1: \documentclass[nofootinbib,twocolumn,aps,amsmath,amssymb]{revtex4}
2: \usepackage{graphics,bm}
3: \usepackage{epsfig}
4: \usepackage{graphicx}
5:
6: \def\sumintdiff{\hbox{$\Delta$}\!\!\!\!\!\!\int}
7:
8: \newcommand{\beq}{\begin{equation}}
9: \newcommand{\eeq}{\end{equation}}
10: \newcommand{\bqa}{\begin{eqnarray}}
11: \newcommand{\eqa}{\end{eqnarray}}
12: \newcommand{\simorder}{\raisebox{-4pt}{$\, \stackrel{\textstyle >}{\sim} \,$}}
13: \newcommand{\simordertwo}{\raisebox{-4pt}{$\, \stackrel{\textstyle<}{\sim}\,$}}
14: \newcommand{\unity}{1\hspace{-1.3mm}1}
15:
16: % slashchar, for slashed characters.
17: \def\slashchar#1{\setbox0=\hbox{$#1$} % set a box for #1
18: \dimen0=\wd0 % and get its size
19: \setbox1=\hbox{/} \dimen1=\wd1 % get size of /
20: \ifdim\dimen0>\dimen1 % #1 is bigger
21: \rlap{\hbox to \dimen0{\hfil/\hfil}} % so center / in box
22: #1 % and print #1
23: \else % / is bigger
24: \rlap{\hbox to \dimen1{\hfil$#1$\hfil}} % so center #1
25: / % and print /
26: \fi}
27:
28: \begin{document}
29:
30: %\voffset=-20truemm
31: %\hoffset=-15truemm
32: %\textwidth=167 truemm
33: %\textheight=210truemm
34: \parindent=20pt
35: \pagestyle{plain}
36: \font\tenrm=cmr10
37:
38:
39: \author{Harmen J. Warringa and Dani\"el Boer} \affiliation{Department
40: of Physics and Astronomy, Vrije Universiteit, De Boelelaan 1081, 1081
41: HV Amsterdam, The Netherlands}
42: \email{harmen@nat.vu.nl, dboer@nat.vu.nl}
43:
44: \title{Color superconductivity
45: vs.\ pseudoscalar condensation in a three-flavor NJL model}
46: \author{Jens O. Andersen}
47: \affiliation{Nordita, Blegdamsvej 17, DK-2100 Copenhagen {\O},
48: Denmark} \email{jensoa@nordita.dk} \ \date{\today}
49:
50: \begin{abstract}
51: We calculate numerically the phase diagram of the three-flavor
52: Nambu--Jona-Lasinio model at zero and finite temperature as a function
53: of the up, down, and strange quark chemical potentials.
54: We focus on the competition between
55: pseudoscalar condensation and color superconductivity. We find that
56: the two types of phases are separated by first-order transitions.
57: \end{abstract}
58:
59: \maketitle
60:
61: %=====================================================================
62: \section{Introduction}
63: In recent years a huge effort has been made in order to understand the
64: phase diagram of strongly interacting matter, both experimentally as
65: well as theoretically. Collisions at the Relativistic Heavy-Ion
66: Collider (RHIC) at BNL and the Large Hadron Collider (LHC) at CERN
67: allow the experimental study of hadronic matter at energy densities
68: exceeding that required to create a quark-gluon plasma. The energy and
69: baryon densities of these experiments correspond to temperatures up to
70: 200 MeV and a baryon chemical potential in the range 0-600 MeV.
71:
72: At zero baryon chemical potential, lattice calculations suggest the
73: existence of a transition from ordinary nuclear matter to a
74: quark-gluon plasma at temperatures of around 160 MeV (see for example
75: Ref.\ {\cite{Karsch2000}}). While there has been progress in
76: generalizing lattice calculations to finite baryon chemical potential
77: \cite{Fodor2002b, deForcrand2002, Allton2002, DElia2003}, highly
78: nontrivial problems remain to be solved, and in practice one is
79: restricted to small baryon chemical potentials.
80:
81: Since one cannot apply lattice gauge theories to the region of the
82: phase diagram where the baryon chemical potential is large, one must
83: employ other methods. One possibility is perturbative QCD.
84: Asymptotic
85: freedom guarantees that this method can be applied at asymptotically
86: large chemical potentials. If the densities are too small to
87: use perturbative QCD, one may apply effective theories. Models such as
88: the instanton liquid model, random matrix models, and
89: Nambu--Jona-Lasinio (NJL) models have been applied to the study of the
90: QCD phase diagram at finite temperature and densities. Despite their
91: shortcomings, it is expected that such models do describe the
92: qualitative features of the QCD phase diagram, in regions not
93: accessible to perturbative or lattice QCD. For example, the existence
94: of a critical point at finite baryon chemical potential has been predicted
95: using effective models~\cite{Halasz1998, Berges1998}. Furthermore, it
96: is expected that quark matter at high baryon chemical potentials and
97: low temperatures can be in a two-flavor color-superconducting phase
98: (2SC) \cite{Bailin1984, Alford1997, Rapp1997} or a color-flavor locked
99: (CFL) phase \cite{Alford1999, Alford1999q}. In these phases gaps on the order of
100: 100 MeV arise.
101:
102: Many results have been obtained for equal up, down and strange quark
103: chemical potentials. However, this may not be directly relevant for
104: heavy-ion collisions or compact stars. For example, to enforce
105: electric and color neutrality and weak equilibrium in compact stars
106: different flavor and color chemical potentials have been
107: introduced~\cite{Alford2002, Steiner2002}. This gives rise to a more
108: complicated phase diagram in which one also finds new 2SC-like
109: \cite{Neumann2002} and gapless superconducting phases
110: \cite{Alford1999b, Shovkovy2003, Alford2004a, Alford2004c, Ruster2004, Abuki2004, Ruster2005,
111: Blaschke2005}. Similarly, in heavy-ion collisions, a difference
112: between the quark chemical potentials arises if the number densities
113: of the different quark flavors are not the same. This difference can
114: cause interesting observable effects such as two critical
115: endpoints~\cite{Klein2003, Toublan2003}. However, instanton induced
116: interactions tend to suppress this effect \cite{Frank2003}.
117:
118: In addition to a more complicated structure of the superconducting
119: phases, different chemical potentials can also trigger pseudoscalar
120: condensation \cite{Son2001, Kogut2001, Barducci2003}. This has been
121: confirmed on the lattice at zero baryon but finite isospin chemical
122: potential \cite{Kogut2002}. Depending on the flavors involved, the
123: charged pion, neutral or charged kaon fields may acquire a vacuum
124: expectation value. Pseudoscalar condensation in the two-flavor NJL
125: model has been studied in Refs.\ \cite{Toublan2003, Barducci2004}, as
126: a function of the different chemical potentials at zero and finite
127: temperature. An extension to three flavors was carried out in Ref.\
128: \cite{Barducci2005}.
129:
130: In this paper, we discuss the phase diagram of the three-flavor NJL
131: model as a function of the different chemical potentials including
132: both pseudoscalar condensation and color superconductivity. We did not
133: apply electric or color neutrality conditions. At zero temperature,
134: pseudoscalar condensation is possible if $\vert \mu_u - \mu_d \vert >
135: m_\pi$ \cite{Son2001} or if $\vert \mu_{u,d} - \mu_s \vert > m_K$
136: \cite{Kogut2001}. On the other hand, color superconducting phases
137: occur if the chemical potentials are large and approximately
138: equal. Therefore, one can imagine scenarios where for example $\mu_u
139: \approx \mu_d$ (the Fermi surfaces of the $u$ and $d$ quark should be
140: sufficiently close for Cooper pairing to occur) and $\mu_u \approx -
141: \mu_s$ (the Fermi surfaces of the $u$ and $\bar{s}$ should be
142: sufficiently close for kaon condensation to occur), with $\vert
143: \mu_{u,d} - \mu_s \vert > m_K$. In such a case a 2SC phase is
144: competing against a phase in which kaons condense. From our
145: calculations it follows that a coexistence phase of pseudoscalar condensation and
146: color superconductivity does not occur for the
147: parameters chosen and that these phases are separated by a first-order
148: transition. However, we do not exclude that other choices of
149: parameters may lead to such a coexistence phase, just as coexistence of
150: color superconductivity and chiral symmetry breaking may occur in the
151: NJL model for specific ranges of parameters \cite{Blaschke2003}. Here we refer
152: to a coexistence phase to mean a phase in which two condensates are nonzero
153: simultaneously.
154:
155: In this work we study pseudoscalar condensation in the quark
156: anti-quark channel. We have not taken into account the pseudoscalar
157: diquark interaction. This interaction is suppressed relative to the
158: scalar diquark interaction due to instantons~\cite{Rapp1997}. However
159: in absence of instanton interactions, if one neutralizes the bulk
160: matter with respect to color and electric charges it is possible to
161: have pseudoscalar diquark condensation with rather large gaps
162: \cite{Buballa2004b}. Pseudoscalar diquark condensation in the NJL
163: model is similar \cite{Buballa2004b} to pseudoscalar condensation in
164: the CFL phase studied with effective chiral models in Refs.\
165: \cite{Casalbuoni1999,Son1999,Casalbuoni2000, Schafer2000,
166: Bedaque2002a, Kaplan2002, Forbes2004}.
167:
168: We emphasize that in this paper we do not impose electric or color neutrality
169: conditions, but that this would qualitatively affect the phase structure,
170: leading for example to the observation that in a macroscopic volume of quark
171: matter the 2SC phase is energetically disfavored~\cite{Alford2002,
172: Steiner2002}. Therefore, our
173: results do not address the qualitative features of the QCD phase structure
174: with neutrality constraints imposed.
175:
176: The paper is organized as follows. In Sec.\ II, we briefly describe
177: the NJL model and the choice of parameters. In Sec.\ III, we discuss the
178: calculations and some of its technical aspects. In Sec.\ IV, we
179: present our results, and in Sec.\ V we conclude.
180:
181: %=====================================================================
182: \section{The NJL model}
183: In the NJL model \cite{Nambu1961}, one treats the interaction between
184: the quarks as a point-like quark color current-current
185: interaction. This naive approximation to QCD works very well in
186: explaining various low-energy observables such as hadron masses
187: \cite{Klevansky1992}. By applying several Fierz transformations to
188: the current-current interaction (see for example Ref.\
189: \cite{Buballa2004a}) and including only terms which give rise to
190: attractive $qq$ and $\bar{q}q$ channels, one obtains the following
191: Lagrangian density
192: \begin{equation}
193: \mathcal{L} = \bar \psi \left(i \gamma^\mu \partial_\mu - M_0
194: + \mu \gamma_0 \right)\psi
195: + \mathcal{L}_{\bar q q} + \mathcal{L}_{qq}
196: \label{eq:lagrnjl} \;,
197: \end{equation}
198: where the quark-antiquark term, $\mathcal{L}_{\bar q q}$, and the
199: diquark interaction term, $\mathcal{L}_{q q}$, are defined below. We
200: have suppressed the color, flavor, and Dirac indices of the fermion
201: fields $\psi$ for notational simplicity. The mass matrix $M_0$ is
202: diagonal and contains the bare quark masses $m_{0u}$, $m_{0d}$ and
203: $m_{0s}$. The matrix $\mu$ is also diagonal and contains the quark
204: chemical potentials $\mu_u$, $\mu_d$ and $\mu_s$. We use the metric
205: $g^{\mu \nu} = \mathrm{diag}(+\,-\,-\,-)$ and the standard representation
206: for the $\gamma$-matrices. The quark-antiquark interaction part of
207: the Lagrangian density is
208: \begin{equation}
209: \mathcal{L}_{\bar q q}
210: = G \left[ \left(\bar \psi \lambda_a \psi \right)^2 +
211: \left(\bar \psi \lambda_a i \gamma_5 \psi \right)^2 \right] \;.
212: \end{equation}
213: The matrices $\lambda_a$ are the $9$ generators of $\mathrm{U}(3)$ and
214: act in flavor space. They
215: are normalized as $\mathrm{Tr}\, \lambda_a
216: \lambda_b = 2 \delta_{a b}$.
217: The diquark interaction term of the Lagrangian density is given by
218: \begin{equation}
219: \mathcal{L}_{qq} =
220: \frac{3}{4} G \left(\bar \psi t_{A} \lambda_{B} C i \gamma_5 \bar \psi^T \right)
221: \left(\psi^T t_{A} \lambda_{B} C i \gamma_5 \psi \right)
222: \;,
223: \end{equation}
224: where $A, B \in \{2, 5, 7\}$ since only the interaction in the color
225: and flavor antisymmetric triplet channel is attractive. The matrices
226: $t_a$ are the generators of $\mathrm{U}(3)$ and act in color space.
227: Their normalization is $\mathrm{Tr}\, t_a t_b = 2 \delta_{a b}$. To
228: remind the reader, the antisymmetric flavor matrices $\lambda_2$,
229: $\lambda_5$ and $\lambda_7$ couple up to down, up to strange and down
230: to strange quarks, respectively. The charge conjugate of a field
231: $\psi$ is denoted by $\psi_c = C \bar \psi^T$ where $C = i \gamma_0
232: \gamma_2$. The coupling strength $3 G /4$ of the diquark interaction
233: is fixed by the Fierz transformation. However, some authors discuss the NJL model
234: with a different diquark coupling constant (see for example Ref.\
235: \cite{Ruster2005} for a comparison).
236:
237: The results that will be presented below are obtained with the
238: following choice of parameters
239: \begin{eqnarray}
240: m_{0u} = m_{0d} = 5.5\;\mathrm{MeV} \;,\;\; m_{0s} =
241: 112\;\mathrm{MeV}
242: \nonumber \\
243: G = 2.319 / \Lambda^2 \;, \;\;\Lambda = 602.3\;\mathrm{MeV} \;.
244: \end{eqnarray}
245: This choice of parameters gives rise to constituent quark masses
246: $M_u=M_d=368$ MeV and $M_s=550$ MeV \cite{Buballa2004a}.
247:
248: %=====================================================================
249: \section{The effective potential}
250: To obtain the phase diagram of the NJL model, we first introduce 18
251: real condensates $\alpha_a$ and $\beta_a$, and 9 complex condensates
252: $\Delta_{AB}$ as follows
253: \begin{eqnarray}
254: \alpha_a &=& - 2 G \left < \bar \psi \lambda_a \psi
255: \right>\;, \\
256: \beta_a &=& - 2 G \left < \bar \psi \lambda_a i
257: \gamma_5 \psi
258: \right> \;, \\
259: \Delta_{AB} &=& \frac{3}{2}G \left < \psi^T t_A \lambda_B C \gamma_5
260: \psi \right> \;.
261: \end{eqnarray}
262: We will assume that all condensates are space-time independent. The
263: crystalline Larkin-Ovchinnikov-Fulde-Ferrell (LOFF) phase
264: \cite{Alford2001b} will not be considered here. The next step is
265: to apply a Hubbard-Stratonovich transformation to eliminate the
266: four-point quark interactions to make the Lagrangian quadratic in the
267: quark fields. Introducing a two-component Nambu-Gorkov field $\Psi^T =
268: \left( \psi^T\!,\;\psi_c^T \right) / \sqrt{2}$ allows for
269: straightforward integration over the quark fields. After going to
270: imaginary time, the thermal effective potential ${\cal V}$ in the mean-field
271: approximation reads
272: \begin{multline}
273: \mathcal{V} = \frac{\alpha_a^2 + \beta_a^2}{4G}
274: + \frac{\left \vert \Delta_{AB}\right \vert^2 }{3 G} \\
275: -\frac{T}{2} \sum_{p_0 = (2n + 1)\pi T} \int \frac{\mathrm{d}^3 p}{
276: \left( 2\pi \right)^3} \log \mathrm{det} K \;,
277: \label{eq:effpotential}
278: \end{multline}
279: where $K$ is a $72 \times 72$ matrix
280: \begin{equation}
281: K =
282: \left(
283: \begin{array}{cc}
284: \unity_c \otimes \mathcal{D}_1
285: &
286: \Delta_{AB}\, t_A \otimes \lambda_B \otimes \gamma_5
287: \\
288: -\Delta^*_{AB}\, t_A \otimes \lambda_B \otimes \gamma_5
289: &
290: \unity_c \otimes \mathcal{D}_2
291: \end{array} \right)\;,
292: \end{equation}
293: and %where
294: \begin{eqnarray}
295: \mathcal{D}_1 &=& \unity_f \otimes (i \gamma_0 p_0 +
296: \gamma_i p_i)
297: - \mu \otimes \gamma_0 \nonumber \\
298: & &
299: - ( M_0 +\alpha_a \lambda_a ) \otimes \unity_d
300: - \beta_a \lambda_a \otimes i \gamma_5
301: \;,
302: \\
303: \mathcal{D}_2 &=& \unity_f \otimes (i \gamma_0 p_0 + \gamma_i
304: p_i)
305: + \mu \otimes \gamma_0
306: \nonumber \\
307: & &
308: - ( M_0 + \alpha_a \lambda_a^T)\otimes \unity_d
309: - \beta_a \lambda_a^T \otimes i \gamma_5
310: \;.
311: \end{eqnarray}
312: The matrix $\unity$ is the identity matrix in color ($c$), flavor ($f$),
313: or Dirac ($d$) space.
314:
315: The values of the condensates and the phase diagram are determined by
316: minimizing the effective potential ${\cal V}$ with respect to the
317: condensates. To make the minimization procedure easier, one can take
318: advantage of the fact that certain condensates must vanish. Firstly,
319: application of QCD inequalities \cite{Weingarten1983, Son2001} shows
320: that in QCD at zero chemical potential chiral symmetry breaking cannot be
321: driven by parity-violating condensates of the type $\left < \bar \psi
322: i \gamma_5 \psi \right>$. Outside the phase in which diquarks condense, we
323: found numerically that this is also correct in the NJL model at finite chemical potentials.
324: Therefore, $\beta_0$, $\beta_3$ and
325: $\beta_8$ are put to zero. Secondly, although
326: perturbative one-gluon exchange cannot distinguish between $\beta_k$
327: and $\alpha_k$ condensation with $k \in \{1,2,4,5,6,7\}$, pseudoscalar
328: condensation is favored due to the instanton interaction
329: \cite{Son2001}. We therefore set all $\alpha_k$'s with $k \in
330: \{1,2,4,5,6,7\}$ to zero. We found numerically that this is correct,
331: despite the fact that the model we consider does not include instanton
332: interactions.
333:
334: One can further simplify the minimization procedure by using the
335: symmetries of the NJL model. In absence of quark masses and chemical
336: potentials, the Lagrangian density has a global
337: $\mathrm{SU(3)}_c\times\mathrm{U(3)}_{V}\times\mathrm{U(3)}_{A}$
338: symmetry. Due to the non-vanishing quark masses, the symmetry is
339: broken down to $\mathrm{SU(3)}_c \times \mathrm{U(3)}_{V}$. Since we
340: consider different quark masses and finite chemical potentials, the
341: symmetry of the Lagrangian density is further reduced to
342: $\mathrm{SU(3)}_c\times\mathrm{U(1)}_u\times\mathrm{U(1)}_d
343: \times\mathrm{U(1)}_s $. The vacuum manifold is invariant under the
344: same transformations as the Lagrangian density, so applying a
345: $\mathrm{U}(1)$-flavor transformation to all condensates leaves the
346: free energy invariant. Therefore, using the $\mathrm{U}(1)$-flavor
347: transformations one can choose the pseudoscalars to condense in the
348: $\beta_2$, $\beta_5$, and $\beta_7$ channels, and set $\beta_1$,
349: $\beta_4$ and $\beta_6$ to zero. The phase in which $\beta_2$,
350: $\beta_5$ or $\beta_7$ is non-vanishing is called the $\pi^+/\pi^-$,
351: $K^0 / \bar K^0$ or $K^+/K^-$ condensed phase, respectively.
352:
353: Because of the global $\mathrm{SU}(3)_c$ symmetry, one can also rotate
354: away several diquark condensates. Without loss of generality, we can
355: minimize with respect to $\Delta_{22}$, $\Delta_{25}$, $\Delta_{55}$,
356: $\Delta_{27}$, $ \Delta_{57}$ and $\Delta_{77}$. In principle, all six
357: diquark condensates can have a phase. It is always possible to remove
358: two of them by using the two diagonal $\mathrm{SU}(3)_c$
359: transformations. As long as there is no pseudoscalar condensation, one
360: can use the $\mathrm{U}(1)$-flavor symmetries to rotate away three
361: other phases. As a result either $\Delta_{25}, \Delta_{55},
362: \Delta_{27}$, or $\Delta_{57}$ has a phase \cite{Buballa2004a}.
363: However, this reduction is not completely possible if pseudoscalar
364: condensation occurs. By choosing the pseudoscalars to condense in the
365: $\beta_2$, $\beta_5$ and $\beta_7$ channels, one breaks the
366: $\mathrm{U}(1)$-flavor symmetry. Hence if pseudoscalar condensation
367: arises in one channel, one can in general rotate away one phase less in the diquark
368: sector. If it occurs in more channels, two phases
369: less can be rotated away.
370: However, numerically we find that allowing for a
371: complex phase leads to diquark condensation only in the $\Delta_{22}$,
372: $\Delta_{55}$ and the $\Delta_{77}$ channels. The $\Delta_{25}$,
373: $\Delta_{27}$, and $\Delta_{57}$ diquark condensates do not arise or
374: can be rotated away. Moreover, we find that pseudoscalar condensation in
375: the quark-antiquark channel
376: does not coexist with color superconductivity, such that one can always
377: take the diquark condensates to be real.
378:
379: The different possible color-superconducting phases are
380: named as follows \cite{Ruster2005}
381: \begin{eqnarray}
382: \Delta_{22}\neq0,\,\;\Delta_{55}\neq0,\,\Delta_{77} \neq 0 & & \mathrm{CFL} \;,
383: \nonumber \\
384: \Delta_{77} =0,\;\; \Delta_{22}\neq0,\,\Delta_{55} \neq 0 & & \mathrm{uSC} \;,
385: \nonumber \\
386: \Delta_{55} = 0,\;\;\Delta_{22}\neq0,\,\Delta_{77} \neq 0 & & \mathrm{dSC} \;,
387: \nonumber \\
388: \Delta_{22} = 0,\;\;\Delta_{55}\neq0,\,\Delta_{77} \neq 0 & & \mathrm{sSC} \;,
389: \nonumber \\
390: \Delta_{22} \neq 0, \;\; \Delta_{55}=0,\, \Delta_{77} = 0 & & \mathrm{2SC} \;,
391: \nonumber \\
392: \Delta_{55} \neq 0, \;\; \Delta_{22}=0,\, \Delta_{77} = 0 & & \mathrm{2SCus} \;,
393: \nonumber \\
394: \Delta_{77} \neq 0, \;\; \Delta_{22}=0,\, \Delta_{55} = 0 & & \mathrm{2SCds} \;.
395: \end{eqnarray}
396:
397: To calculate the effective potential in an efficient way, one can multiply
398: the matrix $K$ with $\mathrm{diag}(\unity_c \otimes \unity_f \otimes
399: \gamma_0, \unity_c \otimes \unity_f \otimes \gamma_0 )$ which leaves
400: the determinant invariant. In this way, one obtains a new matrix $K'$
401: with $i p_0$'s on the diagonal. By determining the eigenvalues of the
402: matrix $K'$ with $p_0 = 0$, one can reconstruct the determinant
403: of $K$ for all values of $p_0$ which is namely
404: $\prod_{i=1}^{72}(\lambda_i - ip_0)$. After summing over Matsubara
405: frequencies, one finds
406: \begin{equation}
407: T \!\!\!\!\!\! \sum_{p_0 = (2 n+1) \pi T} \!\!\!\!\!\! \log
408: \mathrm{det} K = \sum_{i=1}^{72} \left[\frac{\lambda_i}{2} + T
409: \log\left(1 + e^{-\lambda_i / T} \right) \right] \;.
410: \end{equation}
411: All that remains in order to determine the effective potential is to
412: integrate over three-momentum $p$ up to an ultraviolet cutoff $\Lambda$.
413:
414: The speed of the calculation of the effective potential depends
415: heavily on how fast one can compute the eigenvalues. There are several
416: ways to speed up the calculation. Firstly, the determinant of $K$ does
417: not depend on the direction of $\vec p$. Therefore, one can choose
418: $\vec p$
419: to lie in the $z$-direction. Together with the choice of the
420: non-vanishing condensates mentioned above, this implies that
421: $K'(p_0=0)$ becomes a real symmetric matrix, which simplifies the
422: calculation of the eigenvalues. Secondly, one can interchange rows and
423: columns of $K'$ without changing its determinant. By doing so, one can
424: bring $K'$ in a block-diagonal form. One can then determine the
425: eigenvalues of the blocks separately which is significantly faster
426: since the time needed to compute eigenvalues numerically scales
427: cubically with the dimension of the matrix. In the most general case
428: with diquark condensation, one can always reduce the problem to two
429: $36 \times 36$ matrices. Moreover, if there is no diquark
430: condensation, but only pseudoscalar condensation, the problem can be
431: further reduced to computing the eigenvalues of two $6 \times 6$ matrices.
432:
433: We determined the eigenvalues using LAPACK routines \cite{Lapack1999}.
434: After numerical integration over three-momentum $p$ up to the
435: cutoff, the condensates were determined by minimizing the effective
436: potential using MINUIT \cite{James1975}. To be certain that the
437: minimization procedure did not end up in a local minimum, we always
438: checked the continuity of the minimized effective potential as a
439: function of chemical potentials and/or temperature.
440:
441: %=====================================================================
442: \section{Phase diagrams}
443: In this section, we present our results for the phase diagrams of the
444: NJL model with $u$, $d$, and $s$ quarks. We plot the phase diagrams
445: as a function of the chemical potentials and temperature. To determine
446: the locations of the phase boundaries, we examine the behavior of the
447: condensates. If a condensate jumps discontinuously
448: the transition is first-order, and this is indicated by a solid line.
449: If its derivative has a
450: discontinuity, the transition is second order, and this is indicated
451: by a dotted line.
452: If a condensate changes rapidly in a narrow range without
453: vanishing there is a smooth cross-over, and this is indicated by
454: a dashed-dotted line at the point were the condensate varies maximally.
455:
456: \subsection{$\mu_s=0$ and $T=0$}
457:
458: In Fig.\ \ref{fig:muumudt0}, we display the phase diagram of the NJL
459: model for $\mu_s=0$ and $T=0$ as a function of $\mu_u$ and
460: $\mu_d$. Outside the 2SC phase (q), our results agree qualitatively with
461: the two-flavor calculations of Ref.\ \cite{Barducci2004} (see their
462: Fig.\ 1), where color-superconducting phases were not taken into account.
463: Moreover, in Ref.\ \cite{Barducci2004} different parameters were used, in addition
464: to a form factor which mimics asymptotic freedom.
465:
466: \begin{figure}[htb]
467: \includegraphics{muUmuDT0.eps}
468: \caption{Phase diagram for $\mu_s = 0$ and $T=0$ as a function of
469: $\mu_u$ and $\mu_d$. First and second-order transitions are indicated
470: by solid and dotted lines, respectively. The
471: letters denote the different phases, where
472: a: $\bar u u$ + $\bar d d$ + $\bar s s$,
473: c: $\bar u u$ + $\bar s s$,
474: d: $\bar d d$ + $\bar s s$,
475: g: $\bar s s$,
476: i: $\pi^+ / \pi^-$ + $\bar s s$ and
477: q: 2SC + $\bar s s$.}
478: \label{fig:muumudt0}
479: \end{figure}
480:
481: One can clearly see that the phase diagram is symmetric under
482: reflection in the origin. This is because the free energy is invariant
483: under the transformation $(\mu_u, \mu_d, \mu_s) \rightarrow (-\mu_u,
484: -\mu_d, -\mu_s)$ from the symmetry between particles and
485: anti-particles. Fig.\ \ref{fig:muumudt0} is also symmetric under
486: interchange of $u$ and $d$, because of the choice of equal up and down
487: quark masses. This gives rise to the symmetry of the phase boundaries
488: with respect to the diagonals.
489:
490: In general, horizontal and vertical lines in the phase diagrams arise
491: if the pairing of one type of quark is not changed after a
492: transition. In this case, the location of the phase boundary is
493: determined by the properties of other quarks. Therefore, changing the
494: chemical potential of the unchanged quark species cannot have a big
495: influence on the location of the phase boundary. This results in the
496: horizontal and vertical lines. For $T=0$, one always finds these lines
497: near the values of the constituent quark masses, i.e.\ $\mu_u \approx
498: M_u$, $\mu_d \approx M_d$ and $\mu_s \approx M_s$ (see for example Ref.\
499: \cite{Buballa2004a}). The diagonal lines arise
500: because at $T=0$ pion condensation can occur if $\vert \mu_u - \mu_d
501: \vert > m_\pi = 138\;\mathrm{MeV}$ \cite{Son2001}.
502:
503: The diagram shows that if the chemical potentials are different, the
504: transition to the color-superconducting phase (q) remains first order as
505: was concluded in Ref.\ \cite{Bedaque2002}. Moreover, one can see from
506: Fig.\ \ref{fig:muumudt0} that if $\mu_u \neq \mu_d$ it is possible to
507: go through two first-order transitions before entering the 2SC phase (q)
508: (similar to the situation discussed in Ref.\ \cite{Toublan2003}
509: without color superconductivity). We observe that to have such a
510: scenario at zero temperature, a minimum difference between $\mu_u$ and
511: $\mu_d$ is required. In the present case this is about 35 MeV. Pion
512: condensation (i) and the 2SC phase (q) are in this diagram separated by two
513: phase transitions in contrast to the estimated $(\mu_B,\mu_I)$ phase
514: diagram of Ref.\ \cite{He2005}.
515:
516: \subsection{$\mu_d = 0$ and $T=0$}
517: In Fig.\ \ref{fig:muumust0} we display the phase diagram for $\mu_d =
518: 0$ and $T=0$ as a function of $\mu_u$ and $\mu_s$. Since the up and
519: down quark masses are much smaller than the strange quark mass, this
520: diagram is very different from Fig. \ref{fig:muumudt0}. Besides the
521: possibility of pion condensation in (h) and (i), phases in which
522: the charged kaon (k) and the neutral kaon condense (l)/(m) arise. The
523: lines separating the charged kaon phase (k) from the chirally broken
524: phase (a) are diagonal because at $T=0$ kaon condensation can occur if
525: $\vert \mu_s - \mu_{u, d} \vert > m_K = 450\;\mathrm{MeV}$
526: \cite{Kogut2001} (the chosen parameter set gives rise to a somewhat low kaon
527: mass, but this is not relevant for the qualitative features of the phase
528: diagram).
529: In (r) we find the 2SCus phase. This phase is surrounded by
530: phases in which the pions (h)/(i) and the neutral kaons (l)
531: condense. We find that one passes a first-order transition when going
532: from the pion and neutral kaon condensed to the 2SCus phase.
533:
534: \begin{figure}[htb]
535: \includegraphics{muUmuST0.eps}
536: \caption{Phase diagram for $\mu_d = 0$ and $T = 0$ as a function of
537: $\mu_u$ and $\mu_s$. First and second-order transitions are indicated
538: by solid and dotted lines, respectively. The letters denote the
539: different phases, where a: $\bar u u$ + $\bar d d$ + $\bar s s$,
540: b: $\bar u u$ + $\bar d d$,
541: d: $\bar d d$ + $\bar s s$,
542: h: $\pi^+ / \pi^-$, i: $\pi^+ / \pi^-$ + $\bar s s$,
543: k: $K^+/ K^-$ + $\bar d d$,
544: l: $K^0 / \bar K^0$,
545: m: $K^0/\bar K^0$ + $\bar u u$
546: and r: 2SCus + $\bar d d$.}
547: \label{fig:muumust0}
548: \end{figure}
549:
550: \subsection{$\mu_u \approx \mu_d$}
551:
552: In Fig.\ \ref{fig:muudmust0}, we display the phase diagram for $T=0$
553: as a function of the up and down quark chemical potential and the
554: strange quark chemical potential. We have chosen $\mu_u = \mu_d +
555: \epsilon$ where $\epsilon$ is a very small positive
556: number. This $\epsilon$ is necessary because when $\epsilon = 0$ one
557: is just at a first-order phase boundary between the phase in which the charged kaons condense
558: (k) and the one in which the neutral kaons condense (m),
559: as can be seen from Fig.\ \ref{fig:muumust0}. This nonzero value of
560: $\epsilon$ gives rise to a small asymmetry in the phase diagram. If
561: $\epsilon$ is chosen negative, the phases in which the neutral (l)/(m) and the charged (j)/(k) kaon condenses are
562: interchanged.
563:
564: \begin{figure}[htb]
565: \includegraphics{muUDmuST0.eps}
566: \caption{Phase diagram for $T = 0$ as a function of $\mu_u = \mu_d +
567: \epsilon $ and $\mu_s$. First and second-order transitions are
568: indicated by solid and dotted lines, respectively. The letters denote
569: the different phases, where
570: a: $\bar u u$ + $\bar d d$ + $\bar s s$,
571: b: $\bar u u$ + $\bar d d$,
572: j: $K^+ / K^-$,
573: k: $K^+/ K^-$ + $\bar d d$,
574: l: $K^0 / \bar K^0$,
575: m: $K^0/\bar K^0$ + $\bar u u$,
576: n: 2SC,
577: q: 2SC + $\bar s s$ and
578: t: CFL.
579: }
580: \label{fig:muudmust0}
581: \end{figure}
582:
583: Apart from the additional 2SC (n)/(q) and CFL (t) phases, our results agree
584: qualitatively with the three-flavor calculations
585: of Ref.\ \cite{Barducci2005} (see their Fig.\ 7).
586:
587: The authors of Ref.\ \cite{Barducci2005} used different quark masses
588: and a different coupling constant, and in addition employed a form
589: factor to mimic asymptotic freedom. Therefore, one may conclude that
590: the use of such a form factor does not affect the phase diagram
591: qualitatively. We would also like to point out that the phase diagram Fig.\
592: \ref{fig:muudmust0} cannot simply be obtained by a superposition of
593: phase diagrams obtained from a calculation with pseudoscalar
594: condensation, but without superconductivity (such as done in
595: \cite{Barducci2005}), and
596: one with superconductivity, but without pseudoscalar condensation
597: (such as done in \cite{Gastineau2002}).
598: Despite the fact that the two types of phases do
599: not coexist, there is nevertheless competition between them.
600: Figure \ref{fig:muudmust0} shows that the $K^0 / \bar K^0$ (l)/(m)
601: and the $K^+ / K^-$ (j)/(k) phases are separated from the 2SC phase (q) by a
602: first-order transition. This remains the case at finite temperature as
603: is illustrated in Fig.\ \ref{fig:muud550must}. This figure displays
604: the phase diagram as a function of $\mu_s$ and temperature, for fixed
605: $\mu_u = \mu_d = 550\;\mathrm{MeV}$.
606:
607: \begin{figure}[htb]
608: \includegraphics{muUD550muST.eps}
609: \caption{Phase diagram as a
610: function of $\mu_s$ and $T$, for fixed
611: $\mu_u = \mu_d = 550\;\mathrm{MeV}$.
612: First and second-order
613: transitions are indicated
614: by solid and dotted lines respectively, while
615: cross-overs are indicated by
616: dashed-dotted lines. The uppermost phase without label is the
617: restored phase.
618: The letters denote the different phases,
619: where
620: g: $\bar s s$,
621: l: $K^0 / \bar K^0$,
622: n: 2SC,
623: q: 2SC + $\bar s s$ and
624: t: CFL.
625: }
626: \label{fig:muud550must}
627: \end{figure}
628:
629: Returning to the discussion of Fig.\ \ref{fig:muudmust0}; the line
630: $\mu_u = \mu_d = \mu_s$ goes through the phase (a) in which chiral
631: symmetry is spontaneously broken. At some point it enters via a
632: first-order transition the 2SC + $\bar s s$ phase (q), and finally
633: goes into the CFL phase (t), again via a first-order transition. If
634: there is a difference between $\mu_u = \mu_d$ and $\mu_s$, one can see
635: from Fig.~\ref{fig:muudmust0} that as the densities increase, quark
636: matter can go directly from a phase of spontaneous chiral symmetry
637: breaking (a) to a CFL phase (t) without passing the 2SC phase (q) first. This
638: can also occur in compact stars \cite{Alford2002}. One should keep in
639: mind though that the relation between chemical potential and number density
640: is not linear. For example, at a first-order phase boundary, the
641: number density increases discontinuously. At these particular
642: densities, quark matter can be in a mixed state of normal and
643: superconducting matter \cite{Bedaque2002, Lawley2005}.
644:
645: It is also interesting to note that the phases (l)/(j) of kaon
646: condensation can also occur outside the region of spontaneous chiral
647: symmetry breaking. Assuming the phase transition towards chiral
648: symmetry restoration coincides with the deconfinement transition (as
649: appears to be the case in lattice studies at small baryon chemical
650: potential and in some models), this would imply that condensation of a
651: state with quantum numbers of the kaon may persist in the deconfined
652: phase. This was first observed in Ref.~\cite{Son2001}, based on a perturbative
653: calculation at high isospin chemical potential, that is expected to be
654: applicable only in the deconfined region.
655:
656: \subsection{$\mu_u = \mu_s$}
657:
658: In Fig.\ \ref{fig:muusmudt0}, we show the phase diagram at zero
659: temperature as a function of $\mu_u = \mu_s$ and $\mu_d$. This diagram
660: is similar to Fig.\ \ref{fig:muumudt0} for small strange quark
661: chemical potentials (below the kaon mass).
662: At larger strange quark chemical potentials the
663: diagrams differ, exhibiting kaon condensation (l) and
664: diquark condensation involving strange quarks (r)/(o)/(t).
665:
666: \begin{figure}[htb]
667: \includegraphics{muUSmuDT0.eps}
668: \caption{Phase diagram for $T = 0$ as a function of $\mu_u=\mu_s$ and
669: $\mu_d$. First and second-order transitions are indicated
670: by solid and dotted lines, respectively. The letters denote
671: the different phases, where
672: a: $\bar u u$ + $\bar d d$ + $\bar s s$,
673: c: $\bar u u$ + $\bar s s$,
674: d: $\bar d d$ + $\bar s s$,
675: g: $\bar s s$,
676: i: $\pi^+ / \pi^-$ + $\bar s s$ ,
677: l: $K^0 / \bar K^0$,
678: o: 2SCus,
679: q: 2SC + $\bar s s$,
680: r: 2SCus + $\bar d d$ and
681: t: CFL.}
682: \label{fig:muusmudt0}
683: \end{figure}
684:
685: \begin{figure}[t]
686: \includegraphics{muUS550muDT.eps}
687: \caption{Phase diagram as a function of $\mu_d$ and $T$, for fixed
688: $\mu_u = \mu_s = 550\;\mathrm{MeV}$. First and second-order
689: transitions are indicated
690: by solid and dotted lines, respectively, while cross-overs
691: are denoted by dashed-dotted lines. The uppermost phase without label is the
692: restored phase.
693: The letters
694: denote the different phases, where f: $\bar d d$, h: $\pi^+ / \pi^-$,
695: l: $K^0 / \bar K^0$, o: 2SCus, r: 2SCus + $\bar d d$, t: CFL and w:
696: sSC. The lower right corner of this figure is enlarged in Fig.\
697: \ref{fig:muus550mudtr}.}
698: \label{fig:mus550mudt}
699: \end{figure}
700:
701: \begin{figure}[t]
702: \includegraphics{muUS550muDTrd.eps}
703: \caption{Same as Fig.\ \ref{fig:mus550mudt}. The phases that occur are
704: f: $\bar d d $,
705: n: 2SC,
706: o: 2SCus,
707: p: 2SCds,
708: r: 2SCus + $\bar d d$,
709: t: CFL,
710: u: uSC,
711: v: dSC and
712: w: sSC.}
713: \label{fig:muus550mudtr}
714: \end{figure}
715:
716: In Fig.\ \ref{fig:mus550mudt}, we show the phase diagram as a function
717: of $\mu_d$ and $T$, for fixed $\mu_u = \mu_s = 550\;\mathrm{MeV}$. In
718: this figure one can find five critical points. Also, one can see in
719: this figure that the $K^0 / \bar K^0$ phase (l) is separated from the
720: 2SCus phase (r) by a first-order transition for all temperatures.
721: Furthermore, it is interesting that at finite temperature there is a
722: first-order transition from the phase in which the neutral kaons
723: condense (l) to the pion condensed phase (h). In Fig.\
724: \ref{fig:muus550mudtr}, we have enlarged the lower-right corner of
725: Fig.\ \ref{fig:mus550mudt} for clarity. In this figure one can find
726: all possible superconducting phases, including the more exotic uSC (u), dSC (v) and
727: sSC (w) phases. For $\mu_u = \mu_d = \mu_s$ one goes from the
728: CFL phase (t) via the 2SC phase (n) to the chirally restored
729: phase when raising the temperature. However, small differences
730: between $\mu_u = \mu_s$ and $\mu_d$ can cause one to go through completely
731: different phases.
732:
733: %=====================================================================
734: \section{Conclusions}
735: In this paper, we studied the phase diagram of the three-flavor NJL
736: model including pseudoscalar condensation and color superconductivity
737: as a function of the different quark chemical potentials and
738: temperature. The NJL model has a rich and
739: interesting phase structure. The pseudoscalar condensed and color
740: superconducting phases are competing and are separated by a
741: first-order phase transition. As we have discussed, this need not be
742: the case for other (less conventional) choices of the parameters of
743: the model.
744:
745: Furthermore, we concluded that at zero temperature and zero strange
746: quark chemical potential, there is a minimum asymmetry of about 35 MeV
747: between the up and the down quark chemical potentials required in
748: order to have two first-order transitions, when going from the phase
749: with spontaneous chiral symmetry breaking to the 2SC phase.
750:
751: Our results provide a qualitative check and extension of several earlier
752: calculations that appeared in the literature. The new aspects of the
753: phase diagrams are often located in regions, where the quark chemical
754: potentials are large and very different in magnitude for the
755: different flavors. Although such situations are not necessarily
756: realized in compact stars or can be realized in heavy-ion collisions,
757: a comparison with future lattice data may nevertheless provide
758: interesting information. This is especially relevant for pseudoscalar
759: condensation in the phase where chiral symmetry is restored and also
760: for the complicated superconductivity phase structure close to the
761: cutoff of the model.
762:
763: This work can be extended in several ways. For example, one can take
764: into account 't Hooft's instanton-induced interaction
765: \cite{tHooft1976}. If one has pseudoscalar condensation, this is more
766: difficult than in the normal case. Another useful extension would be the
767: inclusion of the neutrality conditions~\cite{Alford2002, Steiner2002},
768: in which case the phase structure
769: changes and for instance gapless phases will occur.
770: It would also be interesting to
771: see how the results depend on the strength of the diquark coupling and
772: also on the quark masses. Furthermore, one could add the LOFF phase
773: \cite{Alford2001b}. In this crystalline phase, quarks of different
774: momenta can pair. One could also include vector interactions. In this
775: case spin-1 diquark condensation (see for example Refs.\
776: \cite{Pisarski1999, Buballa2002}) and an induced Lorentz-symmetry
777: breaking (ISB) phase~\cite{Langfeld1998} are among the
778: possibilities. It would also be worthwhile to take pseudoscalar diquark
779: condensation \cite{Buballa2004b} into account. Finally,
780: one could try to go beyond the mean-field approximation as was done in
781: Ref.\ \cite{Hufner1994}.
782:
783: \section*{Acknowledgments}
784: The research of D.B.~has been made possible by financial support from
785: the
786: Royal Netherlands Academy of Arts and Sciences.
787:
788: \begin{thebibliography}{99}
789:
790: \bibitem{Karsch2000}
791: F.~Karsch, E.~Laermann and A.~Peikert,
792: %``The pressure in 2, 2+1 and 3 flavour QCD,''
793: Phys.\ Lett.\ B {\bf 478}, 447 (2000).
794: %[arXiv:hep-lat/0002003].
795: %%CITATION = HEP-LAT 0002003;%%
796:
797: \bibitem{Fodor2002b}
798: Z.~Fodor and S.~D.~Katz,
799: %``Lattice determination of the critical point of QCD at finite T and mu,''
800: JHEP {\bf 0203}, 014 (2002).
801: %[arXiv:hep-lat/0106002].
802: %%CITATION = HEP-LAT 0106002;%%
803:
804: \bibitem{deForcrand2002}
805: P.~de Forcrand and O.~Philipsen,
806: %``The QCD phase diagram for small densities from imaginary chemical
807: %potential,''
808: Nucl.\ Phys.\ B {\bf 642}, 290 (2002).
809: %[arXiv:hep-lat/0205016].
810: %%CITATION = HEP-LAT 0205016;%%
811:
812: \bibitem{Allton2002}
813: C.~R.~Allton {\it et al.},
814: %``The QCD thermal phase transition in the presence of a small chemical
815: %potential,''
816: Phys.\ Rev.\ D {\bf 66}, 074507 (2002).
817: %[arXiv:hep-lat/0204010].
818: %%CITATION = HEP-LAT 0204010;%%
819:
820: \bibitem{DElia2003}
821: M.~D'Elia and M.~P.~Lombardo,
822: %``Finite density QCD via imaginary chemical potential,''
823: Phys.\ Rev.\ D {\bf 67}, 014505 (2003).
824: %[arXiv:hep-lat/0209146].
825: %%CITATION = HEP-LAT 0209146;%%
826:
827: \bibitem{Halasz1998}
828: M.~A.~Halasz, A.~D.~Jackson, R.~E.~Shrock, M.~A.~Stephanov and J.~J.~M.~Verbaarschot,
829: %``On the phase diagram of {QCD},''
830: Phys.\ Rev.\ D {\bf 58}, 096007 (1998).
831: %[arXiv:hep-ph/9804290].
832: %%CITATION = HEP-PH 9804290;%%
833:
834: \bibitem{Berges1998}
835: J.~Berges and K.~Rajagopal,
836: %``Color superconductivity and chiral symmetry restoration at nonzero baryon
837: %density and temperature,''
838: Nucl.\ Phys.\ B {\bf 538}, 215 (1999).
839: %[arXiv:hep-ph/9804233].
840: %%CITATION = HEP-PH 9804233;%%
841:
842: \bibitem{Bailin1984}
843: D.~Bailin and A.~Love,
844: %``Superfluidity And Superconductivity In Relativistic Fermion Systems,''
845: Phys.\ Rept.\ {\bf 107}, 325 (1984).
846: %%CITATION = PRPLC,107,325;%%
847:
848: \bibitem{Alford1997}
849: M.~G.~Alford, K.~Rajagopal and F.~Wilczek,
850: %``QCD at finite baryon density: Nucleon droplets and color
851: %superconductivity,''
852: Phys.\ Lett.\ B {\bf 422}, 247 (1998).
853: %[arXiv:hep-ph/9711395].
854: %%CITATION = HEP-PH 9711395;%%
855:
856: \bibitem{Rapp1997}
857: R.~Rapp, T.~Sch\"afer, E.~V.~Shuryak and M.~Velkovsky,
858: %``Diquark Bose condensates in high density matter and instantons,''
859: Phys.\ Rev.\ Lett.\ {\bf 81}, 53 (1998).
860: %[arXiv:hep-ph/9711396].
861: %%CITATION = HEP-PH 9711396;%%
862:
863: \bibitem{Alford1999}
864: M.~G.~Alford, K.~Rajagopal and F.~Wilczek,
865: %``Color-flavor locking and chiral symmetry breaking in high density {QCD},''
866: Nucl.\ Phys.\ B {\bf 537}, 443 (1999).
867: %[arXiv:hep-ph/9804403].
868: %%CITATION = HEP-PH 9804403;%%
869:
870: \bibitem{Alford1999q}
871: M.~G.~Alford, J.~Berges and K.~Rajagopal,
872: %``Unlocking color and flavor in superconducting strange quark matter,''
873: Nucl.\ Phys.\ B {\bf 558}, 219 (1999).
874: %[arXiv:hep-ph/9903502].
875: %%CITATION = HEP-PH 9903502;%%
876:
877: \bibitem{Alford2002}
878: M.~Alford and K.~Rajagopal,
879: %``Absence of two-flavor color superconductivity in compact stars,''
880: JHEP {\bf 0206}, 031 (2002).
881: %[arXiv:hep-ph/0204001].
882: %%CITATION = HEP-PH 0204001;%%
883:
884: \bibitem{Steiner2002}
885: A.~W.~Steiner, S.~Reddy and M.~Prakash,
886: %``Color-neutral superconducting quark matter,''
887: Phys.\ Rev.\ D {\bf 66}, 094007 (2002).
888: %[arXiv:hep-ph/0205201].
889: %%CITATION = HEP-PH 0205201;%%
890:
891: \bibitem{Neumann2002}
892: F.~Neumann, M.~Buballa and M.~Oertel,
893: %``Mixed phases of color superconducting quark matter,''
894: Nucl.\ Phys.\ A {\bf 714}, 481 (2003).
895: %[arXiv:hep-ph/0210078].
896: %%CITATION = HEP-PH 0210078;%%
897:
898: \bibitem{Alford1999b}
899: M.~G.~Alford, J.~Berges and K.~Rajagopal,
900: %``Gapless color superconductivity,''
901: Phys.\ Rev.\ Lett.\ {\bf 84}, 598 (2000).
902: %[arXiv:hep-ph/9908235].
903: %%CITATION = HEP-PH 9908235;%%
904:
905: \bibitem{Shovkovy2003}
906: I.~Shovkovy and M.~Huang,
907: %``Gapless two-flavor color superconductor,''
908: Phys.\ Lett.\ B {\bf 564}, 205 (2003).
909: %[arXiv:hep-ph/0302142].
910: %%CITATION = HEP-PH 0302142;%%
911:
912: \bibitem{Alford2004a}
913: M.~Alford, C.~Kouvaris and K.~Rajagopal,
914: %``Gapless color-flavor-locked quark matter,''
915: Phys.\ Rev.\ Lett.\ {\bf 92}, 222001 (2004).
916: %[arXiv:hep-ph/0311286].
917: %%CITATION = HEP-PH 0311286;%%
918:
919: \bibitem{Alford2004c}
920: M.~Alford, C.~Kouvaris and K.~Rajagopal,
921: %``Evaluating the gapless color-flavor locked phase,''
922: Phys.\ Rev.\ D {\bf 71}, 054009 (2005).
923: %[arXiv:hep-ph/0406137].
924: %%CITATION = HEP-PH 0406137;%%
925:
926: \bibitem{Ruster2004}
927: S.~B.~R\"uster, I.~A.~Shovkovy and D.~H.~Rischke,
928: %``Phase diagram of dense neutral three-flavor quark matter,''
929: Nucl.\ Phys.\ A {\bf 743}, 127 (2004).
930: %[arXiv:hep-ph/0405170].
931: %%CITATION = HEP-PH 0405170;%%
932:
933: \bibitem{Abuki2004}
934: H.~Abuki, M.~Kitazawa and T.~Kunihiro,
935: %``How does the dynamical chiral condensation affect the three-flavor neutral
936: %quark matter?,''
937: arXiv:hep-ph/0412382.
938: %%CITATION = HEP-PH 0412382;%%
939:
940: \bibitem{Ruster2005}
941: S.~B.~R\"uster, V.~Werth, M.~Buballa, I.~A.~Shovkovy and D.~H.~Rischke,
942: %``The phase diagram of neutral quark matter: Self-consistent treatment of quark
943: %masses,''
944: arXiv:hep-ph/0503184.
945: %%CITATION = HEP-PH 0503184;%%
946:
947: \bibitem{Blaschke2005}
948: D.~Blaschke, S.~Fredriksson, H.~Grigorian, A.~M.~\"Oztas and F.~Sandin,
949: %``The phase diagram of three-flavor quark matter under compact star
950: %constraints,''
951: arXiv:hep-ph/0503194.
952: %%CITATION = HEP-PH 0503194;%%
953:
954: \bibitem{Klein2003}
955: B.~Klein, D.~Toublan and J.~J.~M.~Verbaarschot,
956: %``The QCD phase diagram at nonzero temperature, baryon and isospin chemical
957: %potentials in random matrix theory,''
958: Phys.\ Rev.\ D {\bf 68}, 014009 (2003).
959: %[arXiv:hep-ph/0301143].
960: %%CITATION = HEP-PH 0301143;%%
961:
962: \bibitem{Toublan2003}
963: D.~Toublan and J.~B.~Kogut,
964: %``Isospin chemical potential and the QCD phase diagram at nonzero temperature
965: %and baryon chemical potential,''
966: Phys.\ Lett.\ B {\bf 564}, 212 (2003).
967: %[arXiv:hep-ph/0301183].
968: %%CITATION = HEP-PH 0301183;%%
969:
970: \bibitem{Frank2003}
971: M.~Frank, M.~Buballa and M.~Oertel,
972: %``Flavor-mixing effects on the QCD phase diagram at non-vanishing isospin
973: %chemical potential: One or two phase transitions?,''
974: Phys.\ Lett.\ B {\bf 562}, 221 (2003).
975: %[arXiv:hep-ph/0303109].
976: %%CITATION = HEP-PH 0303109;%%
977:
978: \bibitem{Son2001}
979: D.~T.~Son and M.~A.~Stephanov,
980: %``QCD at finite isospin density,''
981: Phys.\ Rev.\ Lett.\ {\bf 86}, 592 (2001).
982: %[arXiv:hep-ph/0005225].
983: %%CITATION = HEP-PH 0005225;%%
984:
985: \bibitem{Kogut2001}
986: J.~B.~Kogut and D.~Toublan,
987: %``QCD at small non-zero quark chemical potentials,''
988: Phys.\ Rev.\ D {\bf 64}, 034007 (2001).
989: %[arXiv:hep-ph/0103271].
990: %%CITATION = HEP-PH 0103271;%%
991:
992: \bibitem{Barducci2003}
993: A.~Barducci, G.~Pettini, L.~Ravagli and R.~Casalbuoni,
994: %``Ladder-QCD at finite isospin chemical potential,''
995: Phys.\ Lett.\ B {\bf 564}, 217 (2003).
996: %[arXiv:hep-ph/0304019].
997: %%CITATION = HEP-PH 0304019;%%
998:
999: \bibitem{Kogut2002}
1000: J.~B.~Kogut and D.~K.~Sinclair,
1001: %``Lattice QCD at finite isospin density at zero and finite temperature,''
1002: Phys.\ Rev.\ D {\bf 66}, 034505 (2002).
1003: %[arXiv:hep-lat/0202028].
1004: %%CITATION = HEP-LAT 0202028;%%
1005:
1006: \bibitem{Barducci2004}
1007: A.~Barducci, R.~Casalbuoni, G.~Pettini and L.~Ravagli,
1008: %``A calculation of the QCD phase diagram at finite temperature, and baryon and
1009: %isospin chemical potentials,''
1010: Phys.\ Rev.\ D {\bf 69}, 096004 (2004).
1011: %[arXiv:hep-ph/0402104].
1012: %%CITATION = HEP-PH 0402104;%%
1013:
1014: \bibitem{Barducci2005}
1015: A.~Barducci, R.~Casalbuoni, G.~Pettini and L.~Ravagli,
1016: %``Pion and kaon condensation in a 3-flavor NJL model,''
1017: Phys.\ Rev.\ D {\bf 71}, 016011 (2005).
1018: %[arXiv:hep-ph/0410250].
1019: %%CITATION = HEP-PH 0410250;%%
1020:
1021: \bibitem{Blaschke2003}
1022: D.~Blaschke, M.~K.~Volkov and V.~L.~Yudichev,
1023: %``Coexistence of color superconductivity and chiral symmetry breaking within
1024: %the NJL model,''
1025: Eur.\ Phys.\ J.\ A {\bf 17}, 103 (2003).
1026: %[arXiv:hep-ph/0301065].
1027: %%CITATION = HEP-PH 0301065;%%
1028:
1029: \bibitem{Buballa2004b}
1030: M.~Buballa,
1031: %``NJL-model description of Goldstone boson condensation in the color-flavor
1032: %locked phase,''
1033: Phys.\ Lett.\ B {\bf 609}, 57 (2005).
1034: %[arXiv:hep-ph/0410397].
1035: %%CITATION = HEP-PH 0410397;%%
1036:
1037: \bibitem{Schafer2000}
1038: T.~Sch\"afer,
1039: %``Kaon condensation in high density quark matter,''
1040: Phys.\ Rev.\ Lett.\ {\bf 85}, 5531 (2000).
1041: %[arXiv:nucl-th/0007021].
1042: %%CITATION = NUCL-TH 0007021;%%
1043:
1044: \bibitem{Bedaque2002a}
1045: P.~F.~Bedaque and T.~Sch\"afer,
1046: %``High density quark matter under stress,''
1047: Nucl.\ Phys.\ A {\bf 697}, 802 (2002).
1048: %[arXiv:hep-ph/0105150].
1049: %%CITATION = HEP-PH 0105150;%%
1050:
1051: \bibitem{Kaplan2002}
1052: D.~B.~Kaplan and S.~Reddy,
1053: %``Novel phases and transitions in quark matter,''
1054: Phys.\ Rev.\ D {\bf 65}, 054042 (2002).
1055: %[arXiv:hep-ph/0107265].
1056: %%CITATION = HEP-PH 0107265;%%
1057:
1058: \bibitem{Casalbuoni1999}
1059: R.~Casalbuoni and R.~Gatto,
1060: %``Effective theory for color-flavor locking in high density QCD,''
1061: Phys.\ Lett.\ B {\bf 464}, 111 (1999).
1062: %[arXiv:hep-ph/9908227].
1063: %%CITATION = HEP-PH 9908227;%%
1064:
1065: \bibitem{Son1999}
1066: D.~T.~Son and M.~A.~Stephanov,
1067: %``Inverse meson mass ordering in color-flavor-locking phase of high density
1068: %QCD,''
1069: Phys.\ Rev.\ D {\bf 61}, 074012 (2000).
1070: %[arXiv:hep-ph/9910491].
1071: %%CITATION = HEP-PH 9910491;%%
1072:
1073: \bibitem{Casalbuoni2000}
1074: R.~Casalbuoni, Z.~y.~Duan and F.~Sannino,
1075: %``Low energy theory for 2 flavors at high density QCD,''
1076: Phys.\ Rev.\ D {\bf 62}, 094004 (2000).
1077: %[arXiv:hep-ph/0004207].
1078: %%CITATION = HEP-PH 0004207;%%
1079:
1080: \bibitem{Forbes2004}
1081: M.~M.~Forbes,
1082: %``Kaon condensation in an NJL model at high density,''
1083: arXiv:hep-ph/0411001.
1084: %%CITATION = HEP-PH 0411001;%%
1085:
1086: \bibitem{Nambu1961}
1087: Y.~Nambu and G.~Jona-Lasinio,
1088: %``Dynamical Model Of Elementary Particles Based On An Analogy With
1089: %Superconductivity. I,''
1090: Phys.\ Rev.\ {\bf 122}, 345 (1961).
1091: %%CITATION = PHRVA,122,345;%%
1092:
1093: %\cite{Klevansky:1992qe}
1094: \bibitem{Klevansky1992}
1095: S.~P.~Klevansky,
1096: %``The Nambu-Jona-Lasinio model of quantum chromodynamics,''
1097: Rev.\ Mod.\ Phys.\ {\bf 64}, 649 (1992).
1098: %%CITATION = RMPHA,64,649;%%
1099:
1100: \bibitem{Buballa2004a}
1101: M.~Buballa,
1102: %``NJL model analysis of quark matter at large density,''
1103: Phys.\ Rept.\ {\bf 407}, 205 (2005).
1104: %[arXiv:hep-ph/0402234].
1105: %%CITATION = HEP-PH 0402234;%%
1106:
1107: \bibitem{Alford2001b}
1108: M.~G.~Alford, J.~A.~Bowers and K.~Rajagopal,
1109: %``Crystalline color superconductivity,''
1110: Phys.\ Rev.\ D {\bf 63}, 074016 (2001).
1111: %[arXiv:hep-ph/0008208].
1112: %%CITATION = HEP-PH 0008208;%%
1113:
1114: \bibitem{Weingarten1983}
1115: D.~Weingarten,
1116: %``Mass Inequalities For QCD,''
1117: Phys.\ Rev.\ Lett.\ {\bf 51}, 1830 (1983).
1118: %%CITATION = PRLTA,51,1830;%%
1119:
1120: \bibitem{Lapack1999}
1121: E.~Anderson et.\ al$.$,
1122: LAPACK User's Guide, Third Edition. SIAM, Philadelphia, (1999).
1123:
1124: \bibitem{James1975}
1125: F.~James and M.~Roos,
1126: %``'Minuit' A System For Function Minimization And Analysis Of The Parameter
1127: %Errors And Correlations,''
1128: Comput.\ Phys.\ Commun.\ {\bf 10}, 343 (1975).
1129: %%CITATION = CPHCB,10,343;%%
1130:
1131: \bibitem{Bedaque2002}
1132: P.~F.~Bedaque,
1133: %``Color superconductivity in asymmetric matter,''
1134: Nucl.\ Phys.\ A {\bf 697}, 569 (2002).
1135: %[arXiv:hep-ph/9910247].
1136: %%CITATION = HEP-PH 9910247;%%
1137:
1138: \bibitem{He2005}
1139: L.~He, M.~Jin and P.~Zhuang,
1140: %``Pion superfluidity and meson properties at finite isospin density,''
1141: arXiv:hep-ph/0503272.
1142: %%CITATION = HEP-PH 0503272;%%
1143:
1144: \bibitem{Gastineau2002}
1145: F.~Gastineau, R.~Nebauer and J.~Aichelin,
1146: %``Thermodynamics of the three-flavor Nambu-Jona-Lasinio model: Chiral symmetry
1147: %breaking and color superconductivity,''
1148: Phys.\ Rev.\ C {\bf 65}, 045204 (2002).
1149: %[arXiv:hep-ph/0101289].
1150: %%CITATION = HEP-PH 0101289;%%
1151:
1152: \bibitem{Lawley2005}
1153: S.~Lawley, W.~Bentz and A.~W.~Thomas,
1154: %``The phases of isospin asymmetric matter in the two flavor NJL model,''
1155: arXiv:nucl-th/0504020.
1156: %%CITATION = NUCL-TH 0504020;%%
1157:
1158: \bibitem{tHooft1976}
1159: G.~'t Hooft,
1160: %``Computation Of The Quantum Effects Due To A Four-Dimensional
1161: %Pseudoparticle,''
1162: Phys.\ Rev.\ D {\bf 14}, 3432 (1976).
1163: %[Erratum-ibid.\ D {\bf 18}, 2199 (1978)].
1164: %%CITATION = PHRVA,D14,3432;%%
1165:
1166: \bibitem{Pisarski1999}
1167: R.~D.~Pisarski and D.~H.~Rischke,
1168: %``Color superconductivity in weak coupling,''
1169: Phys.\ Rev.\ D {\bf 61}, 074017 (2000).
1170: %[arXiv:nucl-th/9910056].
1171: %%CITATION = NUCL-TH 9910056;%%
1172:
1173: \bibitem{Buballa2002}
1174: M.~Buballa, J.~Hosek and M.~Oertel,
1175: %``Anisotropic admixture in color-superconducting quark matter,''
1176: Phys.\ Rev.\ Lett.\ {\bf 90}, 182002 (2003).
1177: %[arXiv:hep-ph/0204275].
1178: %%CITATION = HEP-PH 0204275;%%
1179:
1180: \bibitem{Langfeld1998}
1181: K.~Langfeld and M.~Rho,
1182: %``Quark condensation, induced symmetry breaking and color superconductivity at
1183: %high density,''
1184: Nucl.\ Phys.\ A {\bf 660}, 475 (1999).
1185: %[arXiv:hep-ph/9811227].
1186: %%CITATION = HEP-PH 9811227;%%
1187:
1188: \bibitem{Hufner1994}
1189: J.~H\"ufner, S.~P.~Klevansky, P.~Zhuang and H.~Voss,
1190: %``Thermodynamics of a quark plasma beyond the mean field: A generalized
1191: %Beth-Uhlenbeck approach,''
1192: Annals Phys.\ {\bf 234}, 225 (1994).
1193: %%CITATION = APNYA,234,225;%%
1194:
1195: \end{thebibliography}
1196:
1197: \end{document}
1198:
1199: