K$. However, a procedure for obtaining 
333: Dyson equations which sum the leading infrared diagrams for this matrix element
334: has been described in Ref.\cite{ss7}.  
335: An important element in the formalism is the construction of a basis
336: which permits an efficient Fock space expansion. The method proposed in 
337: Ref. \cite{ss7} built this basis with the aid of a gluonic vacuum Ansatz (it is worth stressing that
338: the vacuum need not be highly accurate, it merely provides a starting point for
339: constructing a more general quasiparticle basis, and any sufficiently general
340: vacuum Ansatz will suffice). The Ansatz employed is
341: 
342: \begin{equation}
343: \Psi_0[\A] = \langle \A | \omega \rangle =  \exp\left[-{1\over 2} \int d\k
344:    \A^a(\k)\,\omega(k)\A^a(-\k) \right].
345: \label{varvac}
346: \end{equation}
347: which depends on an unknown trial function, $\omega(k)$. This is the simplest
348: correlation one can build into the vacuum and corresponds to the BCS Ansatz
349: of many-body physics.
350: Note that the perturbative vacuum is obtained when $\omega = |\k|$. 
351: 
352: The trial
353: function is obtained by minimizing the vacuum energy density
354: 
355: \begin{equation}
356:   {{\delta}\over {\delta \omega}} \langle \omega |H |\omega \rangle  = 0.
357:   \label{vev}
358: \end{equation}
359: The vacuum state obtained from this procedure is denoted $|\omega\rangle$.
360: We refer to
361: $\omega$ as the gap function since it is also
362: responsible for lifting the single particle gluon
363: energy beyond its perturbative value.
364: Of course evaluating the matrix element in Eq. \ref{vev} requires an explicit
365: expression for $\langle\omega |\rho K \rho|\omega\rangle$ which is provided 
366: by the Dyson procedure described above. The result is a set of coupled 
367: nonlinear integral equations
368: which describe the gap equation and the Dyson equation for $K^{(0)}$.  Solving
369: these equations yields both the quasigluon dispersion relation, $\omega(k)$,
370: and the quasigluon effective interaction, $K^{(0)}(\x-\y)$. Renormalization
371: is achieved by fitting the latter
372: to the lattice Wilson loop potential. The result is in excellent agreement with
373: the lattice and provides a dynamical mass scale for the quasigluons: $m_g \equiv 
374: \omega(0) \approx 600$ MeV.  
375: It is this large mass scale which permits rapid 
376: convergence of any Fock series expansion in the gluonic sector of QCD and explains
377: why quark degrees of freedom dominate low energy hadronic physics.
378: We remark that the emergence of a confining potential
379: is nontrivial and indicates the robustness of the method.
380: 
381: An analytic approximation to the solution to the coupled equations yields the
382: following form  for the vacuum expectation value of the Coulomb interaction\cite{ss7}:
383: 
384: 
385: 
386: 
387: \begin{equation}
388: K^{(0)}(k) = {12.25\over k^2} \left\{ \begin{array}{ll}
389:                                \left({m_g \over k}\right)^{1.93} & k < m_g\\
390:                                { 0.6588 \log(k^2/m_g^2 + 0.82)^{-0.62} \log(k^2/m_g^2 + 1.41)^{-0.80}} & k > m_g 
391:                                       \end{array} \right.
392: \label{Vss7}
393: \end{equation}
394: This form mimics the numerical solution and the lattice results quite well. The
395: long range potential behaves as $k^{-3.93}$, within 2\% of the expected linear
396: behavior (the deviation is likely to be a numerical artifact). This allows the
397: extraction of a string tension via $6 \pi b \approx 12.25 m_g^2$, which implies
398: $b \approx 0.234$ GeV$^2$. The effect of the smaller exponent is to reduce this
399: string tension to approximately 0.21 GeV$^2$  at physically relevant scales. A
400: fit to the standard `Coulomb plus linear' potential (const - $4/3\cdot \alpha_s/r + br$)
401: yields an effective strong coupling of $\alpha_s = 0.12$. This value is rather
402: small with respect to expectations (in part because there was little lattice data
403: at small distances in the fit of Ref.\cite{ss7}). Thus we have also employed a
404: modified potential of the form
405: 
406: 
407: \begin{equation}
408: K^{(0)}(k) = {12.25\over k^2} \left\{ \begin{array}{ll}
409:                                \left({ m_g \over k}\right)^{1.93} & k < m_g\\
410:                                \left({\log(1+a)\over \log(k^2/m_g^2 + a)}\right)^b & k > m_g 
411:                                       \end{array} \right.
412: \label{Vmod}
413: \end{equation}
414: Here $a$ and $b$ are constants which may be varied. Values of $a = 1.0$ and $b=0.8$ 
415: give a potential which agrees well with lattice data at short distance, 
416: but which otherwise is very close to the original form of Eq.\ref{Vss7}.
417: Both potentials have been employed in the following to test the sensitivity of the 
418: results on the functional form of the potential.
419: 
420: 
421: %\begin{figure}
422: %\includegraphics[scale=0.4,angle=-90]{V140.ps}
423: %\caption{ \label{V}Static $Q{\bar Q}$ ground state potential.
424: %[[[Show models and a fit to lattice]] }
425: %\end{figure}
426: 
427: 
428: 
429: \subsection{Quark Gap Equation}
430: 
431: 
432: Our chief goal is to examine a model of QCD which permits the simultaneous
433: description of heavy quarkonia and chiral pions.  It is clear that this
434: may not be achieved in a potential formalism -- the many-body aspect of
435: QCD is required. Our approach to the quark sector therefore mimics that
436: of the gluon sector, namely we model the quark vacuum with a Gaussian 
437: wavefunctional and determine the quark gap equation. Solving this yields
438: dynamical chiral symmetry breaking and a massless Goldstone boson in the
439: random phase approximation in accord with the Thouless sum rule\cite{thouless}. 
440: We note that this general approach has been used many times in the past, starting with 
441: the classic work of Nambu and Jona-Lasinio \cite{NJL}.  Subsequent work has dealt
442: with renormalization issues \cite{A,FM} or with models which are closer to 
443: QCD \cite{Orsay}. How the constituent quark model may be reconciled with chiral
444: symmetry breaking is explained in Ref. \cite{ss5}. Finally an extensive literature on the Dyson-Schwinger approach to
445: this problem exists\cite{MR}.
446: 
447: 
448: In this approach the gap equation represents a nonperturbative one loop
449: computation and thus must be properly renormalized. As noted in the gluonic
450: sector, this is a nontrivial step whose implementation depends on the subset
451: of diagrams being summed. In the BCS/RPA formalism employed here, we have 
452: found that standard
453: renormalization is sufficient to guarantee finite results. In particular we
454: have added mass renormalization 
455: 
456: $$
457:  \delta H_m = \delta m\int d\x\, \psi^\dagger \beta \psi 
458: $$
459: and wavefunction renormalization
460: $$
461: \delta H_\psi = \left( Z_\psi - 1\right) \int d\x\,\psi^\dagger (-i {\bm \alpha}\cdot \bdel) \psi
462: $$
463: terms to the Hamiltonian of Eq. \ref{h0} and the theory has been truncated at 
464: the scale $\Lambda$. Recall that the effective instantaneous interaction
465: has already been rendered finite.
466: 
467: Proceeding with the standard Bogoliubov or Dyson procedure yields the following
468: quark gap equation
469: 
470: \begin{equation}
471: Z_\psi(\Lambda) p s_p = \left[m(\Lambda) + \delta m(\Lambda)\right]c_p + 
472: {C_F \over 2} \int {k^2 dk \over (2\pi)^3} \left[ V_0(p,k) s_k c_p - V_1(p,k) c_k s_p\right]
473: \label{qgap}
474: \end{equation}
475: 
476: %\begin{equation}
477: %{\rm reg} = -{C_F\over 2} \int [dk] \left( K_0^{IR}(p,k) - K_1^{IR}(p,k) {k \over p}\right)s_p c_p
478: %\end{equation}
479: 
480: \noindent
481: where 
482: $$
483: V_L(p,k) = 2\pi \int d(\hat p \cdot \hat k)\, K^{(0)}(\p-\k) P_L(\hat p \cdot\hat k).
484: $$
485: The functions $s_k$ and $c_k$ are defined in terms of the Bogoliubov angle $\phi(k)$
486: as $s_k = \sin\phi(k)$ and $c-k = \cos \phi(k)$. 
487: The quark gap equation is to be solved for the unknown Bogoliubov angle, which 
488: then specifies the quark
489: vacuum and the quark field mode expansion via spinors of the form
490: 
491: \begin{equation}
492: u_s(k) = \sqrt{1 + s_k \over 2} \left( \begin{array}{c}
493:                                      \chi_s  \\
494:                                       {c_k\over 1 + s_k}  {\bm{\sigma} \cdot \hat k} \chi_s
495:                                  \end{array} 
496:                                   \right).
497: \end{equation}
498: Comparing the quark spinor
499: to the canonical spinor (we use nonrelativistic normalization) permits a simple
500: interpretation of the Bogoliubov angle through the relationship $\mu(k) = k \tan\varphi(k)$ where $\mu$ may 
501: be interpreted as a dynamical momentum-dependent quark mass. Similarly $\mu(0)$
502: may be interpreted as a constituent quark mass.
503: 
504: 
505: 
506: In the case of massless quarks the right hand side of the quark gap equation 
507: diverges logarithmically for potentials
508: obeying the perturbative relation $K^{(0)}(k) \to k^{-2}$ for large $k$. The 
509: divergence 
510:  may be absorbed
511: into the wavefunction renormalization, $Z_\psi = 1 - C_F/(6\pi^2) \log\Lambda$,
512: yielding a finite gap equation.  It is also possible to renormalize by examining
513: the once-subtracted gap equation. 
514: For the massive quark case two logarithmic divergences proportional
515: to the quark mass and momentum appear. It is convenient to absorb these divergences
516: separately into the mass and wavefunction terms respectively.
517: For the study presented here the potential is modified by logarithmic corrections 
518: at short
519: distances, thus all integrals are finite and the cutoff may be removed immediately.
520: We note, however, that it still may be useful to make finite renormalizations.
521: 
522: 
523: The numerical solution for the dynamical quark mass is very accurately represented
524: by the functional form
525: 
526: \begin{equation}
527: \mu(k) = \sigma K^{(0)}(k)\left( 1 - {\rm e}^{-{M /(\sigma K^{(0)}(k)})}\right)
528: \label{muEq}
529: \end{equation}
530: where $M$ is a `constituent' quark mass and $\sigma$ is a parameter related to the
531: quark condensate. Notice that this form approaches  the constituent mass for small 
532: momenta and $\sigma K^{(0)}$ for large momenta. The latter behavior is in accord with
533: the quark gap equation which implies that\cite{P}
534: 
535: \begin{equation}
536: \mu(k) \to {C_F\over 2} \int {d^3q \over (2\pi)^3} K^{(0)}(k) s(q)
537: = -{C_F \over 4 N_c} K^{(0)}(k) \langle \bar \psi \psi \rangle.
538: \label{muk}
539: \end{equation}
540: 
541: 
542: A rough fit to the numerical solution yields $M = 68$ MeV and $\sigma =$ 0.001 GeV$^3$. The constituent quark mass is small compared to typical relativistic
543: constituent quark model masses of roughly 200 MeV.   The value for $\sigma$ implies
544: a quark condensate of approximately (-210 MeV)$^3$, in reasonable agreement with
545: current estimates of (-250 MeV)$^3$. However, we note that direct computations of the condensate 
546: typically yield results of approximately (-110 MeV)$^3$. 
547: These flaws undoubtedly point to inadequacies
548: in the quark vacuum Ansatz. Of course, since we are working with the full QCD
549: Hamiltonian, it is possible to improve the Ansatz (for a coupled cluster approach,
550: see Ref. \cite{SK}; for one loop corrections see Ref. \cite{ss4}). Since one of our chief interests is the implementation of a
551: formalism which respects chiral symmetry, and not detailed numerics, we satisfy
552: ourselves with the present procedure.
553: 
554: 
555: 
556: 
557: 
558: 
559: 
560: 
561: %\begin{figure}
562: %\includegraphics[scale=0.35,angle=-90]{mu2.ps}
563: %\caption{ \label{mu} Constituent Quark Mass. The points are the numerical
564: %solution to the gap equation. The curve is Eq. \ref{muEq}.}
565: %\end{figure}
566: 
567: 
568: 
569: \section{Mesons}
570: 
571: 
572: With explicit expressions for the quasigluon interaction (and hence the constituent
573: quark interaction) and the dynamical quark mass in hand we are ready to obtain mesonic
574:  bound
575: states. As mentioned above, in order to obtain a massless pion one must construct states, $|M\rangle$, in the random phase approximation:
576: 
577: \begin{equation}
578: \langle M| [H,Q^{\dag}_M] | RPA \rangle = (E_M - E_{BCS} ) \langle M| Q^{\dag} |RPA \rangle, 
579: \label{RPA}
580: \end{equation} 
581: where $Q^{\dag}_M$ is defined in terms of the positive and
582: negative energy wavefunctions, $Q^{\dag}_M = \sum_{\alpha\beta}
583: \left[ \psi^+_{\alpha\beta} B^{\dag}_{\alpha} D^{\dag}_{\beta} - \psi^-_{\alpha\beta}
584: D_{\beta} B_{\alpha} \right]$ with $B$ and $D$ being the 
585: quasiparticle operators. It is worthwhile recalling that the RPA  method is
586: equivalent to the Bethe-Salpeter approach with instantaneous interactions\cite{RS}.
587: 
588: 
589: 
590: The RPA and TDA equations includes self energy terms (denoted $\Sigma$)  for each quark line and these must be
591: renormalized.
592: In the zero quark mass case renormalization of the TDA or RPA equations
593: proceeds in the same way as for the quark gap equation. In fact, the
594: renormalization of these equations is consistent and one may show that a 
595: finite gap equation implies a finite RDA or TDA equation. This feature
596: remains true in the massive case. 
597: 
598: The RPA equation in the pion channel reads:
599: 
600: \begin{eqnarray}
601: (E_\pi-E_{BCS}) \psi^+(k) &=& 2\left[m s_k + k c_k + \Sigma(k)\right] \psi^+(k) \nonumber \\
602: && - {C_F\over 2} \int {p^2dp\over (2\pi)^3}\left[ V_0(k,p)(1+s_ks_p) + V_1(k,p)c_kc_p\right] \psi^+(p) \nonumber \\
603: && - {C_F\over 2} \int {p^2dp\over (2\pi)^3}\left[ V_0(k,p)(1-s_ks_p) - V_1(k,p)c_kc_p\right] \psi^-(p)
604: \end{eqnarray}
605: A similar equation for $\psi^-$ holds with $(+ \leftrightarrow -)$ and $E \to -E$. The
606: wavefunctions $\psi^\pm$ represent forward and backward moving components of the
607: many-body wavefunction and the pion itself is a collective excitation with infinitely
608: many constituent quarks in the Fock space expansion.
609: 
610: We also consider a simpler truncation of QCD called the Tamm-Dancoff approximation.
611: This may be obtained from the RPA equation by neglecting the backward wavefunction
612: $\psi^-$.
613: 
614: We have computed the spectrum in the random phase and Tamm-Dancoff approximations
615: and confirm that the pion is massless in the chiral limit.  We also find that 
616: the Tamm-Dancoff approximation yields results very close to the RPA for all states
617: except the pion: the TDA pion mass is 580 MeV while the first excited pion 
618: has an RPA mass of 1410 MeV  and a TDA mass of 1450 MeV. All other mesons have nearly
619: identical RPA and TDA masses.
620: For this reason we simply present  TDA equations and results below.
621: 
622: 
623: The complete hidden flavor meson spectrum in the Tamm-Dancoff approximation is
624: given by the following equations.
625: 
626: 
627: \begin{equation}
628: E \psi_{PC}(k) = 2 \left[m s_k + kc_k + \Sigma(k)\right]\psi_{PC}(k) - {C_F\over 2} \int {p^2dp\over (2\pi)^3} K^{PC}_J(k,p) \psi_{PC}(p)
629: \end{equation}
630: with
631: 
632: \begin{equation}
633: \Sigma(k) = {C_F\over 2} \int {p^2 dp\over (2\pi)^3} \left( V_0 s_k  s_p + V_1 c_k c_p \right)
634: \end{equation}
635: and where $\psi$ is the meson radial wavefunction in momentum space.
636: 
637: An alternate form for the kinetic and self energies which is closer the Schr\"odinger
638: equation may be obtained by substituting the gap equation to obtain:
639: 
640: \begin{equation}
641: \mbox{ kinetic + self energy} = 2\left[ E(k) + \Gamma(k)\right]
642: \end{equation}
643: where
644: \begin{equation}
645: \Gamma(k) = {C_F \over 2} \int {p^2 dp \over (2\pi)^3} V_1 {c_p\over c_k}.
646: \end{equation}
647: and
648: \begin{equation}
649: E(k) = \sqrt{k^2+\mu(k)^2}.
650: \end{equation}
651: 
652: The kernel, $K_J$ in the potential term depends on the meson quantum numbers, $J^{PC}$.
653: In the following possible values for the parity or charge conjugation eigenvalues are
654: denoted by $(J) = +$ if $J$ is even and $-$ if $J$ is odd.
655: 
656: \noindent
657: $\bullet\ $ $0^{++}$
658: 
659: \begin{equation}
660: K(p,k) = V_0 c_p c_k + V_1 (1 + s_p s_k) 
661: \end{equation}
662: 
663: \noindent
664: $\bullet\ $ $J^{(J+1)(J)}$ [$^1J_J$ , $J \geq 0$]
665: 
666: \begin{equation}
667: K_J(p,k) = V_J (1 + s_p s_k) + \left( V_{J-1} {J\over 2J+1} + 
668: V_{J+1} {J+1 \over 2J+1} \right) c_p c_k
669: \label{1jj}
670: \end{equation}
671: 
672: 
673: \noindent
674: $\bullet\ $ $J^{(J+1)(J+1)}$ [$^3J_J$ , $J \geq 1$]
675: 
676: \begin{equation}
677: K_J(p,k) = V_J (1 +s_p s_k) + \left( V_{J-1} {J+1\over 2J+1} 
678: + V_{J+1} {J \over 2J +1} \right) c_p c_k
679: \end{equation}
680: 
681: 
682: \noindent
683: $\bullet\ $ $J^{(J)(J)}$ [$^3(J-1)_J, ^3(J+1)_J$ , $J \geq 1$]
684: 
685: 
686: \begin{eqnarray}
687: K_{11}(p,k) &=& V_J c_p c_k + \left( V_{J-1} {J \over 2J+1} + V_{J+1} 
688: {J+1\over 2J+1} \right) (1 + s_p s_k) \cr
689: K_{22}(p,k) &=& V_J c_p c_k + \left( V_{J-1} {J+1 \over 2J+1} + V_{J+1} 
690: {J\over 2J+1} \right) (1 + s_p s_k) \cr
691: K_{12}(p,k) &=& \left( V_{J-1} - V_{J+1} \right)
692: {\sqrt{J(J+1)}\over 2J+1} \left( s_k + s_p \right)
693: \end{eqnarray}
694: 
695: These interaction kernels have been derived in the quark helicity basis.
696: We remark that in the LS basis the off-diagonal (J-1):(J+1) interaction 
697: for $J^{(J)(J)}$ mesons 
698: is proportional to $1 + s_k s_p - s_k - s_p$ and hence goes to zero in
699: the heavy quark mass limit as expected. Finally, we note that the authors
700: of Ref. \cite{LC} find that the $1^{++}$ and $1^{+-}$ kernels are identical.
701: The likely reason is an error in their $1^{+-}$ kernel which disagrees with 
702: that of Eq. \ref{1jj}.
703: 
704: 
705: 
706: 
707: \subsection{Short Range Behavior}
708: 
709: Our quark interaction is instantaneous, central, and obeys Casimir scaling.
710: However, it is not flavor or spin independent because the full spinor structure of 
711: the interaction has been retained. This spinor structure is specified by the
712: Hamiltonian of QCD and is of a vector nature (specifically $\bar \psi \gamma_0 \psi
713: \otimes \bar \psi \gamma_0 \psi$). 
714: A nonrelativistic reduction of this interaction
715: yields no spin-spin hyperfine or tensor interactions. Thus the present computation
716: cannot correctly
717: describe well known spin splittings such as $J/\psi - \eta_c$ or $\Delta-N$
718: (we describe the extensions necessary to do so below).
719: 
720: 
721: A spin-orbit interaction is present and is given by 
722: 
723: \begin{equation}
724: V_{SO} = {1 \over 2 m^2 r} {d \over dr}\left( -C_F K^{(0)}(r)\right) \vec L \cdot \vec S
725: \end{equation}
726: in the equal quark mass case.
727: 
728: The spin-orbit interaction is famous for its problematic nature in the
729: constituent quark model. Quark model lore states that the spin-orbit interaction
730: generated by one gluon exchange is too strong for phenomenology and must be 
731: softened by the addition of a spin orbit interaction from the confinement
732: term which has an opposite sign. This can only be arranged if confinement
733: has a scalar Dirac structure ($\bar \psi \psi \otimes \bar \psi \psi$). 
734: This is clearly at odds with the formalism presented here which insists that
735: the confinement and Coulomb potentials (here we mean the perturbative $1/r$ tail of the 
736: static potential)  share the same Dirac structure. The
737: resolution to the conundrum is  that it is too simple-minded to ascribe all
738: of spin-orbit interactions to the Dirac structure of instantaneous potentials.
739: Indeed, spin-dependence in QCD is partly generated by nonperturbative
740: mixing with intermediate states, and need not follow the dictates of quark
741: model lore. A specific  realization of this is given in Ref. \cite{ss3}.
742: Finally, we note that a scalar confinement interaction
743: leads to inconsistencies between mesons and baryons: if mesons confine then
744: baryons anticonfine --- clearly an unacceptable situation!
745: 
746: Relativistic interactions generate short range spin-dependent interactions
747: by virtue of their spinor structure. The other sources of spin-dependence are
748: topological effects (for example, an instanton induced interaction) and Fock
749: sector mixing effects. Fock sector mixing is easily incorporated into the
750: current formalism, one need only increase the size of the Fock space being
751: considered (there is one subtlety: the mixing between Fock sectors must be
752: treated carefully since it can involve nonperturbative gluodynamics). One expects
753: the leading higher Fock sectors to be meson-meson (ie., meson loop corrections to 
754: the spectrum) and hybrid. The latter case is the nonperturbative analogue of
755: one-gluon-exchange, and as mentioned above, is phenomenologically important
756: in the light meson spectrum. 
757: 
758: We shall leave the topic of Fock space mixing for a future investigation and
759: press ahead with an examination of the spectrum which arises from the central
760: static potential generated by the non-Abelian Coulomb interaction, keeping in mind
761: that large spin-dependent mass shifts may occur in the light spectrum.
762: 
763: 
764: \section{Results and Discussion}
765: 
766: 
767: Once the gluonic sector of the formalism has been fixed by renormalization 
768: the only remaining parameters are quark masses. In the following we have determined
769: these by fitting the $\Upsilon$, $J/\psi$, and $\phi$ masses. We work in
770: the chiral limit for light ($u$ and $d$ quark) mesons so predictions in this
771: sector are completely fixed.  As a result there is no possibility of adjusting
772: the spectrum presented below. Thus it is possible to test the assumptions of
773: the model throughout the spectrum.
774: 
775: 
776: 
777: 
778: 
779: \subsection{Quarkonia}
780: 
781: 
782: A simple way to obtain the upsilon spectrum from QCD is to insert the
783: lattice Wilson potential into the nonrelativistic Schr\"odinger equation.
784: One argues that the heavy $b$ quark mass validates the Born-Oppenheimer
785: approximation (so that the static potential may be used) and the use
786: of a  nonrelativistic framework. Precisely this approach has been taken by 
787: Juge, Kuti, and Morningstar\cite{JKM} (JKM)
788: and was subsequently justified by comparison 
789: with nonrelativistic lattice computations\cite{JKM2}.   
790: 
791: 
792: Since the 
793: potential we employ is essentially equivalent to that obtained from the
794: Wilson loop 
795: one may expect that the predicted upsilon spectrum will agree very closely 
796: with that of JKM. This expectation relies on two 
797: things: (i) the heavy quark mass must eliminate non-central contributions 
798: in the interaction kernel, (ii) the self energy contribution must be essentially
799: independent of momentum. (The authors of Ref. \cite{JKM} did not
800: include a self energy term in their model because it cannot be extracted from lattice
801: data. Such a term must exist as external fermion propagator  corrections.)  
802: Explicit computations show that these expectations are indeed borne out and indicate
803: that a finite renormalization is sufficient to largely eliminate the self energy term.
804: 
805: Our predicted upsilon spectrum is given in Table I.  
806: A comparison with experiment
807: shows that the radial excitations rise too slowly with radial quantum number, and
808: indeed our results similarly disagree with those of JKM. But, as noted above, the
809: form for the static potential of Eq. \ref{Vss7} is not strongly constrained at
810: large momenta. We have therefore computed the upsilon spectrum with the modified
811: potential of Eq. \ref{Vmod}. The results are shown in Table II and are in much
812: better agreement with JKM and experiment. As expected, the detailed form of the
813: Coulomb tail of the potential is very important for low-lying heavy quark states. 
814: Our results are very similar to typical quark model spectra\cite{GI}; with some deviation
815: (at the percent level) becoming visible higher in the spectrum. This is to 
816: be expected because we employ a lattice potential with a string tension of 0.2-0.25 
817: GeV$^2$ whereas the quark model of Ref. \cite{GI} takes $b = 0.18$ GeV$^2$. 
818: 
819: Overall, the agreement with the experimental upsilon spectrum is impressive considering
820: the simplicity of the model and that the potential was not fit to data. It is possible
821: that deviations are seen higher in the spectrum, and indeed, one may expect this
822: since the open flavor threshold is at 10.56 GeV -- between the third and fourth
823: vector S-wave states. In general mixing with higher Fock components, such as hybrids (one 
824: gluon exchange in perturbative language) and meson-meson channels, will occur. These
825: effects should become more important as one probes higher in the spectrum. 
826: Eventually the quark + potential picture of mesons should break down entirely as
827: an increasing number of gluonic degrees of freedom are excited. 
828: 
829: 
830: The psi spectrum predicted with the canonical potential of Eq. \ref{Vss7} is presented
831: in Table III. The agreement with experiment is on the percent level. This is something
832: of a surprise when considering that the upsilon spectrum required a careful fit to
833: the high momentum potential. The psi spectrum computed from the modified
834: potential (Table IV) compares unfavorably to the data, with deviations at
835: the five percent level. It appears that the linear and Coulomb portions of the
836: potential are equally important to the low-lying psi spectrum and that the small
837: value of the effective strong coupling has largely cancelled against the large
838: value of the string tension. Indeed, as mentioned above, quark models typically employ a much smaller
839: string tension than that of the Wilson loop. We thus have evidence that the
840: effective string tension is reduced for lighter quark masses. This can easily
841: be induced by mixing with virtual meson pairs or by motion of the sources in the
842: Wilson loop. Again, it should be possible to incorporate the physics of string
843: softening in the model through Fock sector mixing.
844: 
845: 
846: 
847: The results of Tables III and IV also indicate that spin splittings are becoming
848: important. For example the $\eta_c$ is roughly 100 MeV lighter than the $J/\psi$
849: and this is not predicted in the model. Again, this fault is easy to remedy once
850: higher Fock components are admitted.  One also sees evidence for tensor splitting in
851: the $J^{++}$ states which are quite well reproduced with the modified potential (Table IV). One concludes that the Dirac structure of the interaction is becoming important
852: and that it does a reasonable job at charm mass scales but that Fock mixing must be
853: accounted for, even for low-lying states.  As indicated in the Introduction, if
854: canonical quark model lore holds, the addition of virtual hybrid states to the
855: model will ruin the predicted tensor splittings since these will increase them
856: an unacceptable amount. We stress, however, that nonperturbative mixing with virtual
857: hybrid states is not equivalent to perturbative gluon exchange which 
858: only becomes relevant in Hamiltonian QCD very high in the spectrum.
859: 
860: 
861: 
862: 
863: 
864: \subsection{Isovector  Mesons}
865: 
866: We have seen that the formalism presented here is quite accurate for
867: bottom quarks and reasonably accurate (with the possibility of being very accurate
868: once simple Fock sector mixing is included) for charm quarks. 
869: The challenge in carrying this success to the light quark sector lies in 
870: (i) the assumption that the static Wilson loop potential is relevant to light
871: quarks, (ii) the importance of chiral symmetry breaking, (iii) large spin-dependences which may be present, (iv) the possibility that topological aspects of QCD become
872: important. One of the reasons Coulomb gauge QCD is useful for constructing models
873: of hadronic physics is that these issues may be addressed in a systematic fashion.
874: The latter three issues have already been discussed; for the first we note that
875: an instantaneous interaction between quarks exists for all quark masses in Coulomb gauge.
876: The viability of this interaction can only be affected by higher order gluonic terms
877: (such as in the operator $K-K^{(0)}$); but we have seen that these contributions  are 
878: suppressed by an energy denominator of the order of 1 GeV. Thus a static interaction
879: should provide a good approximation low in the light spectrum. Virtual light quark
880: loops are another source of nonpotential interactions. But chiral symmetry
881: breaking implies that the light current quarks acquire an effective mass, and this
882: mass assists in dampening such loop effects. The success of the constituent quark
883: model also indicates that loop effects
884: on the interaction can be largely subsumed by renormalization.
885: 
886: 
887: 
888: Our results for the light meson spectrum are presented in Figure 1 and in Table V.
889: The first feature to notice is that the pion is massless as desired. Of course
890: a finite pion mass can be obtained if a finite current quark mass is used in the
891: calculation. We have chosen to keep the current quark mass at zero in order to 
892: test the robustness of the model in a zero parameter computation.
893: 
894: The rho meson mass is predicted to be 772 MeV, in very good agreement with data. 
895: We regard this as somewhat fortuitous since the potential has been fixed by lattice
896: data and no parameter tuning has taken place. The first radial excitation is at 1390
897: MeV, whereas the experimental mass is approximately 1450 MeV. However, we note that
898:  already the possibility
899: of Fock mixing arises since the lowest mass vector hybrid is expected around 1900 MeV.
900: 
901: As seen in the psi spectrum, the tensor $J^{++}$ multiplet is sensitive to short
902: range effects. In this case the $1^{++}$ state is in rough agreement with experiment.
903: However the tensor state lies much above the data. The isoscalar scalar
904: has a mass of 850 MeV. One may be tempted to ascribe this state to the
905: $f_0(980)$, which is famous for being too light in constituent quark models. However,
906: the fact that the tensor state is much too massive is an indication that additional
907: strong spin-orbit forces are required to obtain a satisfactory description of light
908: mesons and that any conclusions concerning the nature of the scalar meson would
909: be premature.
910: 
911: The figure and table also present meson masses for high angular momentum. It is
912: seen that the model severely overestimates  the masses of these states. The likely
913: cause for this is the large (compared to quark model) string tension which we
914: have used.  However, as argued above, there is little freedom in choosing the
915: string tension and one must ascribe the discrepancy to nonpotential effects.
916: The simplest such effects are extra quasigluons in the Fock space expansion of
917: higher lying states. Thus we regard the poor quality of the predictions at high
918: angular momentum as an indication of the range of validity of the potential
919: approach to hadronic physics (of course it is possible to extend this range
920: by allowing more parameter freedom, but the degrees of freedom will be incorrect,
921: and detailed predictions of such models must fail).
922: 
923: 
924: Equations (27) and (28)  make it clear that meson masses form charge conjugation
925: doublets in the pattern $1^{+\pm}$, $2^{-\pm}$, {\it etc.} in the large angular momentum
926: limit.  This pattern is clearly seen in the figure. Such behavior is expected in the
927: heavy quark limit where quark spins decouple and one may expect it to also occur
928: for light quarks in the high angular momentum regime since the quark spin effects
929: are local in the interaction.
930: 
931: 
932: A central feature of chiral symmetry breaking is that quark spinors reduce to massless
933: spinors at momenta much larger than the chiral symmetry breaking scale. Thus one
934: has $c_k \to 1$ and $s_k \to 0$ for $k >> \Lambda_{\chi SB}$. Since the average 
935: quark momentum becomes large as the angular momentum increases we conclude that the
936: interaction kernel approaches 
937: 
938: \begin{equation}
939: K_J \to V_J + {1\over 2}V_{J-1} + {1\over 2} V_{J+1}
940: \end{equation}
941: for all possible $J^{PC}$ (the off-diagonal potential in the $J^{(J)(J)}$ sector
942: goes to zero). Thus the entire spectrum becomes degenerate in the large angular momentum
943: limit; in particular parity doubling occurs, as expected\cite{CG}. 
944: 
945: \begin{figure}
946: \includegraphics[width=14cm]{lightSpec.eps}
947: \caption{ 
948: \label{uu} 
949: The isovector meson spectrum.  Experimental masses\cite{PDG} 
950: are represented by lines to the left of each column. The attached boxes indicate
951: the widths of each state. Lines to the right are the results of this computation
952: in the massless quark limit. Narrower boxes are new results from the Crystal Barrel
953: collaboration\cite{CB}.
954: }
955: \end{figure}
956: 
957: 
958: 
959: \section{Conclusions}
960: 
961: 
962: It is becoming clear that simple constituent quark models are limited in their
963: ability to describe excitations high in the hadronic spectrum. Well known 
964: flaws like nonrelativistic kinematics are exacerbated by the limitations of
965: a fixed particle number formulation. Indeed one expects Fock sector mixing to become
966: increasingly important as experiment probes high in the spectrum. For example 
967: meson loop effects can cause significant mass shifts and commensurate 
968: wavefunction distortion will affect other matrix elements. Furthermore, gluons 
969: must manifest themselves about 1 GeV above the ground state in a given sector and
970: the subsequent state mixing can be important.
971: 
972: The model constructed here is an attempt at going beyond the nonrelativistic quark model.
973: Since it is based on a truncation of QCD in Coulomb gauge, it is heavily constrained, 
974: systematically improvable, and relativistic. Because the instantaneous interaction is 
975: fixed by the lattice gauge Wilson loop potential the model can also fail.
976: Any such failure may be regarded as an indication of the limitations of the potential
977: approach to hadronic physics. Furthermore, the Dirac structure of the interaction is
978: fixed by the Coulomb gauge Hamiltonian of QCD.  This has important phenomenological 
979: implications; for example the spin-orbit interaction is fixed to have the same sign
980: as the Coulomb interaction of the constituent quark model. A happy consequence is that
981: mesons and baryons are treated on equal footing (scalar confinement requires an 
982: additional sign change of the interaction between the sectors).  Finally, the assumption
983: that the operator $K$ generates the leading quark and gluon interaction can be tested
984: on the lattice. Initial results are mixed, with Ref. \cite{ZC} supporting the conjecture
985: while Ref. \cite{G}  finds that the Coulomb string tension is roughly three times larger
986: than the Wilson loop string tension. We find the latter result improbable since the 
987: coupling of gluons with static quarks is suppressed, leaving only the non-Abelian
988: Coulomb interaction to mediate the Wilson loop interaction. Nevertheless, if the 
989: latter result is confirmed the present method will likely have to be abandoned.
990: 
991: The issue of Fock sector mixing will be vital to the success of this program. Such
992: effects are clearly needed in the light quark spectrum and are of some significance
993: for heavy quarks. The most important such effect is the nonperturbative analogue of 
994: one gluon exchange, namely mixing with virtual hybrid states. Implementing this will
995: be a crucial test of the model since spin orbit splittings depend sensitively in the
996: Dirac structure of the mixing terms and of the central potential. It is entirely possible
997: that the formalism proposed here will fail and that some sort of string approach will
998: be required but this remains to be seen.
999: 
1000: 
1001: We have implemented chiral symmetry breaking using
1002: standard many-body or Nambu--Jona-Lasinio methods. This is of course a truncation of
1003: all diagrams which contribute to the bound state Bethe-Salpeter equation, however, it
1004: is convenient and is enough to  demonstrate the chiral nature of the pion and the
1005: importance of chiral symmetry high in the spectrum. It is possible to improve the
1006: computation in a systematic fashion. Analogous improvements in the Dyson-Schwinger
1007: Bethe-Salpeter approach are discussed in Ref. \cite{RR}.
1008: 
1009: 
1010: In future we intend to examine the open flavor spectrum, strong decays, short range
1011: structure and spin splittings, and Fock sector mixing effects.  Further research into
1012: topological aspects of the model and the isoscalar sector will also be of great interest.
1013: 
1014: 
1015: 
1016: \begin{acknowledgments}
1017: 
1018: This work was supported by the DOE under contracts DE-FG02-00ER41135  and  DE-AC05-84ER40150.
1019: \end{acknowledgments}
1020: 
1021: \appendix
1022: 
1023: 
1024: \begin{table}
1025: \caption{$\Upsilon$ Spectrum (MeV)}
1026: \begin{tabular}{rrrrrrrrrrrrrrrrr}
1027: \hline
1028:  $0^{-+}$& $0^{++}$& $1^{--}$& $1^{+-}$& $1^{++}$&
1029:  $2^{--}$& $2^{-+}$& $2^{++}$& $3^{--}$& $3^{+-}$&
1030:  $3^{++}$& $4^{--}$& $4^{-+}$& $4^{++}$& $5^{--}$&
1031:  $5^{+-}$& $5^{++}$\\
1032: \hline
1033:  9460 &  9723 &  9460 &  9731 &  9727 &  9946 &  9948 &  9735 &  9954 &
1034: 10141
1035:   &  10139 &  10318 &  10319 &  10147 &  10326 &  10487 &  10486\\
1036:  9878 &  10070 &  9878 &  10076 &  10073 &  10254 &  10256 &  10079 &
1037: 10261 &
1038:  10426 &  10424 &  10586 &  10587 &  10431 &  10593 &  10744 &  10743\\
1039:  10205 &  10369 &  9941 &  10375 &  10372 &  10536 &  10538 &  10133 &
1040: 10311
1041:   &  10696 &  10694 &  10850 &  10851 &  10478 &  10639 &  11003 &
1042: 11002\\
1043:  10494 &  10646 &  10205 &  10651 &  10649 &  10804 &  10806 &  10378 &
1044: 10541
1045:   &  10957 &  10955 &  11103 &  11104 &  10701 &  10856 &  11247 &
1046: 11246\\
1047:  10761 &  10901 &  10250 &  10908 &  10904 &  11050 &  11052 &  10419 &
1048: 10580
1049:   &  11194 &  11192 &  11336 &  11337 &  10736 &  10889 &  11483 &
1050: 11482 \\
1051: \hline
1052: \end{tabular}
1053: \end{table}
1054: 
1055: 
1056: \begin{table}
1057: \caption{$\Upsilon$ Spectrum. Modified Potential, $a=1.0$, $b=0.8$ Experiment in brackets. (GeV)}
1058: \begin{tabular}{lllll}
1059: \hline
1060: $0^{-+}$ & $0^{++}$ & $1^{--}$\footnote{S-wave only.} & $1^{++}$ & $2^{++}$ \\
1061: \hline
1062: 9.46  & 9.80 (9.86) & 9.46 (fit)     & 9.81 (9.89)   & 9.83 (9.91) \\
1063: 9.97 &  10.20 (10.23) & 9.98 (10.02) & 10.21 (10.25) & 10.22 (10.27) \\
1064: 10.35 &  10.54        & 10.35 (10.35) & 10.54        & 10.29 \\
1065: 10.67 & 10.83         & 10.67 (10.58)   & 10.84      & 10.55 \\
1066: 10.96 & 11.10         & 10.95 (10.86?) & 11.10       & 10.60 \\
1067: 11.22 & 11.35         & 11.21 (11.02?) & 11.35       & 10.85 \\
1068: \hline
1069: \end{tabular}
1070: \end{table}
1071: 
1072: \begin{table}
1073: \caption{$\psi$ Spectrum (MeV)}
1074: \begin{tabular}{rrrrrrrrrrrrrrrrr}
1075: \hline
1076:  $0^{-+}$& $0^{++}$& $1^{--}$& $1^{+-}$& $1^{++}$&
1077:  $2^{--}$& $2^{-+}$& $2^{++}$& $3^{--}$& $3^{+-}$&
1078:  $3^{++}$& $4^{--}$& $4^{-+}$& $4^{++}$& $5^{--}$&
1079:  $5^{+-}$& $5^{++}$\\
1080: \hline
1081:  3061 &  3376 &  3063 &  3424 &  3400 &  3720 &  3736 &  3450 &  3772 &
1082: 4016
1083:   &  4003 &  4264 &  4274 &  4058 &  4321 &  4517 &  4508\\
1084:  3639 &  3878 &  3642 &  3911 &  3895 &  4154 &  4166 &  3932 &  4196 &
1085: 4406
1086:   &  4396 &  4625 &  4633 &  4441 &  4673 &  4851 &  4843\\
1087:  4093 &  4294 &  3684 &  4320 &  4307 &  4531 &  4541 &  3962 &  4218 &
1088: 4753
1089:   &  4745 &  4951 &  4958 &  4458 &  4687 &  5157 &  5150\\
1090:  4480 &  4658 &  4096 &  4679 &  4668 &  4868 &  4876 &  4337 &  4566 &
1091: 5068
1092:   &  5061 &  5250 &  5256 &  4784 &  4993 &  5440 &  5434\\
1093:  4823 &  4985 &  4126 &  5003 &  4994 &  5175 &  5183 &  4362 &  4586 &
1094: 5360
1095:   &  5354 &  5530 &  5536 &  4800 &  5007 &  5710 &  5705\\
1096: \hline
1097: \end{tabular}
1098: \end{table}
1099: 
1100: \begin{table}
1101: \caption{$\psi$ Spectrum. Modified Potential, $a=1.0$, $b=0.8$ Experiment in brackets. (GeV)}
1102: \begin{tabular}{lllll}
1103: \hline
1104: $0^{-+}$ & $0^{++}$ & $1^{--}$ & $1^{++}$ & $2^{++}$ \\
1105: \hline
1106: 3.09 (2.98)  & 3.45 (3.41)&  3.095 (fit) & 3.48 (3.51) & 3.57 (3.55) \\
1107: 3.76 (3.65) & 4.03        & 3.76 (3.686) & 4.05        & 4.11 \\
1108: 4.28        & 4.51       & 3.81 (3.77) & 4.53          & 4.12\\
1109: 4.73        & 4.93       & 4.29 (4.04) & 4.94          & 4.57 \\
1110: 5.12        & 5.30       & 4.32 (4.16) & 5.31          & 4.58 \\
1111: \hline
1112: \end{tabular}
1113: \end{table}
1114: 
1115: 
1116: 
1117: \begin{table}
1118: \caption{Isovector Spectrum (MeV)}
1119: \begin{tabular}{rrrrrrrrrrrrrrrrrrrr}
1120: \hline
1121:  $0^{-+}$& $0^{++}$& $1^{--}$& $1^{+-}$& $1^{++}$&
1122:  $2^{--}$& $2^{-+}$& $2^{++}$& $3^{--}$& $3^{+-}$&
1123:  $3^{++}$& $4^{--}$& $4^{-+}$& $4^{++}$& $5^{--}$&
1124:  $5^{+-}$& $5^{++}$& $6^{--}$& $6^{-+}$& $6^{++}$ \\
1125: \hline
1126:  0 &  842 &  772 &  1317 &  1088 &  1796 &  1902 &  1731 &  2284 &
1127: 2365 &
1128:  2303 &  2713 &  2755 &  2706 &  3065 &  3098 &  3068 &  3387 &  3410 &
1129: 3386\\
1130:  1454 &  1660 &  1389 &  1958 &  1811 &  2331 &  2415 &  1931 &  2377 &
1131: 2808
1132:   &  2755 &  3115 &  3151 &  2760 &  3101 &  3461 &  3434 &  3725 &
1133: 3746 &
1134:  3412\\
1135:  2102 &  2278 &  1582 &  2490 &  2384 &  2797 &  2865 &  2247 &  2723 &
1136: 3206
1137:   &  3161 &  3481 &  3514 &  3101 &  3427 &  3796 &  3771 &  4040 &
1138: 4059 &
1139:  3721\\
1140:  2629 &  2785 &  2042 &  2948 &  2866 &  3210 &  3268 &  2457 &  2828 &
1141: 3570
1142:   &  3529 &  3819 &  3849 &  3162 &  3467 &  4108 &  4086 &  4335 &
1143: 4353 &
1144:  3749\\
1145:  3079 &  3221 &  2190 &  3354 &  3287 &  3583 &  3633 &  2707 &  3121 &
1146: 3904
1147:   &  3868 &  4133 &  4160 &  3462 &  3761 &  4401 &  4380 &  4613 &
1148: 4630 &
1149:  4034 \\
1150: \hline
1151: \end{tabular}
1152: \end{table}
1153: 
1154: %\bibliography{/home/swanson/tex/eric,/home/swanson/tex/others}
1155: % bibtex/latex are not smart enough to merge these ...
1156: 
1157: 
1158: \begin{thebibliography}{99}
1159: 
1160: \bibitem{RS}
1161: J.~Resag and D.~Sch{\"u}tte,
1162: ``RPA Equations and the Instantaneous Bethe-Salpeter Equation'',
1163: arXiv:nucl-th/9312013.
1164: %%CITATION = NUCL-TH 9312013;%%
1165: 
1166: %\cite{LeYaouanc:1984dr}
1167: \bibitem{LeYaouanc:1984dr}
1168: A.~Le Yaouanc, L.~Oliver, S.~Ono, O.~Pene and J.~C.~Raynal,
1169: %``A Quark Model Of Light Mesons With Dynamically Broken Chiral Symmetry,''
1170: Phys.\ Rev.\ D {\bf 31}, 137 (1985).
1171: %%CITATION = PHRVA,D31,137;%%
1172: 
1173: \bibitem{LC}
1174: F.~J.~Llanes-Estrada and S.~R.~Cotanch,
1175: %``Relativistic many-body Hamiltonian approach to mesons,''
1176: Nucl.\ Phys.\ A {\bf 697}, 303 (2002).
1177: %[arXiv:hep-ph/0101078].
1178: %%CITATION = HEP-PH 0101078;%%
1179: 
1180: \bibitem{ZK}
1181: T.~Zhang and R.~Koniuk,
1182: %``Heavy Quark Mesons In A Relativistic Quark Model,''
1183: Phys.\ Rev.\ D {\bf 43}, 1688 (1991).
1184: %%CITATION = PHRVA,D43,1688;%%
1185: 
1186: 
1187: 
1188: \bibitem{bonn}
1189: U.~Loring, B.~C.~Metsch and H.~R.~Petry,
1190: %``The light baryon spectrum in a relativistic quark model with  instanton-induced quark forces: The strange baryon spectrum,''
1191: Eur.\ Phys.\ J.\ A {\bf 10}, 447 (2001);
1192: %[arXiv:hep-ph/0103290].
1193: %%CITATION = HEP-PH 0103290;%%
1194: R.~Ricken, M.~Koll, D.~Merten, B.~C.~Metsch and H.~R.~Petry,
1195: %``The meson spectrum in a covariant quark model,''
1196: Eur.\ Phys.\ J.\ A {\bf 9}, 221 (2000). See also
1197: %[arXiv:hep-ph/0008221].
1198: %%CITATION = HEP-PH 0008221;%%
1199: D.~Ebert, R.~N.~Faustov and V.~O.~Galkin,
1200: %``Properties of heavy quarkonia and B/c mesons in the relativistic quark  model,''
1201: Phys.\ Rev.\ D {\bf 67}, 014027 (2003);
1202: %%CITATION = HEP-PH 0210381;%%
1203: L.~S.~Celenza, B.~Huang, H.~S.~Wang and C.~M.~Shakin,
1204: %``Covariant Confinement Model For The Study Of The Properties Of Light  Mesons,''
1205: Phys.\ Rev.\ C {\bf 60}, 025202 (1999).
1206: %%CITATION = PHRVA,C60,025202;%%
1207: 
1208: 
1209: \bibitem{MR}
1210: P.~Maris and C.~D.~Roberts,
1211: %``Dyson-Schwinger equations: A tool for hadron physics,''
1212: Int.\ J.\ Mod.\ Phys.\ E {\bf 12}, 297 (2003).
1213: %[arXiv:nucl-th/0301049].
1214: %%CITATION = NUCL-TH 0301049;%%
1215: 
1216: 
1217: \bibitem{ss8}
1218: A.~P.~Szczepaniak and E.~S.~Swanson,
1219: ``The low lying glueball spectrum'',
1220: arXiv:hep-ph/0308268.
1221: %%CITATION = HEP-PH 0308268;%%
1222: 
1223: 
1224: 
1225: \bibitem{CL}
1226: N.~H.~Christ and T.~D.~Lee,
1227: %``Operator Ordering And Feynman Rules In Gauge Theories,''
1228: Phys.\ Rev.\ D {\bf 22}, 939 (1980);
1229: T.~D.~Lee, {\sl Particle Physics And Introduction To Field Theory},
1230: (New York, Harwood Academic,  1981).
1231: %%CITATION = PHRVA,D22,939;%%
1232: 
1233: 
1234: \bibitem{schwinger}
1235: J.~Schwinger, Phys.\ Rev.\ {\bf 127}, 324 (1962).
1236: 
1237: 
1238: 
1239: 
1240: \bibitem{ss7}
1241: A.~P.~Szczepaniak and E.~S.~Swanson,
1242: %``Coulomb gauge QCD, confinement, and the constituent representation,''
1243: Phys.\ Rev.\ D {\bf 65}, 025012 (2002).
1244: %[arXiv:hep-ph/0107078].
1245: %%CITATION = HEP-PH 0107078;%%
1246: 
1247: \bibitem{Z}
1248: D.~Zwanziger,
1249: %``Renormalization in the Coulomb gauge and order parameter for  confinement in QCD,''
1250: Nucl.\ Phys.\ B {\bf 485}, 185 (1997).
1251: %%CITATION = NUPHA,B518,237;%%
1252: 
1253: 
1254: 
1255: 
1256: \bibitem{vanbaal}
1257: P. van Baal, ``Global Issues in Gauge Fixing'', arXiv:hep-th/9511119; ``The QCD Vacuum'', arXiv:hep-lat/9709066.
1258: 
1259: \bibitem{Z3}
1260: D.~Zwanziger,
1261: ``Non-perturbative Faddeev-Popov formula and infrared limit of QCD'',
1262: arXiv:hep-ph/0303028.
1263: %%CITATION = HEP-PH 0303028;%%
1264: 
1265: 
1266: \bibitem{gribov}
1267: V.N. Gribov, Nucl. Phys. {\bf B139}, 1 (1978).
1268: 
1269: 
1270: 
1271: \bibitem{Z1}
1272: D.~Zwanziger,
1273: %``No confinement without Coulomb confinement,''
1274: Phys.\ Rev.\ Lett.\  {\bf 90}, 102001 (2003).
1275: %[arXiv:hep-lat/0209105].
1276: %%CITATION = HEP-LAT 0209105;%%
1277: 
1278: \bibitem{ZC}
1279: A.~Cucchieri and D.~Zwanziger,
1280: %``Static color-Coulomb force,''
1281: Phys.\ Rev.\ Lett.\ {\bf 78}, 3814 (1997).
1282: %%CITATION = HEP-TH 9607224;%%
1283: 
1284: 
1285: \bibitem{G}
1286: J.~Greensite and S.~Olejnik,
1287: %``Coulomb energy, vortices, and confinement,''
1288: Phys.\ Rev.\ D {\bf 67}, 094503 (2003).
1289: %[arXiv:hep-lat/0302018].
1290: %%CITATION = HEP-LAT 0302018;%%
1291: 
1292: \bibitem{thouless}
1293: D. J. Thouless, Nucl. Phys. {\bf 22}, 78 (1961).
1294: 
1295: \bibitem{NJL}
1296: Y. Nambu and G. Jona-Lasinio, Phys. Rev. {\bf 122}, 345 (1961).
1297: 
1298: \bibitem{A}
1299: S.~L.~Adler and A.~C.~Davis,
1300: %``Chiral Symmetry Breaking In Coulomb Gauge QCD,''
1301: Nucl.\ Phys.\ B {\bf 244}, 469 (1984).
1302: %%CITATION = NUPHA,B244,469;%%
1303: 
1304: 
1305: \bibitem{FM}
1306: J.~R.~Finger and J.~E.~Mandula,
1307: %``Quark Pair Condensation And Chiral Symmetry Breaking In QCD,''
1308: Nucl.\ Phys.\ B {\bf 199}, 168 (1982).
1309: %%CITATION = NUPHA,B199,168;%%
1310: 
1311: 
1312: 
1313: \bibitem{Orsay}
1314: A.~Le Yaouanc, L.~Oliver, O.~P\`ene and J.~C.~Raynal,
1315: %``Spontaneous Breaking Of Chiral Symmetry For Confining Potentials,''
1316: Phys.\ Rev.\ D {\bf 29}, 1233 (1984).
1317: %%CITATION = PHRVA,D29,1233;%%
1318: 
1319: \bibitem{ss5}
1320: A.~P.~Szczepaniak and E.~S.~Swanson,
1321: %``Chiral extrapolation, renormalization, and the viability of the quark  model,''
1322: Phys.\ Rev.\ Lett.\  {\bf 87}, 072001 (2001).
1323: %[arXiv:hep-ph/0006306].
1324: %%CITATION = HEP-PH 0006306;%%
1325: 
1326: 
1327: \bibitem{P}
1328: This expression may be compared with the one loop perturbative version of
1329: D. Politzer, Nucl. Phys. {\bf B117}, 397 (1976).
1330: 
1331: \bibitem{SK}
1332: A.~P.~Szczepaniak and P.~Krupinski,
1333: %``Spontaneous chiral symmetry breaking in the linked cluster expansion,''
1334: Phys.\ Rev.\ D {\bf 66}, 096006 (2002).
1335: %[arXiv:hep-ph/0204249].
1336: %%CITATION = HEP-PH 0204249;%%
1337: 
1338: \bibitem{ss4}
1339: A.~P.~Szczepaniak and E.~S.~Swanson,
1340: %``From current to constituent quarks: A renormalization group improved  Hamiltonian-based description of hadrons,''
1341: Phys.\ Rev.\ D {\bf 55}, 1578 (1997).
1342: %%CITATION = HEP-PH 9609525;%%
1343: 
1344: \bibitem{ss3}
1345: A.~P.~Szczepaniak and E.~S.~Swanson,
1346: %``On the Dirac structure of confinement,''
1347: Phys.\ Rev.\ D {\bf 55}, 3987 (1997).
1348: %[arXiv:hep-ph/9611310].
1349: %%CITATION = HEP-PH 9611310;%%
1350: 
1351: 
1352: \bibitem{JKM}
1353: K.~J.~Juge, J.~Kuti and C.~J.~Morningstar,
1354: %``Gluon excitations of the static quark potential and the hybrid  quarkonium spectrum,''
1355: Nucl.\ Phys.\ Proc.\ Suppl.\  {\bf 63}, 326 (1998).
1356: %[arXiv:hep-lat/9709131].
1357: %%CITATION = HEP-LAT 9709131;%%
1358: 
1359: \bibitem{JKM2}
1360: K.~J.~Juge, J.~Kuti and C.~J.~Morningstar,
1361: %``The heavy hybrid spectrum from NRQCD and the Born-Oppenheimer  approximation,''
1362: Nucl.\ Phys.\ Proc.\ Suppl.\  {\bf 83}, 304 (2000).
1363: %[arXiv:hep-lat/9909165].
1364: %%CITATION = HEP-LAT 9909165;%%
1365: 
1366: \bibitem{GI}
1367: S.~Godfrey and N.~Isgur,
1368: %``Mesons In A Relativized Quark Model With Chromodynamics,''
1369: Phys.\ Rev.\ D {\bf 32}, 189 (1985).
1370: %%CITATION = PHRVA,D32,189;%%
1371: 
1372: \bibitem{PDG}
1373: ``Review of Particle Properties'', K. Hagiwara {\it et al.}, Phys. Rev. {\bf D66}, 1 (2002).
1374: 
1375: \bibitem{CB}
1376: A.~V.~Anisovich {\it et al.},
1377: % ``Combined Analysis Of Meson Channels With I = 1, C = -1 
1378: % From 1940-Mev To 2410-Mev,''
1379: Phys.\ Lett.\ B {\bf 542}, 8 (2002).
1380: %%CITATION = PHLTA,B542,8;%%
1381: 
1382: \bibitem{CG} 
1383: A general discussion of parity doubling is in E.S. Swanson, `` Parity Doubling in the Meson Spectrum'', arXiv:hep-ph/0309296.
1384: See also 
1385: T.~D.~Cohen and L.~Y.~Glozman,
1386: %``Chiral multiplets versus parity doublets in highly excited baryons,''
1387: Phys.\ Rev.\ D {\bf 65}, 016006 (2002)
1388: %[arXiv:hep-ph/0102206].
1389: %%CITATION = HEP-PH 0102206;%%
1390: and the comments in 
1391: A.~Le Yaouanc, L.~Oliver, S.~Ono, O.~P\`ene and J.~C.~Raynal,
1392: %``A Quark Model Of Light Mesons With Dynamically Broken Chiral Symmetry,''
1393: Phys.\ Rev.\ D {\bf 31}, 137 (1985), especially after Eq. 7.15.
1394: %%CITATION = PHRVA,D31,137;%%
1395: 
1396: \bibitem{RR}
1397: A.~Bender, C.~D.~Roberts and L.~Von Smekal,
1398: %``Goldstone Theorem and Diquark Confinement Beyond Rainbow-Ladder Approximation,''
1399: Phys.\ Lett.\ B {\bf 380}, 7 (1996); A. Bender, W. Detmold, C.D. Roberts,
1400:  and A.W. Thomas, Phys. Rev. C 65 (2002) 065203.
1401: %[arXiv:nucl-th/9602012].
1402: %%CITATION = NUCL-TH 9602012;%%
1403: 
1404: 
1405: \end{thebibliography}
1406: 
1407: \end{document}
1408: