1: \documentclass[nofootinbib,twocolumn]{revtex4}
2: \usepackage{graphicx}
3: \usepackage{latexsym}
4: \def\be{\begin{equation}}
5: \def\ee{\end{equation}}
6: \def\bea{\begin{eqnarray}}
7: \def\eea{\end{eqnarray}}
8:
9: \begin{document}
10: \title{A Single Scalar Field Model of Dark Energy with Equation of
11: State Crossing $-1$}
12: \author{Mingzhe Li}
13: \email{m.li@thphys.uni-heidelberg.de}
14: \affiliation{Institut f\"{u}r Theoretische Physik,
15: Philosophenweg 16, 69120 Heidelberg, Germany}
16: \author{Bo Feng}
17: \email{fengbo@mail.ihep.ac.cn}
18: \author{Xinmin Zhang}
19: \email{xmzhang@mail.ihep.ac.cn} \affiliation{Institute of High
20: Energy Physics, Chinese Academy of Sciences, P.O. Box 918-4, Beijing
21: 100049, People's Republic of China}
22:
23:
24: \begin{abstract}
25: In this paper we study the possibility of building models of dark
26: energy with equation of state across $-1$ and propose explicitly a
27: model with a single scalar field which gives rise to an equation of
28: state larger than $-1$ in the past and less than $-1$ at the present
29: time, consistent with the current observations.
30: \end{abstract}
31:
32: \maketitle
33:
34: \hskip 1.6cm PACS number: 98.80.Cq \vskip 0.4cm
35:
36:
37: Since two groups \cite{Riess98,Perl99} independently showed the
38: accelerating expansion of our universe based on the Type Ia
39: Supernova (SN) observations of the redshift-distance relations in
40: 1998, numerous observations and experiments consistently indicate
41: that our universe is dominated by a component with a negative
42: pressure, dubbed dark energy in the literature. Especially the first
43: year Wilkinson Microwave Anisotropy Probe (WMAP) data
44: \cite{Spergel03} strongly support for a concordance model in which
45: about $73\%$ of the energy of the universe is indwelt in the dark
46: energy ingredient. One essential to understand the nature of the
47: dark energy is to detect the evolution of its equation of state
48: $w=p/\rho$, where $p$ and $\rho$ are the pressure and the energy
49: density of the dark energy respectively. Recent analysis on the data
50: from the Supernova, cosmic microwave background (CMB) and large
51: scale structure (LSS) show that the cosmological constant fits well
52: to the data \cite{seljak}, however current data also mildly favor an
53: evolving dark energy with an equation of state $w$ larger than $-1$
54: in the past and less than $-1$ today \cite{starobinsky,ex} and evolving across
55: $-1$ in the intermediate redshift. If such a result holds on with
56: the accumulation of observational data, this would be a great
57: challenge to the current models of dark energy. Firstly, the
58: cosmological constant as a candidate for dark energy will be
59: excluded and dark energy must be dynamical. Secondly, the simple
60: dynamical dark energy models considered vastly in the literature
61: like the quintessence or the phantom can not be satisfied either.
62:
63: In the quintessence models
64: \cite{wetterich,wetterich2,peebles,steinhardt}, $w$ is evolving
65: however always in the range of $-1 \leq w \leq 1$. In the phantom
66: models \cite{phantom}, one has $w \leq -1$. For the general
67: k-essence models $\mathcal{L}=p(\phi,~X)$ with $X=1/2
68: \nabla_{\mu}\phi\nabla^{\mu}\phi$, and $\nabla_{\mu}$ the covariant
69: derivative, it was shown that it also fails to cross the boundary of
70: $w=-1$ because of the problems of singularity and classical
71: instability in the perturbative equations \cite{zhao,vikman}. Thus
72: it is interesting to ask if theoretically there exists model of a
73: single scalar field which has a $w$ crossing $-1$ during the
74: evolution of the universe \footnote{The dark energy models with the
75: property of crossing the cosmological constant boundary are called
76: quintom models \cite{quintom}, and there are a lot of interests in
77: the phenomenological study of this class of models recently in the
78: literature
79: \cite{wei,vikman,zhang,guo,hu,xia,michael,li,cai,xfzhang}.}.
80: In this paper we propose a toy model of this type.
81:
82: In the general k-essence models, the Lagrangian is a function of the
83: field $\phi$ and its first derivative $\nabla_{\mu}\phi$. In this
84: paper we will propose a dark energy model which includes higher
85: derivative operators in the Lagrangian, for example, \be
86: \mathcal{L}=\mathcal{L}(\phi,~X,~\Box\phi\Box\phi,~\nabla_{\mu}\nabla_{\nu}\phi\nabla^{\mu}\nabla^{\nu}\phi)~,
87: \ee where $\Box\equiv \nabla_{\mu}\nabla^{\mu}$ is the d'Alembertian
88: operator. The higher derivative terms in the effective Lagrangian
89: above may be derived from fundamental theories. In fact it has been
90: shown in the literature that this type of operators does appear as
91: the quantum corrections or due to the non-local physics in the
92: string theory \cite{simon,woodard,gross}. With the higher derivative
93: terms to the Einstein gravity, the theory is shown to be
94: renormalizable \cite{stelle} which has attracted many attentions. Recently, higher derivative operators
95: have been considered to stabilize the linear fluctuations in the scenario of ``ghost condensation" \cite{ghost}. In
96: addition the canonical form for the higher derivative theory has
97: been put forward by Ostrogradski about one and a half century ago
98: \cite{ostrogradski}. In short, it is interesting and worthwhile to
99: study the implications of models with higher derivatives in
100: cosmology.
101:
102: In this paper we will show this type of theories provides a class of
103: models of dark energy where the equation of state can get across
104: $-1$ during evolution, consistent with the current observations. We
105: consider the simple model with the Lagrangian given by \footnote{We
106: take the convention for the sign of the metric with $(+,~-,~-,~-)$
107: .} \be\label{lagrangian} \mathcal{L}=-{1\over
108: 2}\nabla_{\mu}\phi\nabla^{\mu}\phi+{c\over
109: 2M^2}\Box\phi\Box\phi-V(\phi)~, \ee where $M$ is a constant with
110: mass dimension and $c$ is a dimensionless constant. One can see that
111: this model differs from the phantom model \cite{phantom} due to the
112: dimension-6 operator. Our analysis below will show that this model
113: gives rise to an equation of state which transit from $w>-1$ to
114: $w<-1$ as the redshift decreases, furthermore it is equivalent to
115: the two scalar fields model of the quintom dark energy studied in
116: Ref.\cite{quintom,guo,xfzhang}.
117:
118:
119: Given the Lagrangian (\ref{lagrangian}) we have the
120: energy-momentum tensor:
121: \bea\label{stress} T^{\mu\nu}&=&[{1\over
122: 2}\nabla_{\rho}\phi\nabla^{\rho}\phi+{c\over
123: 2M^2}\Box\phi\Box\phi+{c\over
124: M^2}\nabla^{\rho}\phi\nabla_{\rho}(\Box\phi)\nonumber\\ & &+V]g^{\mu\nu}
125: -\nabla^{\mu}\phi\nabla^{\nu}\phi- {c\over
126: M^2}\nabla^{\nu}\phi\nabla^{\mu}(\Box\phi)\nonumber\\ & &- {c\over
127: M^2}\nabla^{\mu}\phi\nabla^{\nu}(\Box\phi), \eea and the equation
128: of motion: \be\label{eqm} -\Box\phi-{c\over
129: M^2}\Box^2\phi+\frac{dV}{d\phi}=0~. \ee
130: The energy-momentum tensor (\ref{stress}) can be rewritten as
131: \bea\label{astress} T^{\mu\nu}&=&{1\over
132: 2}[\nabla_{\rho}\psi\nabla^{\rho}\psi-\nabla_{\rho}\chi\nabla^{\rho}\chi
133: +2V(\psi-\chi)+\nonumber\\
134: & & {M^2\over c}\chi^2]g^{\mu\nu}
135: -\nabla^{\mu}\psi\nabla^{\nu}\psi+\nabla^{\mu}\chi\nabla^{\nu}\chi~,
136: \eea where $\chi$ and $\psi$ are defined by
137: \bea
138: \chi &=&\frac{c}{M^2}\Box\phi\label{change}~,\\
139: \psi &=&\phi+\chi~. \eea
140: It is not difficult to see that
141: the energy-momentum tensor
142: (\ref{astress}) can be derived from the following Lagrangian
143: \be\label{alagrangian} \mathcal{L}= -{1\over
144: 2}\nabla_{\mu}\psi\nabla^{\mu}\psi+{1\over
145: 2}\nabla_{\mu}\chi\nabla^{\mu}\chi -V(\psi-\chi)-{M^2\over
146: 2c}\chi^2~, \ee
147: with $\psi$ and $\chi$ being two independent
148: fields. The variations of the Lagrangian in (\ref{alagrangian}) with
149: respect to the fields $\psi$
150: and $\chi$ respectively give rise to the following equations of motions,
151: \bea
152: & &\Box\psi-V'=0~,\\
153: & &\Box\chi+\frac{M^2}{c}\chi-V'=0~, \eea where the prime denotes the
154: derivative with respect to $\psi-\chi$.
155:
156: When we take $\psi =\phi+\chi$ in (8) and (9),
157: we recover the equation of motion (\ref{eqm}) of the single field
158: model and the transformation equation (\ref{change}). So, one can
159: see from the derivations above that our single field model
160: (\ref{lagrangian}) is equivalent to the one with two fields (\ref{alagrangian}) where one is the
161: normal (quintessence-like) scalar field $\chi$ and another is the ghost
162: (phantom-like) field $\psi$.
163: In general, the normal mode $\chi$ gives rise to the equation of
164: state $w_1 \geq -1$ and the ghost mode has $w_2 \leq -1$,
165: consequently the energy density in $\chi$ would decrease and that in
166: $\psi$ would increase with the expansion of the universe. Assuming
167: $\chi$ dominates the dark energy sector in the early time, the
168: equation of state for the system will be larger than $-1$ during
169: that period of time. When the ghost mode becomes dominant over the
170: normal one (and will continue to be dominant among all the
171: components in the universe), the equation of state of the system
172: will become $<-1$ and cross the cosmological constant boundary at a
173: intermediate redshift. Quantitatively the moment when the equation
174: of state crosses $-1$ depends on the values of the model parameters.
175:
176: Now we study the perturbations of our model.
177: As is well known, in the usual quintessence and phantom models,
178: the field and metric perturbations are stable
179: on the small length scales. For both the quintessence and the phantom
180: models the sound speed as indicated in Ref. \cite{kinflation} is given by
181: \be
182: c_s^2= \frac{\partial p/ \partial X}{\partial \rho/ \partial X}=1~,
183: \ee
184: where
185: $X\equiv (1/2)\nabla_{\mu}\phi\nabla^{\mu}\phi$.
186: This can be seen clearly from the dispersion relations. For the phantom
187: field, the frequency $\omega$ relates to the momentum $\textbf{k}$
188: in the same way as in the case of the quintessence,
189: $\omega^2=\textbf{k}^2+...$. And the sound speed is often defined as
190: $\omega^2=c_s^2\textbf{k}^2+...$.
191: Since we have demonstrated that our model is equivalent to a model with
192: a quintessence-like and a phantom-like scalar fields, the general
193: arguments for the quintessence and the phantom hold here, which indicate
194: there will be no exponentially
195: instabilities of the perturbations on small scales in our model.
196:
197: In general the potential term $V(\psi-\chi)$ (equivalently $V(\phi)$
198: in our single field model) should include the interactions between
199: the two fields $\psi$ and $\chi$. For some specific choices of the
200: potential $V$, however these two modes could decouple. As an example
201: we consider $V=(1/2)m^2\phi^2$. Then the Lagrangian
202: (\ref{alagrangian}) can be ``diagonalized" as
203: \be\label{aalagrangian} \mathcal{L}={1\over
204: 2}\nabla_{\mu}\phi_1\nabla^{\mu}\phi_1 -{1\over
205: 2}\nabla_{\mu}\phi_2\nabla^{\mu}\phi_2-{1\over 2}m_1^2\phi_1^2
206: -{1\over 2}m_2^2\phi_2^2~, \ee through the transformation \be \left(
207: \begin{array}{c} \psi \\ \chi\end{array} \right)
208: =\left(\begin{array} {cc} -a_1 & a_2\\-a_2 & a_1\end{array}\right)
209: \left( \begin{array}{c} \phi_1\\ \phi_2
210: \end{array} \right)~.
211: \ee In (12) and (13) \bea & & a_1={1\over 2}(1+\frac{4c
212: m^2}{M^2})^{-1/4}
213: (\sqrt{1+\frac{4c m^2}{M^2}}-1)~,\nonumber\\
214: & & a_2={1\over 2}(1+\frac{4c m^2}{M^2})^{-1/4}(\sqrt{1+\frac{4c m^2}{M^2}}+1)~,
215: \eea
216: and
217: \bea\label{mass}
218: & &m_1^2=\frac{M^2}{2c}(\sqrt{1+\frac{4cm^2}{M^2}}+ 1)~,\nonumber\\
219: & &m_2^2=\frac{M^2}{2c}(\sqrt{1+\frac{4cm^2}{M^2}}- 1)~.
220: \eea
221: Hence, one can see that our model
222: (\ref{lagrangian}) with $V=(1/2)m^2\phi^2$ is
223: equivalent to the uncoupled system
224: (\ref{aalagrangian}). The two modes
225: $\phi_1$ and $\phi_2$ evolve independently in the
226: universe. The positivity of the parameter $c$
227: guarantees the positivity of $m^2_1$ ($m^2_2$ is always positive
228: as long as $1+\frac{4c m^2}{M^2}>0$). In the limit of $m\ll M$, the
229: masses of the two modes are approximately $m_1\simeq M/\sqrt{c}$ and
230: $m_2\simeq m$.
231: In fact, $\phi_1$ and $\phi_2$ are the eigenfunctions of the
232: d'Alembertian operator, $\Box$, with the
233: eigenvalues $-m_1^2$ and $m_2^2$ respectively. The solution to the equation of motion
234: (\ref{eqm}) $\phi$ is decomposed by these eigenfunctions as
235: $\phi=(1+\frac{4c m^2}{M^2})^{-1/4}(\phi_1+\phi_2)$.
236:
237:
238: As we mensioned above, the perturbations in the phantom component $\phi_2$ present no unphysical instabilities in the classical level.
239: In fact, the spatial fluctuation of $\phi_2$ with wavenumber $k>m_2$ is stable, the solution to the equation of motion
240: is oscillatory with time. The time-rising behavior only presents on very large scales, $L>1/m_2$. If the phantom mass $m_2$ of the phantom field is smaller than $H$, there will be no instability within the horizon \cite{hsu}.
241: The rising behaviors of the super-horizon modes (including the background mode, which corresponds to $k=0$) of phantom do not matter in cosmology. This is because the universe is expanding instead of static, generally the components in the universe are evolving. Furthermore, the Hubble expansion provides a friction force which prevents these modes from increasing exponentially. In another word, such an instability is ``benign" \cite{smilga}.
242:
243:
244: In Fig.\ref{equivalence} we plot the equation of state of our model as a function of $\ln a$.
245: One can see clearly $w$ crosses $-1$ during evolution. Furthermore,
246: we made a test on the equivalence of the model (\ref{lagrangian})
247: with the two-field quintom model numerically for $V=(1/2)m^2\phi^2$.
248: If the background spacetime is described by the
249: Friedmann-Robertson-Walker metric,
250: $ds^2=dt^2-a^2(t)\delta_{ij}dx^idx^j$ and the field $\phi$ is
251: homogeneous, the d'Alembertian operator will depend only on the time
252: derivatives, $\Box=\partial^2/\partial t^2+3H\partial/\partial t$.
253: The equation of motion (\ref{eqm}) becomes: \be
254: -\ddot\phi-3H\dot\phi+m^2\phi-{c\over
255: M^2}(\frac{\partial^2}{\partial t^2}+3H\frac{\partial}{\partial
256: t})(\ddot\phi+3H\dot\phi)=0~, \ee where dot represents the
257: derivative with respect to time. The energy density and the pressure
258: are \be\label{energy} \rho=-{1\over 2}\dot\phi^2+{c\over
259: 2M^2}(\ddot\phi+3H\dot\phi)^2-{c\over M^2}\dot\phi\frac{\partial}
260: {\partial t}(\ddot \phi+3H\dot \phi)+{1\over 2}m^2\phi^2~; \ee
261: \be\label{pressure} p=-{1\over 2}\dot\phi^2-{c\over
262: 2M^2}(\ddot\phi+3H\dot\phi)^2-{c\over
263: M^2}\dot\phi\frac{\partial}{\partial t} (\ddot \phi+3H\dot
264: \phi)-{1\over 2}m^2\phi^2~, \ee respectively.
265: \begin{figure}[htbp]
266: %\begin{center}
267: \includegraphics[scale=0.85]{OK.EPS}
268: \caption{\label{equivalence}Plots of the equation of state
269: for our model and the two-field model with one
270: quintessence and one phantom field. In the numerical calculations we
271: choose the initial conditions given
272: by Eq. (\ref{equiv}) and have set $m=2\times 10^{-61} M_{pl}$,
273: $\frac{c}{M^2}=10^{120} M_{pl}^{-2}$. The equivalent initial
274: conditions for the two scalar fields are $\phi_1(0)=3\times
275: 10^{-2}M_{pl}$, $\phi_2(0)=0.18M_{pl}$, and
276: $\dot{\phi_1}(0)=0$, $\dot{\phi_2}(0)=0$.
277: $\Omega_{DE}\approx 0.73$ and $h\approx 0.68$. }.
278: %\end{center}
279: \end{figure}
280: The initial conditions picked up in the numerically calculations are
281: \bea\label{equiv}
282: \phi(0)&=& g[\phi_1(0)+\phi_2(0)]~,\nonumber\\
283: \dot\phi(0)&=& g[\dot{\phi}_1(0)+\dot{\phi_2}(0)]~,\nonumber\\
284: \ddot \phi(0)&=& g[-3H(0)(\dot{\phi}_1(0)+\dot{\phi}_2(0))-m_1^2\phi_1(0)+
285: m_2^2\phi_2(0)],\nonumber\\
286: \dddot \phi(0)&=& g[(9H(0)^2-3\dot H(0))(\dot{\phi}_1(0)+\dot{\phi}_2(0))+\nonumber\\
287: &&3H(0)(m_1^2\phi_1(0)-m_2^2\phi_2(0))-\nonumber\\
288: & &(m_1^2\dot{\phi}_1(0)-m_2^2\dot{\phi}_2(0))]~,\nonumber\\
289: H^2(0)&=& \frac{8\pi G}{3}[\rho_m(0)+\rho_r(0)+{1\over 2}\dot{\phi_1^2}(0)-{1\over 2}
290: \dot{\phi_2^2}(0)
291: +\nonumber\\ & &{1\over 2}m_1^2\phi_1^2(0)+
292: {1\over 2}m_2^2\phi_2^2(0)]~,\nonumber\\
293: \dot H(0)&=&-4\pi G
294: [\rho_m(0)+\frac{4}{3}\rho_r(0)+\dot{\phi_1^2}(0)-\dot{\phi_2^2}(0)]~,
295: \eea where $0$ in the bracket means the initial time, $\rho_m$ and
296: $\rho_r$ are the energy densities of matter and radiation
297: respectively, and $g$ is defined as \be g\equiv (1+\frac{4c
298: m^2}{M^2})^{-1/4}~. \ee We depicte the numerical result for the
299: comparison in Fig.\ref{equivalence}. One can see from this figure
300: the equation of state from (\ref{energy}) and (\ref{pressure}) of
301: our single scalar field model does coincide with that of the
302: two-field model (\ref{aalagrangian}).
303:
304: In summary, we in this paper have proposed a single scalar field
305: model of dark energy which has the property of crossing the
306: cosmological constant boundary. This is achieved by considering
307: higher derivative operators. And this class of model is free from
308: the difficulties of the singularity and the gravitational
309: instabilities in the general k-essence-like models. We should point
310: out that although the energy-momentum tensor of this single scalar
311: field model is equivalent to two independent scalar
312: fields\footnote{After our paper appeared, the authors of
313: Ref.\cite{binf} used the "$\Box$" term only and dropped the
314: conventional kinetic term as well the potential, which cannot be
315: identified with two scalar fields. }, when we consider the
316: interactions they may possibly show some different behaviors. It
317: might be theoretically possible to solve the inherent problems, the
318: quantum instabilities exhibited in the model of
319: phantom\cite{Phtproblms} and is worth studying further.
320:
321: Our study here has the following implications: 1) in the general
322: phenomenological fitting appearing in the literature of
323: the dark energy models
324: to the observational data such as SN, CMB, LSS an issue is how
325: to consistently include the perturbations of the dark energy. The key to
326: this issue is to build theoretical models of dark energy with the equation
327: of state crossing $-1$. Our model proposed in this paper serves as an
328: example for this class of
329: models of dark energy to study the cosmological implications numerically;
330: 2) similar to the k-inflation our model can be
331: applied for inflation. And in this case one would expect non-vanishing isocurvature
332: perturbations even though this model includes explicitly only one inflaton
333: field.
334:
335:
336: {\bf Acknowledgements:} M.L. would like to thank Michael Doran and
337: Christof Wetterich for discussions and to acknowledge the support
338: from the Alexander von Humboldt Foundation. B.F. and X.Z. would like
339: to thank Xiao-Jun Bi, Hong Li and Yun-Song Piao for discussions and
340: are supported in part by the National Natural Science Foundation of
341: China under the grant No. 90303004 and also by the Ministry of
342: Science and Technology of China under grant No. NKBRSF G19990754.
343:
344: \begin{thebibliography}{}
345:
346: \bibitem{Riess98}
347: Riess A G {\it et al.}, 1998 {\it Astron. J.} {\bf 116} 1009 [astro-ph/9805201]
348: \bibitem{Perl99}
349: Perlmutter S {\it et al.}, 1999 {\it Astrophys. J.} {\bf 517} 565 [astro-ph/9812133]
350: \bibitem{Spergel03}
351: Spergel D N {\it et al.}, 2003 {\it Astrophys. J.} Suppl. {\bf 148} 175
352: [astro-ph/0302209]
353: \bibitem{seljak} Seljak U {\it et al.}, 2005 {\it Phys. Rev. D} {\bf 71} 103515 [astro-ph/0407372]
354: \bibitem{starobinsky}
355: Alam U, Sahni V and Starobinsky A A, 2004 JCAP 0406 008 [astro-ph/0403687];
356: Huterer D and Cooray A, 2005 {\it Phys. Rev. D} {\bf 71} 023506 [astro-ph/0404062];
357: Gong Y, 2005 {\it Class. Quant. Grav.} {\bf 22} 2121 [astro-ph/0405446]
358:
359: \bibitem{ex} Corasaniti P S, Kunz M, Parkinson D, Copeland E J
360: and Bassett B A, 2004 {\it Phys. Rev. D} {\bf 70} 083006 [astro-ph/0406608];
361: Hannestad S and Mortsell E, 2004 JCAP {\bf 0409} 001 [astro-ph/0407259];
362: Upadhye A, Ishak M and Steinhardt P J, 2005 {\it Phys. Rev. D} {\bf 72} 063501 [astro-ph/0411803]
363:
364: \bibitem{wetterich} Wetterich C, 1988 {\it Nucl. Phys. B} {\bf 302} 668
365: \bibitem{wetterich2} Wetterich C, 1995 {\it Astron. Astrophys.} {\bf
366: 301} 321 [hep-th/9408025]
367: \bibitem{peebles} Ratra B and Peebles P J E, 1988 {\it Phys. Rev. D} {\bf
368: 37} 3406
369: \bibitem{steinhardt} Caldwell R R, Dave R and Steinhardt P J, 1998
370: {\it Phys. Rev. Lett.} {\bf 80} 1582 [astro-ph/9708069]
371: \bibitem{phantom} Caldwell R R, 2002 {Phys. Lett. B} {\bf 545} 23 [astro-ph/9908168]
372: \bibitem{zhao} Zhao G -B, Xia J -Q, Li M, Feng B and Zhang X, 2005 {\it Preprint} astro-ph/0507482
373: \bibitem{vikman} Vikman A, 2005 {\it Phys. Rev. D} {\bf 71} 023515 [astro-ph/0407107]
374: \bibitem{quintom} Feng F, Wang X and Zhang X, 2005 {\it Phys. Lett. B} {\bf 607}
375: 35 [astro-ph/0404224]; Feng B, Li M, Piao Y -S and
376: Zhang X, 2004 {\it Preprint} astro-ph/0407432
377: \bibitem{wei} Wei Y -H and Tian Y, 2004 {\it Class. Quant. Grav.} {\bf 21} 5347 [gr-qc/0405038]
378:
379: \bibitem{zhang} Zhang X, 2004 {\it Preprint} hep-ph/0410292
380: \bibitem{guo} Guo Z -K, Piao Y -S, Zhang X and Zhang Y -Z, 2005
381: {\it Phys. Lett. B} {\bf 608} 177 [astro-ph/0410654]
382: \bibitem{hu} Hu W, 2005 {\it Phys. Rev. D} {\bf 71} 047301 [astro-ph/0410680]
383: \bibitem{xia} Xia J -Q, Feng B and Zhang X, 2005 {\it Mod. Phys. Lett. A} {\bf 20} 2409 [astro-ph/0411501]
384: \bibitem{michael} Caldwell R R and Doran M, 2005 {\it Phys. Rev. D} {\bf 72} 043527 [astro-ph/0501104]
385: \bibitem{li} Li B J, Chu M -C, Cheung K C and Tang A, 2005 {\it Preprint} astro-ph/0501367
386: \bibitem{cai} Wei H and Cai R -G, 2005 {\it Class. Quant. Grav.} {\bf 22} 3189 [hep-th/0501160]
387: \bibitem{xfzhang} Zhang X F, Li H, Piao Y -S and Zhang X, 2005
388: {\it Preprint} astro-ph/0501652
389: \bibitem{simon}
390: Simon J Z, 1990 {\it Phys. Rev. D} {\bf 41} 3720
391: \bibitem{woodard}
392: Elizer A and Woodard R P, 1989 {\it Nucl. Phys. B} {\bf 325} 389
393: \bibitem{gross}
394: Erler T G and Gross D J, {\it Preprint} hep-th/0406199
395: \bibitem{stelle}
396: Stelle K S, 1977 {\it Phys. Rev. D} {\bf 16} 953
397: \bibitem{ghost}
398: Arkani-Hamed N, Cheng H -C, Luty M A and Mukohyama S, 2004 JHEP 0405 074 [hep-th/0312099];
399: Arkani-Hamed N, Creminelli P, Mukohyama S and Zaldarriaga M, 2004 JCAP 0404 001 [hep-th/0312100];
400: Anisimov A and Vikman A, 2005 JCAP 0504 009 [hep-ph/0411089];
401: Mukohyama S, 2005 {\it Phys. Rev. D} {\bf 71} 104019 [hep-th/0502189];
402: Arkani-Hamed N, Cheng H -C, Luty M A, Mukohyama S and Wiseman T, 2005 {\it Preprint} hep-ph/0507120
403: \bibitem{ostrogradski}
404: Ostrogradski M, 1850 {\it Mem. Ac. St. Petersbourg VI} {\bf 4} 385
405: \bibitem{kinflation} Garriga J and Mukhanov V F, 1999 {\it Phys. Lett. B} {\bf 458}
406: 219 [hep-th/9904176]
407: \bibitem{hsu} Hsu S D H, Jenkins A and Wise M B, 2004 {\it Phys. Lett. B} {\bf 597} 270 [astro-ph/0406043];
408: Buniy R V and Hsu S D H, 2005 {\it Preprint} hep-th/0502203
409: \bibitem{smilga} Smilga A V, 2005 {\it Nucl. Phys. B} {\bf 706} 598 [hep-th/0407231]; 2005 {\it Preprint} hep-th/0503213
410:
411: \bibitem{binf} Anisimov A, Babichev E and Vikman A, 2005 JCAP 0506 006 [astro-ph/0504560]
412:
413: \bibitem{Phtproblms}
414: Carroll S M, Hoffman M and Trodden M, 2003 {\it Phys. Rev. D} {\bf 68}
415: 023509 [astro-ph/0301273]; Cline J M, Jeon S -Y and Moore G D, 2004 {\it Phys. Rev. D}
416: {\bf 70} 043543 [hep-ph/0311312]
417:
418:
419: \end{thebibliography}
420:
421: \end{document}
422: