hep-ph0009115/text
1: \documentstyle[sprocl]{article}
2: 
3: \font\eightrm=cmr8
4: 
5: \input{psfig}
6: 
7: 
8: \bibliographystyle{unsrt} %for BibTeX - sorted numerical labels by
9:                           %order of first citation.
10: 
11: \arraycolsep1.5pt
12: 
13: %\def\Journal#1#2#3#4{{#1} {\bf #2}, #3 (#4)}
14: \def\Journal#1#2#3#4{{#1} {\bf #2}, #4 (#3)}
15: 
16: % Some useful journal names
17: \def\NPB{{\em Nucl. Phys.} B}
18: \def\PLB{{\em Phys. Lett.}  B}
19: \def\PRL{\em Phys. Rev. Lett.}
20: \def\PRD{{\em Phys. Rev.} D}
21: 
22: 
23: 
24: \begin{document}
25: 
26: \title{GLUEBALLS, HYBRID AND EXOTIC MESONS AND STRING BREAKING}
27: 
28: \author{C. MICHAEL}
29: 
30: \address{Theoretical Physics Division, Dept. of Math. Sci.,\\
31:  University of Liverpool, Liverpool, L69 2BX, U.K. \\ E-mail:
32: c.michael@liv.ac.uk}
33: 
34: 
35: \maketitle\abstracts{ We review lattice QCD results for glueballs
36: (including  a discussion of mixing with scalar mesons), hybrid mesons
37: and exotic mesons (such  as $B_s B_s$ molecules). We also discuss
38: string breaking as a mixing between  colour flux states and $B \bar{B}$
39: states. }
40: 
41: \section{Introduction}
42: 
43:  
44: 
45: The most systematic approach to non-perturbative QCD is via lattice
46: techniques. 
47:   Lattice QCD needs as input the quark masses and an overall scale
48: (conventionally  given by $\Lambda_{QCD}$). Then any Green function can
49: be evaluated by taking an average of suitable combinations of the
50: lattice fields in the vacuum samples. This allows masses to be studied 
51: easily and matrix elements (particularly those of weak or
52: electromagnetic currents)  can be extracted straightforwardly.
53:   Unlike experiment, lattice QCD can vary the quark masses and can also 
54: explore different boundary conditions and sources. This allows a wide
55: range of  studies which can be used to diagnose the health of
56: phenomenological models as well as casting light on experimental data.
57: 
58: One limitation of the  lattice approach  to QCD is  in exploring
59: hadronic decays because the  lattice, using Euclidean
60: time, has no concept of asymptotic  states. One feasible strategy is to
61: evaluate the mixing between states of the same  energy - so giving some
62: information on on-shell hadronic decay amplitudes.
63: 
64:  For comparison with models and for ease of computation, the special 
65: case  of infinitely heavy sea quarks (namely neglect of quark effects in
66: the vacuum:  the  quenched approximation) is often used. We shall also 
67: present results from including sea quark effects - usually two flavours
68: of  degenerate sea quark of mass equivalent to strange quarks or
69: heavier.
70: 
71: 
72:  The quark model gives a good overall description of hadronic spectra.
73: Here I  will discuss lattice results for states which go beyond the
74: quark model:  glueballs, exotic mesons and hybrid mesons. I also discuss
75: the related phenomenon of string breaking.
76: 
77: 
78: \section{Glueballs}
79: 
80: Glueballs are defined to be hadronic states made primarily from gluons.
81: The full non-perturbative gluonic interaction is included in quenched
82: QCD.  In the quenched approximation, there is no mixing between such
83: glueballs  and quark - antiquark mesons. A study of the glueball
84: spectrum in quenched QCD  is thus of great value. This will allow
85: experimental searches to be  guided as well as providing calibration for
86: models of glueballs. A non-zero glueball mass in quenched QCD is the 
87: ``mass-gap'' of QCD. To prove this rigourously is one of the major
88: challenges  of our times. Here we will explore the situation using
89: computational techniques.
90: 
91: In principle, lattice QCD can study the meson spectrum as the sea quark
92: mass  is decreased towards experimental values. This will allow the
93: unambiguous glueball states in the quenched approximation to be tracked
94: as the sea quark effects are increased.  It may indeed turn out that no
95: meson in the physical spectrum is primarily a glueball - all states are 
96: mixtures of glue,  $q \bar{q}$, $q \bar{q} q \bar{q}$, etc. We shall
97: later discuss  lattice results on the mixing of glueballs and scalar
98: mesons (ie $q \bar{q}$ states). 
99: 
100: In lattice studies, dimensionless ratios of  quantities are obtained. To
101: explore the glueball masses, it is appropriate to combine  them with
102: another very accurately measured quantity to have a dimensionless 
103: observable. Since the potential between static quarks is very accurately
104: measured from the lattice, it is now conventional to use $r_0$ for this
105: comparison.  Here $r_0$ is implicitly defined by $r^2 dV(r)/dr = 1.65$
106: at $r=r_0$ where $V(r)$ is  the potential energy between static quarks
107: which is easy to determine accurately  on the lattice.  Conventionally 
108: $r_0 \approx 0.5$ fm.
109: 
110: 
111: 
112:  Theoretical analysis  indicates that for  Wilson's discretisation of
113: the gauge fields in the quenched approximation,  the dimensionless ratio
114: $mr_0$ will differ from the continuum  limit value by corrections of
115: order $a^2$.  Thus in fig.~1 the mass of the $J^{PC}$=$0^{++}$  glueball
116: is plotted versus the lattice spacing $a^2$. The straight line then
117: shows the continuum limit obtained  by extrapolating to $a=0$. As can be
118: seen, there is essentially no need for data  at even smaller $a$-values
119: to further fix the continuum value. The value shown  corresponds to
120: $m(0^{++})r_0=4.33(5)$.  Since several lattice
121: groups~\cite{DForc,MTgl,ukqcd,gf11} have measured these  quantities, it
122: is reassuring to see that the purely lattice observables are in 
123: excellent agreement. The publicised difference of quoted $m(0^{++})$
124: from  UKQCD~\cite{ukqcd} and GF11~\cite{gf11} comes entirely from
125: relating quenched lattice  measurements to values in GeV.
126: 
127: 
128: 
129: \begin{figure}[bt] 
130: \vspace{7cm} %
131: \special{psfile=gbr0.ps voffset=-15 hoffset=40 hscale=55 vscale=40}
132:  \caption{ The value of mass of the  $J^{PC}=0^{++}$  glueball state
133: from quenched data ($N_F=0$){\protect\cite{DForc,MTgl,ukqcd,gf11}}
134: in units of $r_0$ where $r_0 \approx 0.5$ fm. The straight line  shows a
135:  fit describing the  approach to the continuum limit as $a \to 0$.
136:  Results~{\protect\cite{sesam,bali,lat99}} with $N_F=2$ flavours of sea
137: quarks are also shown.
138:    }
139: \end{figure}
140:    
141: 
142: In the quenched approximation, different hadronic observables differ
143: from experiment  by factors of up to 10\%. Thus using one quantity or
144: another to set the scale, gives an overall systematic error.  Here I
145: choose to set the scale by taking the conventional value of the string
146: tension, $\sqrt{\sigma}=0.44$ GeV, which then corresponds to
147: $r_0^{-1}=373$ MeV. An overall systematic error of 10\% is then to be
148: included to any  extracted mass. This yields $m(0^{++})=1611(30)(160)$
149: MeV where the second error is the
150: systematic  scale error. Note that this is the  glueball mass in the
151: quenched approximation -  in the real world significant mixing with $q
152: \bar{q}$ states could modify this value substantially.
153: 
154: 
155: In the Wilson approach, the next lightest glueballs
156: are~\cite{MTgl,ukqcd} the tensor $m(2^{++})r_0=6.0(6)$  (resulting in  
157: $m(2^{++})=2232(220)(220)$ MeV) and the pseudoscalar $m(2^{++})r_0=
158: 6.0(1.0)$. Although the Wilson discretisation provides a definitive
159: study of the lightest ($0^{++}$)  glueball in the continuum limit, other
160: methods are competitive for the determination of the mass  of heavier
161: glueballs.  Namely, using an improved gauge discretisation which has 
162: even smaller discretisation errors than the $a^2$ dependence of the
163: Wilson discretisation,  so allowing a relatively coarse lattice spacing
164: $a$ to be used. To extract mass values, one has to explore the time
165: dependence of correlators and for this reason,  it is optimum to use a
166: relatively small time lattice spacing. Thus an asymmetric  lattice
167: spacing is most appropriate.  The  results~\cite{mpglue}  are shown in
168: fig.~2 and for low lying states are that $m(0^{++})r_0=4.21(11)(4)$, 
169: $m(2^{++})r_0=5.85(2)(6)$, $m(0^{-+})=6.33(7)(6)$ and
170: $m(1^{+-})r_0=7.18(4)(7)$.  Another recent study~\cite{nrw} has used an
171: improved discretisation based on the perfect action  approach (without a
172: space-time asymmetry) and obtains  results consistent with earlier work.
173: 
174: 
175: 
176: One signal of great interest would be  a glueball with $J^{PC}$ not
177: allowed for $q \bar{q}$ - a spin-exotic glueball or {\em oddball} -
178: since it would  not mix with $q \bar{q}$ states. These states are
179: found~\cite{MTgl,ukqcd,mpglue} to be  high lying: considerably above
180: $2m(0^{++})$. Thus they are  likely to be in a region very difficult to
181: access unambiguously by experiment. 
182: 
183: 
184: 
185:  Within the quenched approximation, the glueball states are unmixed 
186: with $q \bar{q},\ q \bar{q} q \bar{q}$, etc. Furthermore, the  $q
187: \bar{q}$ states have degenerate flavour singlet and non-singlet states
188: in the quenched approximation.  Once quark loops are allowed in the
189: vacuum, for the favour-singlet states of any given $J^{PC}$,  there will
190: be mixing between the  $s \bar{s}$ state, the  $u \bar{u}+d \bar{d}$
191: state  and the glueball. 
192:  One way to explore this is to measure directly the scalar  mass
193: eigenstates in a study with $N_f=2$ flavours of sea-quark.
194:  Most studies show no significant change of the glueball spectrum as
195: dynamical quark effects are added - but  the sea quark masses used are
196: still rather large~\cite{sesam,bali}. A recent study~\cite{lat99},
197: however,  does find evidence for a reduced mass, albeit with a rather
198: large lattice spacing,  see fig.~1.
199:  This effect could be due to mixing of scalar mesons and glueballs, as
200: we discuss  below, or might just be a sign of an enhanced order $a^2$ 
201: correction at the  relatively large lattice spacing used. 
202: 
203: Let us now discus the mixing of the scalar glueball and scalar mesons.
204: The  mass spectrum of $q \bar{q}$ states has been determined on a 
205: quenched lattice and  the scalar mesons  are found to lie somewhat
206: lighter than the tensor states~\cite{livhyb}. These  $2^{++}$ mesons are
207: experimentally almost unmixed and  so  will be quite close to the
208: quenched mass determination. This suggests  that the quenched scalar
209: masses from the lattice are at around 1.2 GeV and 1.5 GeV (for $n
210: \bar{n}$ and $s \bar{s}$ respectively). An independent
211: study~\cite{weinss,weinssg}  suggests that the scalar $s \bar{s}$ state
212: is about 200 MeV lighter than the glueball which is a broadly compatible
213: conclusion.  Thus the glueball, at around 1.6 GeV, lies heavier than the
214: lightest $q\bar{q}$ scalar states. This information can then be combined
215: with mixing strengths to give the resulting scalar spectrum. 
216: 
217: It is possible to measure the mixing strength on a quenched lattice even
218: though no mixing actually occurs. On a rather coarse lattice ($a^{-1} 
219: \approx 1.2$ GeV), two groups  have attempted this~\cite{weinssg,lat99}.
220: Their results expressed as the mixing for two degenerate quarks of mass
221: around the strange quark mass  are similar, namely $E \approx 0.3$
222: GeV~\cite{weinssg} and 0.5 GeV~\cite{lat99}. From this evaluation of the
223: mixing strength, one can use  a mass matrix to estimate the mass shift
224: induced in the glueball and  scalar meson.
225:  The relevant mass matrix   is (in a glueball, $q \bar{q}$ basis in GeV
226: units):
227: 
228: \begin{math} 
229:  \left( \begin{array}[h]{cc}
230:     1.2  &    0.5 \\
231:     0.5  &   1.8 \\
232: \end{array} \right)
233: \end{math}
234: 
235: \noindent which would give a downward shift of the glueball mass by
236: 20\%. 
237:  This is in qualitative agreement with our direct determination with 
238: $N_f=2$ flavours of sea-quark that the lightest scalar mass is reduced 
239: significantly at this lattice spacing as shown in fig.~1.
240: 
241: 
242:  Note that at this coarse lattice spacing the quenched glueball mass is 
243: reduced (see fig.~1) below the canonical value of 1.6 GeV. Thus a study 
244: at smaller lattice spacing is needed. An exploratory attempt to
245: extrapolate to  the continuum~\cite{weinssg} gave a very small mixing of
246: 86(64) MeV, while the  other determination~\cite{lat99} uses clover
247: improvement so order $a$ effects in the extrapolation to the continuum
248: are suppressed and one would not expect a significant decrease in going
249: to  the continuum limit. 
250:  What this discussion shows is that precision studies of the mixing on 
251: a lattice have not yet been achieved.
252: 
253: 
254: 
255: \begin{figure}[t]
256: \psfig{figure=continuum_glueballs.eps,height=7cm,width=9cm}
257: \vspace{-0.5cm}
258:  \caption{ The continuum glueball spectrum{\protect\cite{mpglue}}. 
259:  }
260:  \label{gbr0}
261: \end{figure}
262: 
263: 
264: 
265: As well as this mixing of the glueball with $q \bar{q}$ states, there
266: will be  mixing  with $q \bar{q} q \bar{q}$ states which will be
267: responsible for the  hadronic decays. A first attempt to study
268: this~\cite{gdecay} yields an estimated width for decay to two
269: pseudoscalar mesons from the scalar glueball of order 100 MeV.  A more
270: realistic study  would involve taking account of mixing with the $n
271: \bar{n}$ and $s \bar{s}$ scalar mesons as  well.
272: 
273: 
274: \section{Exotic states}
275: 
276: 
277:  By exotic state we mean any state which is not dominantly a $q \bar{q}$
278:  or $qqq$ state. Such examples have been known for a long time: the
279: deuteron  is a proton-neutron molecule for example. It is very weakly
280: bound (2 MeV)  and is quite extended. Similar molecular states involving
281: two mesons have been conjectured. 
282: 
283:  One case which is relatively easy to study is the $BB$ system,
284: idealised as two  static quarks and two light quarks. Then a potential
285: as a function of the separation  $R$ between the static quarks can be
286: determined.  Because the static quark spin is irrelevant, the states can
287: be classified by the light quark spin and  isospin.  Lattice
288: results~\cite{cmpp}  (using a light quark mass close to strange) have
289: been obtained for the potential energy for $I_q=0,1$ and $S_q=0,1$. For
290: very  heavy quarks, a potential below $2M_B$  will imply binding of the
291: ${BB}$ molecules with these quantum numbers and $L=0$. For the
292: physically relevant case  of $b$ quarks of around 5 GeV, the kinetic
293: energy will not be negligible and the binding energy of the ${ BB}$
294: molecular states is less  clear cut. One way to estimate the kinetic
295: energy for the ${ BB}$ case with reduced mass circa 2.5 GeV is to use
296: analytic approximations to the  potentials found. For example the
297: $I_q,S_q$=(0,0) case shows a deep  binding at $R=0$ which  can be
298: approximated as a Coulomb potential of $-0.1/R$ in GeV units. This will
299: give a di-meson binding energy of only 10 MeV.  For the other
300: interesting case, $(I_q,S_q)$=(0,1), a  harmonic oscillator potential in
301: the radial coordinate of form $-0.04[ 1- (r-3)^2/4]$ in GeV units leads
302: to a kinetic energy  which completely cancels the potential energy
303: minimum, leaving zero  binding. This harmonic oscillator approximation
304: lies above the estimate of  the potential, so again we expect weak
305: binding of the di-meson system.
306: 
307:  Because of these very small values for the di-meson binding energies, 
308: one needs to retain corrections to the heavy quark approximation to 
309: make more definite predictions, since these corrections are known to 
310: be of magnitude 46 MeV from the $B$, $B^*$ splitting. It will also be 
311: necessary to extrapolate the  light quark mass from strange to 
312: the lighter $u,\ d$ values to make more definite predictions 
313: about the binding of $BB$ molecules.
314: 
315: Models for the binding of two $B$ mesons involve, as in the case of the
316: deuteron,  pion exchange. The lattice study~\cite{cmpp} is able to make a
317: quantitative comparison of lattice pion  exchange with the data
318: described above and excellent agreement is obtained at larger $R$ 
319: values, as expected.
320: 
321: 
322: 
323:  \section{Hybrid Mesons}
324: 
325: 
326:  By hybrid meson, I mean a meson in which the gluonic degrees of freedom
327: are  excited non-trivially. 
328:  I first discuss hybrid mesons with static heavy quarks where the
329: description  can be thought of as an excited colour string. I then
330: summarise the situation  concerning light quark hybrid mesons. 
331: 
332: 
333: Consider $Q \bar{Q}$ states with static quarks  in which the gluonic
334: contribution may be excited. We  classify the gluonic fields according
335: to the symmetries of the system.  This discussion is very similar to the
336: description of electron wave functions in  diatomic molecules. The
337: symmetries are  (i) rotation around the separation axis $z$ with
338: representations labelled by $J_z$ (ii) CP with representations labelled
339: by $g(+)$ and $u(-)$ and (iii) C$\cal{R}$. Here  C interchanges $Q$ and
340: $\bar{Q}$, P is parity and $\cal{R}$ is a rotation  of $180^0$ about the
341: mid-point around the $y$ axis. The C$\cal{R}$ operation is only relevant
342:  to classify states with $J_z=0$. The convention is to label states of
343: $J_z=0,1,2$ by $ \Sigma, \Pi, \Delta$  respectively. The ground state
344: ($\Sigma^+-g$) will have $J_z=0$ and $CP=+$.
345: 
346:  The exploration of the energy levels  of other representations has a
347: long history in lattice studies~\cite{liv,pm}. The first excited state
348: is found  to be the $\Pi_u$.  This can be visualised  as the symmetry of
349: a string bowed out in the $x$ direction minus the same  deflection in
350: the $-x$ direction (plus another component of  the two-dimensional
351: representation with the transverse direction $x$ replaced by $y$),
352: corresponding to flux  states from a lattice  operator which is the
353: difference of U-shaped paths from quark to antiquark of the form $\,
354: \sqcap - \sqcup$.
355: 
356: 
357: 
358: 
359: Recent lattice studies~\cite{jkm}  have used an asymmetric space/time
360: spacing which enables excited states to be  determined comprehensively.
361:  These results confirm the finding that 
362: the $\Pi_u$ excitation is the lowest lying and hence of most relevance 
363: to spectroscopy.
364: 
365:  From the potential corresponding to these excited gluonic states, one
366: can  determine the spectrum of hybrid quarkonia using the Schr\"odinger
367: equation in the Born-Oppenheimer approximation.  This approximation will
368: be good if the heavy quarks move very little in the  time it takes for
369: the potential between them to become established. More  quantitatively,
370: we require that the potential energy of gluonic excitation is much
371: larger than the typical energy of orbital or radial excitation.  This is
372: indeed the case~\cite{liv}, especially for $b$ quarks. Another nice
373: feature of this approach is that the  self energy of the static sources
374: cancels in the energy difference between this  hybrid state and the
375: $Q \bar{Q}$ states. Thus the lattice approach gives directly the
376: excitation energy  of each gluonic excitation.
377: 
378:   The $\Pi_u$ symmetry state corresponds to  excitations of the gluonic
379: field in quarkonium called magnetic (with $L^{PC}=1^{+-}$) and
380: pseudo-electric (with $1^{-+}$) in contrast to the usual  P-wave orbital
381: excitation which has $L^{PC}=1^{--}$. Thus we expect different quantum
382: number assignments from those of the gluonic ground state. Indeed
383: combining with the heavy quark spins, we get a degenerate  set of 8
384: states with    $J^{PC}=1^{--}$, $ 0^{-+}$, $ 1^{-+}$, $ 2^{-+}$ and  
385: $1^{++},\ 0^{+-},\ 1^{+-},\ 2^{+-}$  respectively. Note that of these, 
386: $J^{PC}=  1^{-+},\ 0^{+-}$ and   $2^{+-}$  are spin-exotic and hence
387: will not mix with $Q\bar{Q}$ states. They thus form a very attractive
388: goal for experimental searches for hybrid  mesons.
389: 
390: 
391:  The eightfold degeneracy of the static approach will be broken by 
392: various corrections. As an example, one of the eight degenerate  hybrid
393: states is a pseudoscalar with the heavy quarks in a spin triplet.  This
394: has the same overall quantum numbers as the S-wave  $Q \bar{Q}$ state
395: ($\eta_b$) which, however, has the heavy quarks in a spin singlet. So
396: any  mixing between these states must be mediated by spin dependent
397: interactions.  These spin dependent interactions will be smaller for
398: heavier quarks. It is  of interest to establish the strength of these
399: effects for $b$ and $c$ quarks. Another topic of interest is the
400: splitting  between the spin exotic hybrids which will come from the
401: different  energies  of the magnetic and pseudo-electric gluonic
402: excitations.
403: 
404: 
405:  One way to go beyond the static approach is to use the NRQCD
406: approximation which then enables  the spin dependent effects to be
407: explored.  One study~\cite{jkm} finds that the  $L^{PC}=1^{+-}$ and
408: $1^{-+}$ excitations  have no statistically significant splitting 
409: although the $1^{+-}$  excitation does lie a little lighter. This would
410: imply, after adding in heavy quark spin, that  the $J^{PC}=1^{-+}$
411: hybrid was the lightest spin exotic. Also a relatively large spin
412: splitting was found~\cite{cppacs} among the triplet states considering,
413: however,   only
414:  magnetic gluonic excitations. 
415: 
416: 
417: 
418: \begin{figure}[th]
419: \psfig{file=pot1575.ps,height=7cm,width=9cm}
420: \vspace{-0.5cm}
421:  \caption{The potential energy for quenched and 2 flavours of sea quark
422: for the ground state and first excited gluonic
423: state~{\protect\cite{bali}}.
424:  }
425:  \label{balif}
426: \end{figure} 
427: 
428: 
429:  
430:  Confirmation of the ordering of the spin exotic states also comes from
431:  lattice studies with propagating quarks~\cite{livhyb,milc,sesamhyb}
432: which  are able to measure masses for all 8 states. We  discuss this
433: evidence in more detail below.
434: 
435:  Within the quenched approximation,  the lattice evidence  for
436: $b\bar{b}$ quarks points to a  lightest hybrid spin exotic with
437: $J^{PC}=1^{-+}$ at an energy given by $(m_H-m_{2S})r_0$ =1.8 (static
438: potential~\cite{pm}); 1.9 (static potential~\cite{jkm},
439: NRQCD~\cite{cppacs}); 2.0 (NRQCD~\cite{jkm}). These results can be
440: summarised as       $(m_H-m_{2S})r_0=1.9 \pm 0.1$.
441:  Using the experimental mass of the $\Upsilon(2S)$, this implies that
442: the lightest spin exotic  hybrid is at $m_H=10.73(7)$ GeV including a
443: 10\% scale error.  Above this energy there will be many more hybrid 
444: states, many of which will be spin exotic. A  discussion of hybrid decay 
445: channels has been given~\cite{hf8}.
446: 
447: 
448:  The  excited gluonic static potential has also been determined
449: including sea quarks  ($N_f=2$ flavours) and no significant difference
450: is seen~\cite{bali}- see fig.~\ref{balif}. Thus the quenched estimates
451: given above are not superseded. 
452: 
453: 
454: 
455:  I now  focus on lattice results for hybrid mesons made from light
456: quarks using fully relativistic propagating quarks.  There will be no
457: mixing with $q \bar{q}$ mesons for  spin-exotic hybrid mesons  and these
458: are of special interest. The first study of this area was by the  UKQCD
459: Collaboration~\cite{livhyb} who used operators motivated by the  heavy
460: quark studies referred to above to study all 8 $J^{PC}$ values coming
461: from $L^{PC}=1^{+-}$ and $1^{-+}$ excitations. The  resulting mass
462: spectrum  gives the $J^{PC}=1^{-+}$ state as the lightest spin-exotic
463: state. Taking account of the systematic scale errors in the lattice
464: determination, a  mass of 2000(200) MeV is quoted for this hybrid meson
465: with $s \bar{s}$ light quarks. Although not directly measured, the
466: corresponding light quark hybrid meson would be expected to be around
467: 120 MeV lighter. 
468: 
469: 
470: 
471: A second lattice group has also evaluated hybrid meson spectra with
472: propagating quarks from quenched lattices. They obtain~\cite{milc}
473: masses of the $1^{-+}$ state with statistical and various systematic
474: errors of  1970(90)(300) MeV, 2170(80)(100)(100) MeV and 4390(80)(200)
475: MeV for $n \bar{n}$,  $s \bar{s}$ and $c \bar{c}$ quarks respectively.
476: For the  $0^{+-}$ spin-exotic state they have a noisier signal but
477: evidence that it is heavier. They also explore mixing matrix elements
478: between spin-exotic hybrid  states and 4 quark operators. 
479: 
480:  A first attempt has been made~\cite{sesamhyb} to determine the hybrid
481: meson spectrum using  full QCD. The sea quarks used have several
482: different masses and an extrapolation  is made to the limit of physical
483: sea quark masses, yielding a mass of 1.9(2) GeV for the lightest 
484: spin-exotic hybrid meson, which again is found to be the $1^{-+}$. In
485: principle this  calculation should take account of sea quark effects
486: such as the mixing  between such a hybrid meson and $q \bar{q} q
487: \bar{q}$ states such as $\eta \pi$, although it is possible that the sea
488: quark  masses used are not light enough to explore these features.
489: 
490: 
491: The three independent lattice calculations of the light hybrid spectrum
492: are  in good agreement with each other. They imply that the natural
493: energy  range for spin-exotic hybrid mesons is around 1.9 GeV. The
494: $J^{PC}=1^{-+}$  state is found to be lightest. It is not easy to
495: reconcile these lattice results  with experimental
496: indications~\cite{expt} for resonances at 1.4 GeV and 1.6 GeV,
497: especially the  lower mass value.  Mixing  with  $q \bar{q} q \bar{q}$
498: states such as $\eta \pi$ is not included for realistic quark masses in
499: the  lattice calculations. This can be interpreted, dependent on one's
500: viewpoint,  as either that the lattice calculations  are incomplete or
501: as an indication that the experimental states may have an  important
502: meson-meson component in them. 
503: 
504: 
505:  
506:  \section{String breaking}
507: 
508: 
509:  Although not directly related to exotic states as such, a discussion of
510: string  breaking is relevant to a study of the confinement mechanism.
511: Here I concentrate  on the situation of a static quark and antiquark at
512: separation $R$ with a  gluonic field in a given representation labelled
513: by $J_z$ and $CP$ where the  separation axis is $z$. The ground state
514: gluonic configuration has $J_z=0$  and $CP=+$ (also $C \cal{R}=+$, see
515: discussion above).
516: 
517: 
518: 
519:  If a light quark and antiquark  are to be produced, they must respect
520: these symmetries. When these light quarks  have no angular momentum
521: about the separation axis, they will  either be in  a singlet state
522: ($S_z=0,\ CP=-$) or a triplet state ($S_z=0 \ {\rm or}\ S_z=1,  \
523: CP=+$). Thus the only possible transition is to the $S_z=0$ triplet
524: state. This  is known as the $^3P_0$ model, although here it is exact.
525: Having established the allowed transitions, I now discuss their
526: strength.
527: 
528: Consider the energy of a static $Q \bar{Q}$ versus $R$ for two different
529: situations: (i) a gluonic flux in the ground state  (this is the static
530: potential $V(R)$) and (ii)   the energy  of two ground state mesons ($Q
531: \bar{q}$ plus $\bar{Q}q$) at separation $R$ (ie as $B$ and $\bar{B}$ 
532: mesons but with static heavy quark). We expect  (i) to be lowest lying
533: at small $R$ and (ii) to be lowest lying at large $R$.  Where these
534: energies become degenerate is the string-breaking  separation $R_B$.
535: This is around 1.2 fm - see fig.~\ref{balif}. As has been emphasised in
536: a study of adjoint string breaking~\cite{cmadj}, the lattice enables
537: string breaking to be studied as a mixing phenomenon.
538: 
539: 
540: In the quenched approximation there will be no string breaking  but the
541: matrix element can be estimated. In full QCD studies with sea quarks, 
542: string breaking is enabled and there will be mixing between these
543: states. If  this mixing were substantial, the ``potential energy''
544: determined from Wilson loops  should show a saturation at energy
545: $V(R)=2m_{Q\bar{q}}$ for $R > R_B$. Unfortunately, this effect has not
546: yet  been seen directly in lattice studies. 
547: 
548:  A comprehensive lattice study~\cite{cmppsb} has been made with static
549: quarks and  light quarks of mass around the strange mass. This study
550: finds that  observation of the direct splitting at $R \approx R_B$ has
551: errors too large to be conclusive. The magnitude of the mixing matrix 
552: element can, however, be evaluated with adequate accuracy. This suggests
553: that indeed  the mixing is weak (energy mixing strength of a few tens of
554: MeV only). Such a weak mixing would explain why no direct evidence has
555: been seen  for string breaking from lattice studies of Wilson loops. 
556: 
557:  This discussion of string breaking can be extended~\cite{hf8} to cover
558: hybrid meson decays as well. 
559: 
560: 
561:  \section{Conclusions}
562: 
563:  Quenched lattice QCD is well understood and accurate predictions in the
564: continuum limit  are increasingly becoming available. The lightest
565: glueball  is scalar with mass  $m(0^{++})=1611(30)(160)$ MeV where the
566: second error is an overall scale error. The excited glueball spectrum is
567: known too. The quenched approximation  also gives information on
568: quark-antiquark scalar mesons and their mixing with glueballs. This
569: determination of the mixing in the quenched approximation  also sheds
570: light on results for the  spectrum directly  in full QCD where the
571: mixing will be enabled. There is also some lattice information  on the
572: hadronic decay  amplitudes of glueballs. 
573: 
574: 
575:  Evidence exists for a  possible $B_s B_s$ molecular state.
576: 
577:  For hybrid mesons, there will be no mixing with $q \bar{q}$ for 
578: spin-exotic states and these are the most useful predictions. The
579: $J^{PC}=1^{-+}$ state is expected at 10.73(7) GeV for $b$ quarks,
580:  2.0(2) GeV for $s$ quarks and 1.9(2) GeV 
581: for $u,\ d$ quarks. Mixing of spin-exotic hybrids with
582: $q\bar{q}q\bar{q}$ or equivalently with meson-meson  is  allowed and
583: will modify the  predictions from the quenched approximation.
584: 
585:  String breaking at a separation of 1.2 fm has been studied and the 
586: breaking matrix element is found to be weak.
587: 
588: 
589: 
590: \section*{References}
591: \begin{thebibliography}{99}
592: 
593: %glueball
594: 
595: \bibitem{DForc} P. De Forcrand et al., {\em Phys.\ Lett.} {\bf B152}, 
596: 107 (1985).
597: 
598: \bibitem{MTgl} C. Michael and M. Teper, {\em Nucl.\ Phys.} {\bf B314},
599: 347 (1989). 
600: 
601: \bibitem{ukqcd}  UKQCD collaboration, G. Bali, et al.,
602: {\em Phys.\ Lett.} {\bf B309}, 378 (1993). 
603: 
604: \bibitem{gf11} H. Chen et al.,
605: {\em Nucl.\ Phys. B (Proc.\ Suppl.)} {\bf 34}, 357 (1994); 
606: A. Vaccarino and D. Weingarten, \Journal{\PRD}{60}{1999}{114501}.
607: 
608: \bibitem{sesam}G. Bali et al., 
609: {\em Nucl.\ Phys. B (Proc.\ Suppl.)} {\bf 63}, 209 (1998).
610: 
611: \bibitem{lat99}    C. Michael, M. S. Foster and C. McNeile,
612: {\em Nucl. Phys. B (Proc. Suppl.)} {\bf 83-84}, 185 (2000).
613: 
614: \bibitem{mpglue} C. Morningstar and M. Peardon, {\em Phys. Rev.} {\bf
615: D56}, 4043 (1997); {\em ibid.}, {\bf D60}, 034509 (1999) .
616: 
617: 
618: \bibitem{nrw} F. Niedermeyer, P. R\"ufenacht and U. Wenger,
619: hep-lat/0007007.
620: 
621: \bibitem{livhyb} UKQCD Collaboration,
622:   P. Lacock, C. Michael, P. Boyle  and P. Rowland, 
623: {\em Phys.\ Rev.} {\bf D54},  6997 (1996); 
624: {\em Phys.\ Lett.} {\bf B401},  308 (1997); 
625:  {\em Nucl. Phys. B (Proc. Suppl.)} {\bf 63},  203 (1998).
626: 
627: \bibitem{weinss} W. Lee and D. Weingarten, {\em Nucl. Phys. B (Proc.
628: Suppl)} {\bf 53}, 236(1997); {\em Nucl. Phys. B (Proc.  
629: Suppl)} {\bf 73}, 249 (1999).
630: 
631: \bibitem{weinssg}  W. Lee and D. Weingarten, {\em Nucl. Phys. B (Proc.
632: Suppl)} {\bf 63},  194 (1998);  hep-lat/9805029;
633: \Journal{\PRD}{61}{2000}{014015}.
634: 
635: \bibitem{gdecay} J. Sexton,  A.  Vaccarino  and D.  Weingarten,
636: {\em Nucl.\ Phys.\ B (Proc.\ Suppl.)} {\bf 42},  279 (1995).
637: 
638: \bibitem{cmpp} C. Michael and P. Pennanen, {\em Phys. Rev.} {\bf D60},
639: 054012(1999).
640: 
641: %hybrid
642: 
643: \bibitem{liv}  L.A. Griffiths, C. Michael and  P.E.L. Rakow,
644: {\em  Phys.\ Lett.} {\bf B129},  351 (1983).
645: 
646: \bibitem{pm} S. Perantonis and C. Michael, {\em Nucl.\ Phys.} {\bf B347},
647: 854 (1990) .
648: 
649: \bibitem{jkm} K. Juge , J. Kuti and C. Morningstar, 
650: \Journal{\PRL}{82}{4400}{1999}; {\em Nucl. Phys. B (Proc. Suppl)} {\bf
651: 83}, 304 (2000),hep-lat/9909165.
652: 
653: \bibitem{cppacs}CP-PACS Collaboration, T. Manke et al., 
654: \Journal{\PRL}{82}{1999}{4396}; {\em Nucl. Phys. B (Proc. Suppl)} {\bf
655: 86}, 397 (2000), hep-lat/9909038; {\em Nucl. Phys. B (Proc. Suppl)} {\bf
656: 83}, 319 (2000), hep-lat/9909133.
657:      
658: \bibitem{milc} C. Bernard et al., {\em Nucl. Phys. B (Proc. Suppl.)}
659: {\bf 53},  228 (1996); {\em Phys. Rev.} {\bf D56},  7039 (1997);
660: {\em Nucl. Phys. B (Proc. Suppl.)} {\bf 73}, 264 (1999), hep-lat/9809087.
661: 
662: \bibitem{sesamhyb}  P. Lacock and K. Schilling,
663: {\em Nucl. Phys. B (Proc. Suppl.)}
664: {\bf  73},  261 (1999),
665: hep-lat/9809022
666: 
667: \bibitem{bali} SESAM and T{$\chi$}L Collaboration, 
668: G. Bali et al., hep-lat/0003012.% hep-lat/9901023
669: 
670: \bibitem{expt}  D. Thompson et al., \Journal{\PRL}{ 79}{1997} {1630};
671: S. U. Chung et al., \Journal{\PRD}{60}{1999}{092001};
672: D. Adams et al., \Journal{\PRL}{81}{1998}{5760}
673: 
674: \bibitem{cmadj}    C. Michael,   {\em   Nucl. Phys. B (Proc. Suppl.)}
675: {\bf 6}, 417 (1992).
676: 
677: \bibitem{cmppsb} P. Pennanen and C. Michael, hep-lat/0001015  
678: 
679: \bibitem{hf8} C. Michael, {\em Proceedings of Heavy Flavours 8}, 
680: Southampton 1999,  JHEP, hep-ph/9911219.
681: 
682: \end{thebibliography}
683: 
684: \end{document}
685: 
686: #!/bin/csh -f
687: # this uuencoded Z-compressed .tar file created by csh script  uufiles
688: # for more information, see e.g. http://xxx.lanl.gov/faq/uufaq.html
689: # if you are on a unix machine this file will unpack itself:
690: # strip off any mail header and call resulting file, e.g., figs.uu
691: # (uudecode ignores these header lines and starts at begin line below)
692: # then say        csh figs.uu
693: # or explicitly execute the commands (generally more secure):
694: #    uudecode figs.uu ;   uncompress figs.tar.Z ;
695: #    tar -xvf figs.tar
696: # on some non-unix (e.g. VAX/VMS), first use an editor to change the
697: # filename in "begin" line below to figs.tar_Z , then execute
698: #    uudecode figs.uu
699: #    compress -d figs.tar_Z
700: #    tar -xvf figs.tar
701: #
702: uudecode $0
703: chmod 644 figs.tar.Z
704: zcat figs.tar.Z | tar -xvf -
705: rm $0 figs.tar.Z
706: exit
707: 
708: