1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %Version from May 2003
3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
4: %
5: \documentclass[aps,floats,superscriptaddress,showpacs]{revtex4}
6: %\documentclass[prb,preprint,showpacs]{revtex4}% Physical Review B
7: %\documentclass[twocolumn,aps,floats,superscriptaddress,showpacs]{revtex4}
8: %\documentclass[prl,showpacs,preprintnumbers,psfig,amsmath,amssymb]{revtex4}
9: %\documentclass[aps,floats,superscriptaddress,showpacs]{revtex4}
10: \usepackage{amsmath}
11: \renewcommand{\topfraction}{0.85}
12: \renewcommand{\textfraction}{0.1}
13: \renewcommand{\floatpagefraction}{0.75}
14: \usepackage{epsfig}
15: \usepackage{graphicx}% Include figure files
16: \usepackage{dcolumn}% Align table columns on decimal point
17: \usepackage{bm}% bold math
18: \renewcommand{\dag}{^{\dagger}}
19: \newcommand{\dl}{\partial_\ell}
20: \def\gapp{\lower.35em\hbox{$\stackrel{\textstyle>}{\sim}$}}
21: \def\lapp{\lower.35em\hbox{$\stackrel{\textstyle<}{\sim}$}}
22: \begin{document}
23: \bibliographystyle{apsrev}
24: %
25: \title{Pomeranchuk instability in doped graphene}
26: %
27: \author{Bel\'en Valenzuela }
28: \affiliation{Departamento de F\'isica de la Materia Condensada\\
29: Facultad de Ciencias, Universidad Aut\'onoma de Madrid,\\
30: Cantoblanco, E-28049 Madrid, Spain.}
31: %
32: \author{Mar\'{\i}a A. H. Vozmediano}
33: \affiliation{Instituto de Ciencia de Materiales de Madrid,\\
34: CSIC, Cantoblanco, E-28049 Madrid, Spain.}
35: %
36: \date{\today}
37: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
38: \begin{abstract} The density of states of graphene has Van
39: Hove singularities that can be reached by chemical doping and
40: have already been explored in photoemission experiments. We show
41: that in the presence of Coulomb interactions the system at the Van
42: Hove filling is likely to undergo a Pomeranchuk instability
43: breaking the lattice point group symmetry. In the presence of an
44: on--site Hubbard interaction the system is also unstable towards
45: ferromagnetism. We explore the competition of the two
46: instabilities and build the phase diagram. We also suggest that,
47: for doping levels where the trigonal warping is noticeable, the
48: Fermi liquid state in graphene can be stable up to zero
49: temperature avoiding the Kohn--Luttinger mechanism and providing
50: an example of two dimensional Fermi liquid at zero temperature.
51: \end{abstract}
52: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
53: %
54: \pacs{75.10.Jm, 75.10.Lp, 75.30.Ds}
55:
56: \maketitle
57:
58: \section{Introduction}
59:
60: The recent synthesis of single or few layers of
61: graphite\cite{Netal05,Zetal05} has permitted to test the singular
62: transport properties predicted in early theoretical
63: studies\cite{S84,Hal88,GGV96,Khv01} and experiments\cite{Ketal03}.
64: The discovery of a substantial field effect \cite{Netal04} and the
65: expectations of ferromagnetic behavior\cite{Eetal03} has risen
66: great expectations to use graphene as a reasonable replacement of
67: nanotubes in electronic applications.
68:
69: The Fermi level of graphene can be tuned over a wide energy range
70: by chemical doping \cite{MBetal07,BOetal07}, or by gating
71: \cite{Netal05,Zetal05}. The full dispersion relation of graphene
72: (Fig. 2) shows several regions of special interest. The low doping
73: region near half filling has been the object of much attention due
74: to the behavior of quasiparticles as massless Dirac electrons. The
75: Fermi surface at low filling (cusps of the dispersion relation)
76: begins being circular. For increasing values of the electron
77: density the trigonal distortion appears and, finally, several Van
78: Hove singularities (VHS) develop at energies the order of the
79: hopping parameter $E\sim 2.7 eV$. The possibility of finding
80: interesting physics around these densities have been put forward
81: recently in an experimental paper reporting on angle resolved
82: photoemission (ARPES) results \cite{MBetal07}. Chemical doping of
83: graphene with up to the VHS has been demonstrated in
84: \cite{MBetal07}. The VHS are of particular interest in the
85: structure of graphene nanoribbons \cite{E07,NRC08} and nanotubes
86: where observation of Van Hove singularities using scanning
87: tunneling microscopy have been reported \cite{Ketal99}. It is
88: known that around the Dirac point due to the vanishing density of
89: states at the Fermi level, short range interactions such as an
90: on--site Hubbard term $U$ are irrelevant in the renormalization
91: group sense \cite{GGV94,GGV99} and should not give rise to
92: instabilities at low energies or temperatures. The possibility to
93: stabilize a ferromagnetic \cite{PGC05b} or superconducting
94: \cite{GGV01,UC07,H07} phase near half filling is very unlikely.
95: Alternatively, at densities around the VHS the physics is
96: dominated by the high density of states and should resemble the
97: one discussed before in the framework of the high-$T_c$
98: superconductors \cite{M97}. This is the physical situation to look
99: for ferromagnetic or superconductivity \cite{G08} in graphene --or
100: graphite--. In ref. \cite{GGV00} it was argued that the most
101: likely instabilities of a system whose Fermi surface has VHS and
102: no special nesting features are p--wave superconductivity and
103: ferromagnetism. A very interesting competing instability that may
104: occur when the Fermi surface approaches singular points is a
105: redistribution of the electronic density that induces a
106: deformation of the Fermi surface breaking the symmetry of the
107: underlying lattice. This phenomenon called Pomeranchuk instability
108: \cite{P58} has been found in the squared lattice at the VHS
109: filling \cite{HM00,VV01,G01,LCG08} and has played an important
110: role in the physics of the cuprate superconductors
111: \cite{YK00,YK00b} and in general in layered materials \cite{Y08}.
112: It has also been discussed in a more general context in \cite{V03}
113: where it was argued that opening of an anisotropic gap at the
114: Fermi surface is an alternative to cure the infrared singularities
115: giving rise to Pomeranchuk instabilities. More recently the
116: Pomeranchuk instability has been studied in relation with the
117: possible quantum critical points in strongly correlated systems
118: \cite{NC05}. Due to the special symmetry of the Fermi surface of
119: the honeycomb lattice, the pattern of symmetry breaking phases can
120: be richer than these of the square lattice.
121:
122: In this paper we study the Van Hove filling in graphene modelled
123: with a single band Hubbard model with on--site U and exchange
124: Coulomb interactions V. We perform a mean field calculation and
125: show that Pomeranchuk instability occurs very easily in the system
126: in the presence of the exchange interaction V. Adding the on--site
127: interaction U allows for ferromagnetic ground states. We study the
128: coexistence or competition of the two and build a phase diagram
129: in the (U, V) space of parameters. In section \ref{method} we
130: construct the exchange V and on--site U interactions in the
131: honeycomb lattice and describe the method of calculation and the
132: results obtained.
133: %In \ref{results} we discuss the competition of the
134: %Pomeranchuk instability with ferromagnetism when both interactions
135: %are included. A phase diagram is constructed as a function of the
136: %interactions.
137: In the last section we discuss some points related with the
138: physics of graphene at high doping levels. In the context of the
139: original Kohn--Luttinger instabilities \cite{KL65} that
140: establishes that all Fermi liquids will be unstable at low enough
141: temperatures, we suggest that graphene at intermediate fillings
142: can provide an example of stable metallic system at zero
143: temperature \cite{FKT04a,FKT04b,FKT04c}. Finally we set the lines
144: for future developments.
145:
146: \section{The model}
147: \label{method} The deformation of the Fermi surface by the
148: electronic interactions is a classical problem in condensed matter
149: dating back to founding work of Luttinger \cite{L60}. We model the
150: system by a single band Hubbard model in the honeycomb lattice
151: with an on--site interaction $U$ and a nearest neighbor Coulomb
152: interaction $V$, and perform a self--consistent calculation along
153: the lines of refs. \cite{VV01,RLG08}.
154:
155: The Hubbard hamiltonian is
156: %
157: \begin{equation}
158: H=t\sum_{{\bf i};\sigma} c^+_{{\bf i}\sigma} c_{{\bf i}\sigma}
159: \;+\;U\sum_{\bf i} n_{{\bf i}\uparrow} n_{{\bf i}\downarrow}\;+
160: V\sum_{<{\bf i} {\bf j};\sigma \sigma'>}n_{{\bf i \sigma} }n_{{\bf j
161: \sigma'}} \;\;. \label{ham-fs}
162: \end{equation}
163: %
164: \begin{figure}
165: \begin{center}
166: \includegraphics[width=5cm]{figure1.eps}
167: \caption{(Color online) The honeycomb lattice of graphene.}
168: \label{lattice}
169: \end{center}
170: \end{figure}
171: %
172: The hopping parameter $t$ of graphene is $t\sim2.7eV$. The
173: next-to-nearest neighbor hopping $t'$ is estimated to be much
174: smaller \cite{Oetal07} and it will be ignored. Due to the special
175: topology of the honeycomb lattice with two atoms per unit cell the
176: Hamiltonian is a $2\times 2$ matrix. We will work out in some
177: detail the different terms. The operators $\hat{a}, \hat{b}$
178: associated to the two triangular sublattices A, B are
179: %
180: \begin{equation}
181: a^+_{i A\sigma}=\frac{1}{\sqrt{N_A}}
182: %
183: \sum_k e^{i{\bf k. R_{i A}}}a^+_{{\bf k}\sigma} \;\;\;,\;\;\;
184: b^+_{i B\sigma}=\frac{1}{\sqrt{N_B}}
185: %
186: \sum_k e^{i{\bf k.R_{i B}}}b^+_{{\bf k}\sigma} \;\;\;,\;\;\;
187: \vec{R}_{i A}=\vec{R}_{i B}+\vec{\delta},
188: \end{equation}
189: %
190: where $N_A=N_B=N$ is the number of lattice cells and $\delta_i$
191: are the three vectors connecting a point of sublattice A with its
192: three neighbors in sublattice B (see Fig. \ref{lattice}).
193:
194: In term of these operators the free Hamiltonian reads
195: %
196: \begin{equation}
197: H_0=t\sum_{k\sigma}[\phi({\bf k})a^+_{k\sigma}b_{k\sigma}
198: +\phi^*({\bf k})b^+_{k\sigma}a_{k\sigma}]
199: +\mu\sum_{k\sigma}(a^+_{k\sigma}a_{k\sigma}+b^+_{k\sigma}b_{k\sigma}),
200: \end{equation}
201: %
202: where we have included a chemical potential $\mu$.
203:
204: In matrix form the non--interacting Hamiltonian is
205: \begin{eqnarray}
206: H_0=\left(%
207: \begin{array}{cc}
208: \mu & t\phi({\bf k}) \\
209: t\phi^*({\bf k}) & \mu\\
210: \end{array}%
211: \right), \label{H0}
212: \end{eqnarray}
213: %
214: where
215: %
216: \begin{equation}
217: \phi(\vec{k})=\sum_i e^{-i\vec{k} .\vec{\delta}_i},
218: \end{equation}
219: %
220: The Hamiltonian (\ref{H0}) gives rise to the
221: dispersion relation
222: %
223: \begin{equation}
224: \varepsilon^0({\bf k})=\mu\pm t\sqrt{\vert\phi(\vec{k})\vert^2} =
225: \mu\pm t\sqrt{1+4\cos^2\frac{\sqrt{3}}{2}ak_x+
226: 4\cos\frac{\sqrt{3}}{2}ak_x\cos\frac{3}{2}ak_y} \;,
227: \label{disprel}
228: \end{equation}
229: %
230: whose lower band is shown in the left hand side of Fig.
231: \ref{band2}.
232: %
233: \begin{figure}
234: \begin{center}
235: \includegraphics[width=8cm]{figure2a.eps}
236: \includegraphics[width=6cm]{figure2b.eps}
237: \caption{(Color online) Left: Lower band of the dispersion
238: relation of graphene. The black line marks the Fermi surface for
239: the Van Hove filling. Right: Evolution of the Fermi surface of
240: graphene with doping. At half filling the Fermi surface consists
241: on the six vertices of the hexagonal Brillouin zone (black
242: points). Only two are independent, the rest being equivalent by
243: lattice vectors. Emptying the half filled band the Fermi surface
244: develops circular hole pockets that undergo the trigonal
245: distortion (in red) and become the yellow triangles at the Van
246: Hove fillig $\mu=t$. } \label{band2}
247: \end{center}
248: \end{figure}
249: %
250:
251:
252: \section{Coulomb interaction V and Pomeranchuk instability}
253:
254: We will begin by studying the influence of the interaction $V$
255: omitting the spin of the electrons. The on-site U will be added to
256: study the competition of the Pomeranchuk instability with
257: ferromagnetism. The interaction V takes place between nearest
258: neighbors belonging to opposite sublattices. It reads:
259: %
260: \begin{equation}
261: H_V=V\sum_{{\bf kk'q}}[a^+_{\bf k}a_{ {\bf k+q}}b^+_{{\bf
262: k'}}b_{{\bf k'-q}}\phi^*({\bf q})+b^+_{{\bf k}}b_{{\bf
263: k+q}}a^+_{{\bf k'}}a_{{\bf k'-q}}\phi({\bf q})].
264: \end{equation}
265: %
266: The mean field Hamiltonian becomes
267: %
268: \begin{equation}
269: H_{MF}=\sum_{\bf k}[\tilde{\phi}({\bf k}) a^+_{{\bf k}}b_{{\bf
270: k}}+\tilde{\phi}^*({\bf k}) b^+_{{\bf k}}a_{\bf k}]+\mu\sum_{\bf
271: k}(a^+_{\bf k}a_{\bf k}+b^+_{\bf k}b_{\bf k}), \label{HMF}
272: \end{equation}
273: %
274: where
275: %
276: \begin{equation}
277: \tilde{\phi}({\bf k})=t\phi({\bf k})-\sum_{{\bf q}} V^*({\bf q})
278: <b^+_{{\bf k+q}}a_{{\bf k+q}}> \;\;\;,\;\;\;
279: %
280: V({\bf q})=\frac{V}{N}\sum_{{\bf q}}\phi^*({\bf q}),
281: \end{equation}
282: and $N$ is the number of lattice cells. In matrix form and
283: including the chemical potential the mean field Hamiltonian reads:
284: %
285: \begin{eqnarray}
286: H_{MF}=\left(%
287: \begin{array}{cc}
288: \mu & t\phi({\bf k}) -\sum_{\bf q}V^*({\bf q})<b^+_{k+q}a_{k+q}> \\
289: t\phi^*({\bf k})-\sum_{\bf q}V({\bf q})<a^+_{k+q}b_{k+q}> & \mu\\
290: \end{array}
291: %
292: \right).
293: \label{HM}
294: \end{eqnarray}
295:
296: The mean field Hamiltonian (\ref{HM}) gives rise to the dispersion
297: relation
298: %
299: \begin{equation}
300: E({\bf k})=\mu\pm[t^2\vert\phi({\bf k})\vert^2+F(V, {\bf k})]^{1/2},
301: \label{EV}
302: \end{equation}
303: %
304: where $F(V, {\bf k})$ is
305: %
306: \begin{eqnarray}
307: F(V,{\bf k})&= -t V\sum_{{\bf q}} \phi^*({\bf
308: k})\phi({\bf q}) <a^+_{k+q}b_{k+q}> -t V \sum_{{\bf q}}\phi({\bf
309: k})\phi^*({\bf q})
310: <b^+_{k+q}a_{k+q}> \\ \nonumber
311: &+
312: %
313: V^2\sum_{{\bf q q'}}\phi^*({\bf q})\phi({\bf q'})
314: <a^+_{k+q}b_{k+q}><b^+_{k+q'}a_{k+q'}>.
315: \end{eqnarray}
316:
317: We look for a self--consistent solution of eq. (\ref{EV}) imposing
318: the Luttinger theorem {\it i. e.} that the area enclosed by the
319: interacting Fermi line is the same as the one chosen as the
320: initial condition. We begin with an initial (free) Fermi surface,
321: as an input, add the interaction and let it evolve until a
322: self--consistent solution is found with a given "final"
323: interacting Fermi surface. The important constraint is that the
324: total number of particles should remain fixed. This procedure has
325: been used in the same context in \cite{VV01,RLG08}.
326:
327: \begin{figure}
328: \begin{center}
329: \includegraphics[width=5cm]{figure3.eps}
330: \caption{Pomeranchuk instability suffered by an initial Fermi
331: surface at the Van Hove singularity (upper part) when an exchange
332: interaction V=1t is added (lower part).}
333: \label{deformed}
334: \end{center}
335: \end{figure}
336: %
337: The Fermi surface of graphene around the Van Hove singularity is
338: shown in Fig. \ref{band2}. To better visualize the Pomeranchuk
339: instability we keep the image of the neighboring Brillouin zones.
340: Fig. \ref{deformed} shows the initial Fermi surface sitting at the
341: VHS (upper part) and the spontaneous deformation obtained
342: self--consistently from eq. (\ref{EV}) for a value of the exchange
343: interaction $V=t$.
344:
345: \section{Ferromagnetism and competition of ferromagnetism and
346: Pomeranchuk instabilities} \label{FM}
347:
348: In order to study a possible ferromagnetic instability we will
349: add to the free Hamiltonian an on--site Hubbard term
350: %
351: \begin{equation}
352: H_U=U\sum_i[a^+_{i\uparrow}a_{i\uparrow}a^+_{i\downarrow}a_{i\downarrow}+
353: b^+_{i\uparrow}b_{i\uparrow}b^+_{i\downarrow}b_{i\downarrow}],
354: \end{equation}
355: %
356: A mean field ferromagnetic state will be characterized by
357: %
358: \begin{equation}
359: a^+_{i\sigma}a_{i\sigma}=<a^+_{i\sigma}a_{i\sigma}>+\delta
360: (a^+_{i\sigma}a_{i\sigma})
361: \end{equation}
362: %
363: with
364: \begin{equation}
365: <a^+_{i\sigma}a_{i\sigma}>=\frac{n}{2}+\sigma\frac{m}{2}\;\;\;,\;\;\;
366: <b^+_{i\sigma}b_{i\sigma}>=\frac{n}{2}+\sigma\frac{m}{2}\;\;\;,\;\;\;\sigma=\pm,
367: \end{equation}
368: %
369: what produces the mean field Hamiltonian
370: \begin{equation}
371: H^U_{MF}=U\sum_{i\sigma}(\frac{n}{2}+\sigma\frac{m}{2})
372: (a^+_{i\sigma}a_{i\sigma}+b^+_{i\sigma}b_{i\sigma})-\frac{1}{2}U(n^2-m^2)N.
373: \end{equation}
374: %
375: Adding the kinetic term we get the Hamiltonian
376: %
377: \begin{eqnarray}
378: H_0+H^U_{MF}=\left(%
379: \begin{array}{cc}
380: \mu +\frac{U}{2}(n+\sigma m)& t\phi({\bf k}) \\
381: t\phi^*({\bf k}) & \mu+\frac{U}{2}(n+\sigma m)\\
382: \end{array}
383: %
384: \right) \label{HUM}
385: \end{eqnarray}
386: %
387: whose dispersion relation
388: %
389: \begin{equation}
390: E_{k\sigma}=\mu_\sigma\pm t\vert\phi({\bf k})\vert\;\;\;,\;\;\;
391: \mu_\sigma=\mu+\frac{U}{2}(n+\sigma m),
392: \end{equation}
393: is solved self--consistently to get the ferromagnetic ground
394: state.
395: %
396: \begin{figure}
397: \begin{center}
398: \includegraphics[width=5.0cm]{figure4a.eps}
399: \includegraphics[width=5.0cm]{figure4b.eps}
400: \includegraphics[width=5.0cm]{figure4c.eps}
401: \caption{(Color online) Left: Initial polarization of the Fermi
402: surface with the spin up electrons (upper part) less populated
403: than the spin down (lower part): $n_\downarrow-n_\uparrow=0.14$,
404: $\mu_\uparrow=-0.9t$, $\mu_\downarrow=-1.1t$. Center: Final state
405: obtained for the values of the parameters U=1t and V=2t. A
406: Pomeranchuk instability is clearly visible in both spin bands.
407: Right: Final state with enhanced ferromagnetic polarization
408: obtained for the values of the parameters U=2t and V=1t.}
409: \label{ferro}
410: \end{center}
411: \end{figure}
412: %
413:
414:
415: The competition of the Pomeranchuk instability found in the previous
416: section with a possible ferromagnetic instability is studied with
417: the same procedure using the full mean field Hamiltonian
418: $H_0+H^V_{MF}+H^U_{MF}$ and solving self-consistently the equations
419: %
420: %
421: \begin{equation}
422: E_\uparrow({\bf k})=\mu+U n_\downarrow -\sqrt{t^2\vert\phi({\bf
423: k})\vert^2+F(V, {\bf k})_\uparrow},
424: \end{equation}
425: %
426: \begin{equation}
427: E_\downarrow({\bf k})=\mu+U n_\uparrow-\sqrt{t^2\vert\phi({\bf
428: k})\vert^2+F(V, {\bf k})_\downarrow},
429: \end{equation}
430: %
431: where $F(V, {\bf k})_\sigma$ is given in eq. (12) and
432: $n_\sigma=1/2(n+\sigma m)$.
433:
434:
435: \section{Summary of the results}
436: \label{results}
437:
438: The competition of the Pomeranchuk instability with ferromagnetism
439: is exemplified in Fig. \ref{ferro}. The figure in the left side
440: represents the initial free Fermi surface where the spin up (upper
441: side) and down (lower side) electrons have different populations
442: around the Van Hove filling. In particular in the example given,
443: the chemical potential of the spin up (down) electrons is set
444: slightly below (above) the VHS: $n_\downarrow-n_\uparrow=0.14$,
445: $\mu_\uparrow=-0.9t$, $\mu_\downarrow=-1.1t$. The image in the
446: center represents the renormalized Fermi surface for the two spin
447: polarizations when the exchange interaction V is bigger than U:
448: $V=2U=2t$. We can see that the ferromagnetism has disappeared: the
449: final Fermi surface is the same for the two spin polarization and
450: presents a Pomeranchuk deformation. The opposite case is shown in
451: the figure at the right: with the same initial configuration the
452: final state for the values of the interactions $U=2V=2t$ has an
453: enhanced ferromagnetism with no signal of deformation.
454:
455: The phase diagram of the dominant instability as a function of the
456: interaction strength U and V (measured in units of the hopping
457: parameter t) is shown in Fig. \ref{phased2}. In all cases the
458: initial Fermi surface consists of a slightly polarized state
459: around the Van Hove singularity as the one shown in the left hand
460: side of Fig. \ref {ferro}. The symbols in Fig. \ref{phased2}
461: denote calculated points with the following meaning: the circles
462: appearing for low values of U denote an unpolarized final state
463: with Pomeranchuk deformation as the one shown in the center of
464: Fig. \ref{ferro}. Crosses appearing for large values of U
465: represent a final state where the polarization is bigger than the
466: initial one and the Fermi surface has the original symmetry as the
467: one in the right hand side of Fig. \ref{ferro}. The points denoted
468: by $\bigotimes$ correspond to values where both instabilities
469: coexist and the final Fermi surface is deformed and spin
470: polarized. Finally the crosses at low values of U with V=0
471: represent final states which are still polarized but where the
472: spin polarization is smaller than the initial one. This region is
473: denoted by $FM\downarrow$ in the figure phase diagram. PI (FM)
474: denotes the region where the Pomeranchuk (ferromagnetic)
475: instability dominated respectively.
476:
477: The ferromagnetic evolution of the system for V=0 is as follows:
478: Below a critical value of $1.5 < U_c < 2$ a free slightly
479: polarized system at fillings near (but not at) the VHS evolves
480: towards an unpolarized one in agreement with \cite{PAB04,PAB06}.
481: The Coulomb exchange V induces a deformation of the Fermi surface
482: already at very small values $V_c\sim 0.4 t$ when the free initial
483: state is near the Van Hove singularity. Increasing U enhances the
484: magnetic polarization of the interacting system until the critical
485: value $U\sim 1.75 t$ is reached where the final state corresponds
486: to a ferromagnetic system with un-deformed Fermi surface. Unlike
487: what happens in the square lattice, the critical value of U above
488: which ferromagnetism prevails does not depend on V (vertical line
489: in Fig. \ref{phased2}). In the blank region in the upper part of
490: the figure corresponding to high values of V around the critical U
491: we have not been able to reach a self--consistent solution.
492:
493: We note that at V=0, the critical U for ferromagnetism is zero at
494: the Van Hove filling and changes very rapidly around it in a rigid
495: band model \cite{PAB04,PAB06}. This behavior is due to the
496: divergent density of states at the VHS that would be smeared by
497: temperature effects or disorder in real samples. Based on the
498: previous studies if the square lattice we are confident that the
499: results obtained in the present work are robust and the phase
500: diagram shown in fig. \ref{phased2} will remain qualitatively when
501: more detailed calculations are done.
502:
503: The realization of one or another phase in real graphene samples
504: depends on the values that the effective Coulomb interaction
505: parametrized by $U, V$ has at the VHS filling. Besides the ARPES
506: experiments \cite{MBetal07} which do not comment on the strength
507: of the interaction, we are not aware of other experiments at these
508: doping levels. Conservative estimates can be obtained from the
509: graphite intercalated compounds \cite{TH92} on the basis of the
510: similarities found in \cite{MBetal07} between the Fermi surfaces
511: of the two. Values of $U\sim 2.5 t$, $V\sim t$ can be very
512: reasonable and lie in the range discussed in the present work.
513:
514:
515:
516: %
517: \begin{figure}
518: \begin{center}
519: \includegraphics[width=8cm]{figure5.eps}
520: \caption{Phase diagram showing the competition of Pomeranchuk and
521: ferromagnetic instabilities as a function of the interactions U
522: and V measured in units of the hopping parameter t.}
523: \label{phased2}
524: \end{center}
525: \end{figure}
526: %
527:
528:
529: \section{Discussion and open problems}
530: \label{future}
531:
532:
533: In this work we have examined some physical issues to be expected
534: in graphene at high doping. We have seen that at the Van Hove
535: filling in the presence of an exchange Coulomb interaction the
536: Fermi surface is {\it softer} and becomes very easily deformed as
537: compared with a similar analysis of the square lattice.
538:
539: A natural extension of the work presented in this article is to
540: study the competition of the instabilities studied in this work
541: with superconductivity along the lines of ref. \cite{YM07}. A
542: complete renormalization group analysis of the competing low
543: energy instabilities of the Van Hove filling as a function of the
544: doping and the couplings U and V is also to be done. Due to the
545: special geometry of the Fermi surface in the honeycomb lattice the
546: symmetry breaking phases can present a richer variety than those
547: of the square lattice. A fairly complete analysis of the short
548: range interactions around the Dirac point was done in the early
549: paper \cite{GGV01}. Due to the existence of two Fermi points the
550: classification of the possible low--energy couplings is similar to
551: the g--ology of the one dimensional models. The RG classification
552: of the low--energy couplings in the Van Hove filling is richer
553: since there are three independent VH points to consider and work
554: in this direction is in progress. An analysis of the physics of
555: the Fermi surface around the trigonal warping can lead also to
556: very interesting results in the light of the evolution of the
557: Fermi surface anisotropies done in \cite{GGV97,RLG06}.
558:
559: On the view of the analysis of this problem done in the square
560: lattice we can expect that RG calculation will enrich the phase
561: diagram but the two phases discussed here will stay.
562: Ferromagnetism is a likely possibility due to the Stoner criterium
563: although it will compete with antiferromagnetism at high values of
564: U. A very important parameter to this problem is the next to
565: nearest neighbor hopping $t'$ that suppresses the nesting of the
566: bare Fermi surface and affects the shape of the phase transition
567: lines \cite{PAB04,PAB06}. These references focuss on the magnetism
568: of the Hubbard model (U) on the honeycomb lattice at finite
569: dopings and found that non--homogeneous (spiral) phases are the
570: most stable configurations around the Van Hove filling. It would
571: be interesting to analyze the competition and stability of these
572: non--homogeneous ferromagnetic configurations in the presence of
573: a V interaction.
574:
575: Of particular interest is the possibility of coexistence or
576: competition of the Pomeranchuk instability with possible
577: superconducting instabilities \cite{G08}. A very interesting
578: suggestion has been made in studies of the square lattice that
579: superconductivity changes the nature of the Pomeranchuk transition
580: going from first to second order \cite{YM07}. This feature will
581: probably be maintained in the honeycomb lattice and we are
582: actually exploring this problem.
583:
584:
585: %\subsection
586: %{ A two dimensional Fermi liquid at zero temperature?}
587:
588: As it is known, the Kohn--Luttinger instability in its original context
589: \cite{KL65} suggests that no Fermi liquid will be stable at
590: sufficiently low temperatures. In the case of a 3D electron system
591: with isotropic Fermi surface there is an enhanced scattering at
592: momentum transfer $2k_F$ which translates into a modulation of
593: the effective interaction potential with oscillating behavior
594: similar to the Friedel oscillations. This makes possible the
595: existence of attractive channels, labelled by the angular momentum
596: quantum number. The issue of the stability of Fermi liquids was
597: studied rigorously in two dimensions in a set of papers
598: \cite{FKT04a,FKT04b,FKT04c} with the conclusion that a Fermi
599: liquid could exist in two space dimensions at zero temperature
600: provided that the Fermi surface of the system obeys some
601: "asymmetry" conditions. In particular the non--interacting Fermi
602: surface had to lack inversion symmetry
603: %
604: $$\epsilon(-{\bf k})\neq\epsilon({\bf k})$$
605: %
606: at all points k and be otherwise quite regular (the Van Hove
607: singularities can not be present). The dispersion relation of
608: graphene obeys such a condition for fillings where the trigonal
609: warping is already noticeable and below the Van Hove filling
610: provided that disorder and interactions do not mix the Van Hove
611: points. This is a very common assumption in the graphene physics
612: around the Dirac points where the inversion symmetry in k--space
613: acts as a time reversal symmetry \cite{MGV07}. We find interesting
614: to note that the --quite restrictive--conditions of refs.
615: \cite{FKT04a,FKT04b,FKT04c} can be fulfilled in a real --and very
616: popular--system.
617:
618:
619:
620:
621: \section{Acknowledgments.}
622:
623: We thank A. Cortijo, M. P. L\'opez-Sancho and R. Rold\'an for very
624: interesting discussions. This research was supported by the
625: Spanish MECD grant FIS2005-05478-C02-01 and by the {\it
626: Ferrocarbon} project from the European Union under Contract 12881
627: (NEST).
628: %
629: \bibliography{VH}
630: \end{document}
631: