1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: % TeX file of the article: Electronic Properties of %
3: % Two-Dimensional Carbon %
4: % %
5: % %
6: % Title: %
7: % authors: N. M. R Peres, F Guinea, and A.H. Castro Neto %
8: % %
9: % %
10: % %
11: % %
12: % e-mail: peres@fisica.uminho.pt %
13: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
14:
15:
16:
17: \documentclass[prl,twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
18:
19: \usepackage{epsfig}
20: \usepackage{graphicx}% Include figure files
21: \usepackage{dcolumn}% Align table columns on decimal point
22: \usepackage{bm}% bold math
23:
24: \newcommand{\la}{\langle}
25: \newcommand{\ra}{\rangle}
26: \newcommand{\ua}{\uparrow}
27: \newcommand{\da}{\downarrow}
28: \newcommand{\udl}{\underline}
29: \newcommand{\rar}{\rightarrow}
30: \newcommand{\e}{\epsilon}
31: \newcommand{\om}{\omega}
32: \newcommand{\vk}{{\vec k}}
33: \newcommand{\vkp}{{\vec k\,'}}
34: \newcommand{\vq}{{\vec q}}
35: \newcommand{\vqp}{{\vec q\,'}}
36: \newcommand{\vp}{{\vec p}}
37: \newcommand{\vpp}{{\vec p\,'}}
38: \newcommand{\vf}{v_{\rm F}}
39: \newcommand{\tinin}{\tau^{-1}_{in}}
40: \newcommand{\svac}{\Sigma_{\rm vac}}
41: \newcommand{\cv}{c_{\rm vac}}
42: \newcommand{\lv}{l_{\rm vac}}
43: \newcommand{\oo}{\omega_0}
44: \newcommand{\ef}{\epsilon_{\rm F}}
45: \newcommand{\tsci}{\tau_{\rm elastic}^{-1}}
46: \newcommand{\nimp}{n_{\rm i}}
47: \newcommand{\nstar}{n^*}
48: \newcommand{\lB}{\ell_{\rm B}}
49: \newcommand{\sdc}{\sigma_{\rm dc}}
50: \newcommand{\kf}{k_{\rm F}}
51: \newcommand{\Phio}{\Phi_0}
52:
53: \begin{document}
54:
55: \title{Electronic Properties of Two-Dimensional Carbon}
56: \author{N. M. R. Peres$^{1,2}$,
57: F. Guinea$^{1,3}$, and A. H. Castro Neto
58: $^2$}
59:
60: \affiliation{$^1$Department of Physics, Boston University, 590
61: Commonwealth Avenue, Boston, MA 02215,USA}
62: \affiliation{$^2$Center of Physics and Department of Physics,
63: Universidade do Minho, P-4710-057, Braga, Portugal}
64: \affiliation{$^3$Instituto de Ciencia de Materiales de Madrid, CSIC,
65: Cantoblanco E28049 Madrid, Spain}
66:
67: \begin{abstract}
68: We present a theoretical description of the electronic properties of
69: graphene in the presence of disorder, electron-electron interactions,
70: and particle-hole symmetry breaking.
71: We show that while particle-hole asymmetry, long-range Coulomb interactions,
72: and extended defects lead to the phenomenon of
73: self-doping, local defects determine the transport and spectroscopic properties.
74: Our results explain recent experiments in graphitic devices and predict
75: new electronic behavior.
76: \end{abstract}
77: \pacs{81.05.Uw; 71.55.-i; 71.10.-w}
78:
79: \maketitle
80:
81:
82: Carbon with sp$^2$ bonding has many allotropic forms
83: with different dimensionality: three-dimensional graphite \cite{BCP88},
84: one-dimensional nanotubes \cite{nanotubes}, zero-dimensional fullerenes
85: \cite{buckyball}, and two-dimensional Carbon, also known as graphene.
86: Graphene can be considered the {\it materia prima}
87: for the other forms of Carbon that can be obtained from it
88: either by stacking (graphite),
89: wrapping (nanotubes), or creation of topological defects (fullerenes).
90: Hence, the electronic response of many Carbon based systems depends fundamentally
91: on the basic physics of graphene.
92: Recent experiments in graphene-based devices have shown that it is possible
93: to control their electrical properties, such as carrier type (particle
94: or hole), by the application of external gate voltage \cite{Netal04,outros}.
95: These experiments not only have opened doors to carbon-based
96: nano-electronics but also pose new questions on the
97: nature of the electronic properties in these systems.
98:
99: Being two-dimensional, true long range positional order of the
100: Carbon atoms is not possible at any finite temperature,
101: since thermal fluctuations introduce topological lattice defects such as
102: dislocations (the Hohenberg-Mermim-Wagner theorem).
103: Furthermore, because of the particular structure of the honeycomb lattice,
104: the dynamics of lattice defects in graphene belongs to
105: the generic class of kinetically constrained models \cite{DS00,RS03},
106: where defects are never completely annealed since their
107: number is only weakly dependent on the annealing time \cite{DS00}.
108: Indeed, defects are ubiquitous in carbon allotropes with sp$^2$ coordination,
109: as confirmed by recent experiments \cite{Hetal04b}.
110: Furthermore, lattice defects induced by proton
111: bombardment have been correlated to the appearance of magnetism in graphitic
112: samples \cite{magnet}, indicating the interplay between electron-electron
113: interactions and disorder.
114:
115: Besides dislocations, graphene can sustain other types of extended
116: defects such as disclinations, edges, and micro-cracks. It is
117: known that certain edges (such as the zig-zag edge)
118: lead to the appearance of localized states at the Fermi level
119: \cite{WS00,Vetal05}. Other defects such as pentagons and heptagons
120: (lattice disclinations) also admit localized states \cite{GGV92}.
121: Furthermore, tight-binding calculations show that a combination of a
122: five-fold and seven-fold ring (a lattice dislocation) or a Stone-Wales defect
123: (made up of two pentagons and two heptagons) also lead to a finite density of
124: states at the Fermi level\cite{CEL96,MA01,DSL04}. The presence of
125: states at the Fermi level generated by defects
126: can be tracked down to the particular
127: electronic structure of graphene. Each $\pi$-orbital contributes with
128: one electron (a half-filled band) and
129: the low-energy physics is described by a two-dimensional Dirac equation
130: with a vanishing electronic density of states at the Fermi level. The
131: vanishing of the density of states has two very important consequences:
132: the enhancement of the electron-electron interactions because of the
133: absence of electronic screening, and the formation of defect states
134: at the Fermi level. In this respect, our work complements the extensive literature on defects on
135: electronic systems described by the two dimensional Dirac
136: equation\cite{hirsch}.
137:
138: In the present work, we show that unscreened Coulomb interactions,
139: particle-hole symmetry breaking and defects play a fundamental role
140: in the electronic properties of graphene. In particular, graphene
141: presents the phenomenon of {\it self-doping}, that is, charge
142: transfer between extended defects and the bulk. In this case,
143: depending on particle-hole symmetry breaking, graphene samples
144: can either have electron or hole pockets. While self-doping
145: derives from extended defects and controls the number and type of
146: charge carriers, transport properties also
147: depend on point-like defects such as vacancies, and
148: surface ad-atoms. As we show in what follows, magneto-transport properties,
149: such as Shubnikov-de Hass oscillations and
150: quantum Hall effect (QHE), as well as spectroscopic quantities, such
151: as quasiparticle lifetimes and infrared reflectivity, can be
152: completely explained within this framework. Besides explaining published
153: experimental data, we also make new experimental predictions
154: that can be used as tests of our theory.
155:
156: The bulk Hamiltonian of graphene can be written as:
157: ${\cal H} = {\cal H}_0 + {\cal H}_V + {\cal H}_U$, where,
158: \begin{equation}
159: {\cal H}_0 = - t \sum_{\sigma ; \langle i,j \rangle} c^\dag_{i,\sigma} c_{j, \sigma} + t'
160: \sum_{\sigma ; \langle \langle i,j \rangle \rangle } c^\dag_{i,\sigma} c_{j, \sigma} + h. c.
161: \label{hamil}
162: \end{equation}
163: where $c_{i,\sigma}$ ($c_{i,\sigma}^{\dag}$) annihilates (creates) electrons
164: at the site ${\bf R}_i$ with spin $\sigma$ ($\sigma=\uparrow,\downarrow$),
165: $t$ and $t'$ are the nearest neighbor and next-nearest
166: neighbor hopping energies, respectively. For ${\bf k} \to 0$ the electronic dispersion
167: is $\epsilon_{\bf k} \approx 3 t' + v_{\rm F} | {\bf k} | +
168: (9 t' a^2 |{\bf k}|^2) / 4$. Notice that $t'$ does not change the structure of the
169: extended electronic wavefunctions,
170: that remain described by the two-dimensional Dirac equation with
171: a Fermi-Dirac velocity $v_{\rm F} = (3 t a)/2$ ($a$ is the lattice spacing).
172: ${\cal H}_V$ describes the disorder,
173: \begin{eqnarray}
174: {\cal H}_V = \sum_{j} V_j n_j \, ,
175: \label{disorder}
176: \end{eqnarray}
177: where $V_j$ is the potential strength (for vacancies, $V_j \to \infty$ at the
178: vacant site and is zero, otherwise), and $n_j = \sum_{\sigma} c^\dag_{j,\sigma}
179: c_{j, \sigma}$. The electron interactions read,
180: \begin{eqnarray}
181: {\cal H}_U = (e^2/\epsilon_0) \sum_{i,j \, (i \neq j)} n_i n_j/|{\bf R}_i-{\bf R}_j| \, ,
182: \label{interaction}
183: \end{eqnarray}
184: $e$ is the electric charge, and $\epsilon_0$ is the dielectric constant.
185:
186: Standard derivations of localized states near defects assume electron-hole
187: symmetry, that is, $t'=0$. In this case, the localized states lie exactly at the Fermi level.
188: In order to ensure charge neutrality,
189: these states are half filled, and there is no charge transfer between extended
190: and localized states. Electron-hole symmetry is broken by terms in the
191: Hamiltonian that allow for hopping between sites in the same sublattice, that
192: is, when $t' \neq 0$. In this case the band of localized states
193: near an extended defect acquires a bandwidth of order $t'$ and
194: is shifted from Fermi energy.
195: Using (\ref{hamil}) and (\ref{interaction}), we have studied in the
196: Hartree approximation the induced charge transfer
197: between the edge and the bulk in a long graphene ribbon of width $W$.
198: %, that is, we rewrite (\ref{interaction}) as
199: %$n_i n_j \to \langle n_i \rangle n_j$ and solve the self-consistent problem
200: %numerically for a {\it local} value of $\langle n_i \rangle$, away from the edge.
201: In Fig.[\ref{self_doping}] we present the results for the induced
202: electrostatic potential, charge density, and charge transferred from a
203: zig-zag edge to (delocalized) bulk states. We have studied zig-zag edges
204: as long as $0.1 \mu$ m in graphene ribbons with up to $10^5$ atoms.
205: As one can clearly see in Fig.[\ref{self_doping}], self-doping effects are
206: suppressed as the width of the edge increases. The Coulomb energy
207: at the edge induced by a constant doping $\delta$ per Carbon atom
208: is approximately $ \sim (\delta e^2/a) (W/a)$.
209: The charge transfer is arrested when the potential
210: shifts the band of localized states to the Fermi energy,
211: that is, when $t' \approx (e^2/a) (W/a) \delta$. Hence the
212: self-doping is $\delta \sim ( t' a^2 ) / ( e^2 W )$.
213: The numerical results are consistent with this estimate. Note that the sign
214: of the induced charge depends on the sign of $t'$.
215: The extended defects lead to a scattering
216: rate, $\tsci \sim \vf / W$. The pocket induced by the self-doping
217: has a Fermi energy of
218: $\ef \sim ( \hbar \vf \sqrt{\delta}) / a \sim t \sqrt{( t' a^2 ) / ( e^2 W)}$
219: and thus, $\hbar \tsci \ll \ef$ in wide ribbons.
220: Hence, disorder due to extended defects does not
221: smooth out the details of the small Fermi surface induced by the self-doping.
222: For $t = 2.7$ eV \cite{BCP88}, $t' =0.2 t \approx 0.54$ eV (see below),
223: and $a=1.42 \AA$, we get $\delta \sim 10^{-5} -10^{-4}$ per Carbon atom in a system with
224: domains of typical dimensions $W \sim 0.1 - 1 \, \, \mu$ m, consistent with
225: experiments \cite{Netal04}.
226:
227: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
228: % Figure
229: %%%%%%%%%%%%%%%
230: \begin{figure}[htf]
231: \begin{center}
232: \includegraphics[width=9cm]{Fig_Nature_sd.eps}
233: \end{center}
234: \caption{
235: \label{self_doping}
236: Top: Electrostatic potential (dashed line) and electronic charge (continuous
237: line) as a function of position
238: along a long graphene ribbon terminating on a zig-zag edge.
239: Energies are given in units $t$, and distances in units of $a$.
240: We have used $t' = 0.2 t$, and $e^2 / ( \epsilon_0 t a ) = 0.5$.
241: The inset shows the details of the electronic density near the edge. Bottom:
242: Total additional charge per C atom in extended states as function of
243: the ribbon width, $W$.}
244: \end{figure}
245:
246: We can use the ribbon geometry described above to analyze the properties of
247: graphene in the integer QHE.
248: A magnetic field $B$ changes the phases of the hopping terms in the
249: Hamiltonian (\ref{hamil}), leading
250: to the appearance of bands of degenerate Landau levels. In the continuum
251: limit the Landau levels have energy
252: $\epsilon_n = \pm v_{\rm F} l_{\rm B}^{-1} \sqrt{n}$, where $n$ is a positive
253: integer, $l_{\rm B} = \sqrt{\Phi_0 / B}$ is the cyclotron radius, and
254: $\Phi_0 =h/e$ is flux quanta. The Hall conductance can be obtained from
255: the number of bulk states that cross the Fermi level due to the presence
256: of an edge \cite{bert}. The calculated energy levels of a graphene ribbon
257: in a magnetic field are shown in Fig.[\ref{ribbon}].
258: Notice that the momentum along the ribbon fixes the
259: position of the levels. When the position approaches the edges, the positive
260: energy levels rise in energy as for electron-like Landau levels in the 2D
261: electron gas, while the negative energy levels behave in a hole-like
262: fashion. There is a zero energy mode that splits into a set of electron and hole-like levels. Finally, the localized states at the edges are quite insensitive
263: to the applied magnetic field. If we fix the chemical potential above (below)
264: the zero mode energy, all bulk electron-like (hole-like) levels of lower energy give
265: rise to crossings, and contribute to the Hall conductance.
266: Each Landau level is doubly degenerated since the honeycomb lattice gives
267: rise to two inequivalent sets of Dirac fermions \cite{schakel}. Hence, the Hall conductance
268: arising from the bulk modes is:
269: \begin{eqnarray}
270: \sigma_{{\rm Hall}} = 2 (2 N +1) e^2/h \label{Hall} \, ,
271: \end{eqnarray}
272: and thus the Hall conductivity is quantized in odd number of $2 e^2/h$ (the
273: factor of $2$ comes from the spin degeneracy since cyclotron energy,
274: $\hbar \omega_c = \sqrt{2} v_{\rm F} \hbar/l_{\rm B}$, is much larger than the Zeeman
275: energy, $g \mu_B B$, where $g \approx 2$ and $\mu_B$ the Bohr magneton).
276: This analysis neglects corrections due to the additional band of localized
277: states at the edges induced by the structure of the ribbon.
278:
279: \hspace{1.5cm}
280: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
281: % Figure
282: %%%%%%%%%%%%%%%
283: \begin{figure}[htf]
284: %\begin{center}
285: \includegraphics[width=7cm,angle=-90]{boundary_field_0.2_3.ps}
286: \caption{
287: \label{ribbon}
288: Energy levels of a graphene ribbon of width 600 $a$ with
289: zig-zag edges as function of the momentum parallel to the edges.
290: Energy in units of $t$, and $t'=0.2t$. Momentum in units of $1/ (2 \sqrt{3} a)$.
291: Top: zero magnetic field; Centre: a magnetic flux per C atom of
292: $\Phi= 0.00025 \, \Phi_0$; Bottom: $\Phi = 0.0005 \, \Phi_0$.}
293: \end{figure}
294: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
295:
296:
297:
298:
299: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
300: % Figure
301: %%%%%%%%%%%%%%%
302: \begin{figure}[htf]
303: \begin{center}
304: \includegraphics*[width=8cm]{Fig_Nature_dos_kernel.eps}
305: \end{center}
306: \caption{
307: \label{dos} {\bf (a)}: Density of states as function of vacancy
308: concentration ($D$ is a high energy cut-off); {\bf (b)}:
309: As in (a), in the presence of an applied field. Dashed lines mark the
310: positions of the Landau levels in the absence of disorder;
311: {\bf (c)} and {\bf (d)}: $\sigma(\mu)$ as function of the Fermi energy $\mu$.
312: $t'=0$ in all cases.}
313: \end{figure}
314: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
315:
316: While extended defects induce self-doping they do not change
317: significantly the electronic properties. However, point
318: defects such as vacancies or ad-atoms can change the transport
319: properties even if they do not lead to self-doping. It is
320: known that weaker types of disorder do not change the semi-metallic nature of
321: a system described by the Dirac equation \cite{hirsch}.
322: Nevertheless, point defects in the unitary limit can lead to a sharp
323: resonance at the Fermi energy \cite{hirsch}.
324: We describe the effect of vacancies in the transport properties
325: using the Coherent Phase Approximation (CPA), that
326: gives a good description of the spectral and
327: transport properties of graphene,
328: except perhaps within a small region near Fermi energy where electronic
329: localization becomes important \cite{fisher}. Within CPA, the
330: one-particle spectral function is written as:
331: \begin{equation}
332: A ( {\bf k} , \omega ) = {\rm Im} \left[
333: \omega - \Sigma_{\rm CPA} ( \omega ) -
334: \epsilon_{\bf k} \right]^{-1}
335: \label{CPA}
336: \end{equation}
337: where electron self-energy, $\Sigma_{\rm CPA} ( \omega )$, has to be determined self-consistently
338: from (\ref{hamil}) and (\ref{disorder}). The electronic density of states,
339: $\rho(\omega) = \sum_{{\bf k}} A({\bf k},\omega)$,
340: as function of applied magnetic field and impurity concentration, in the continuum
341: limit, is shown in
342: Fig.[\ref{dos}](a)-(b). Unlike for weaker forms of disorder, vacancies induce a
343: finite density of states at the Fermi level. Furthermore, ${\rm Im}
344: \Sigma_{\rm CPA} ( \omega )$ is finite at the Fermi level, and is monotonically
345: {\it increasing} as $\omega \to 0$ indicating that vacancies have a strong effect
346: on the Dirac fermions. This function can be considered an intrinsic linewidth
347: as measured in ARPES. Besides the effect of disorder, the lifetime of quasiparticles
348: has also an intrinsic contribution from interaction effects associated with eq.~(\ref{interaction})
349: giving a contribution
350: $e^2 / v_{\rm F} \, | \omega |$\cite{GGV96}, which is monotonically {\it decreasing}
351: as $\omega \to 0$. Therefore, the final result, as shown in Fig.[\ref{lifetime}], predicts that
352: that the total linewidth of quasiparticles shows a minimum.
353:
354: \begin{figure}[htb]
355: \begin{center}
356: \includegraphics[width=6cm]{trimmer.eps}
357: \end{center}
358: \caption{\label{lifetime}$|\Sigma''(\omega)|$, as a function of energy
359: $\omega$. Inset: $A({\bf k},\omega)$.
360: %(momentum
361: %can be converted into energy by multiplying it by $\hbar v_F$)
362: }
363: \end{figure}
364:
365:
366: The frequency dependent conductivity can be written as:
367: \begin{equation}
368: \sigma(\omega,T) = 2e^2/(\hbar \pi)
369: \int_{- \infty}^\infty d \epsilon K (\epsilon,\omega)
370: [f(\epsilon,T)-f(\epsilon+\omega,T)]/\omega
371: \label{conductivity_dc}
372: \end{equation}
373: where $K (\epsilon,\omega)$ is a conductivity kernel,
374: and $f(\epsilon,T)$ is the Fermi-Dirac distribution
375: function. At low temperatures, the d.c. conductivity,
376: $\sigma(\mu)=\sigma(\omega=0,T)$,
377: is approximately proportional to $K(\mu,\omega=0)$, and can be experimentally
378: measured by the application of a gate voltage, $V$, to a graphene plane
379: (in the presence of a voltage $V$ the chemical potential shifts from $\mu$
380: to $\mu+e V$). In Fig. [\ref{dos}] (c)-(d)
381: we show $\sigma(\mu)$. A non-conventional
382: $\mu$ dependence is found when compared to other 2D electronic systems,
383: in agreement with experiments \cite{Netal04}. As the temperature rises, the
384: d.c. conductivity is found to increase with
385: temperature, as the thermally excited carriers contribute (as in the
386: case of a narrow gap semiconductor) and is also in agreement with the
387: experimental data \cite{Netal04}. When $\mu+e V=0$ the low
388: temperature conductance per plane has a universal value, independent of
389: disorder and magnetic field \cite{F86}, $\sdc = 4 e^2 / ( \pi h )$. A similar
390: universal behavior was predicted for d-wave superconductors \cite{L93}.
391: Because of particle-hole symmetry breaking, $t' \neq 0$,
392: we expect an asymmetry in the
393: Shubnikov-de Haas oscillations as function of gate voltage.
394: We can quantify this asymmetry by looking at Landau levels $j^*$ and $-j^*-1$ with
395: $j^* \gg 1$. It is possible to show \cite{next} that:
396: \begin{equation}
397: \left|t'/t \right| \approx \sqrt{2}/12 (j^*+3/2)^{-1} (j^*+1)^{-1/2} (\lB/a) \,.
398: \end{equation}
399: For $j^* \approx 6$ for $B = 12$ T ($\lB = 52$ \AA) \cite{Netal04},
400: we find $| t' / t | \approx 0.2$, which is the value used in this work.
401:
402: In Fig.[\ref{sigma_W_T_B}] we show $\sigma(\omega)$ in the presence
403: of impurities and finite magnetic field.
404: The existence oscillations in the conductivity at low temperatures is
405: clearly seen. An analysis of the transitions to the different peaks shows that
406: those involving the $j=0$ Landau level are suppressed and that the
407: energy of the lowest
408: Landau levels is significantly shifted by the disorder.
409:
410: \begin{figure}[htf]
411: //
412: \begin{center}
413: \includegraphics[width=7cm]{Fig_Nature_Landau_omega.eps}
414: \end{center}
415: \caption{
416: \label{sigma_W_T_B} $\sigma(\omega)$
417: for different magnetic fields and $n_i =0.001$,
418: using the parameters given in the text.
419: The lower right graph shows a scaling of the curves as
420: function of $\omega/\sqrt{B}$.}
421: \end{figure}
422:
423: We have analyzed the influence of lattice defects on the electronic
424: properties of graphene. Our results show that:
425: ({\it i}) Extended defects lead to the existence of localized states, or resonances, near the
426: Fermi level of the pure system. In the absence of electron-hole symmetry,
427: these states lead to self-doping, and to the existence of electron or hole Fermi
428: pockets with $10^{-4} - 10^{-5}$ electrons per unit cell for domains of sizes $0.1 - 1 \mu$m.
429: ({\it ii}) An integer QHE with a quantized Hall conductivity in odd values
430: of $2 e^2/h$.
431: ({\it iii}) Point defects lead to a finite inverse scattering time in undoped
432: graphene, which decreases with increasing energy. As the inelastic
433: contribution rises linearly with temperature, a minimum in the lifetime
434: observed in ARPES experiments is predicted.
435: ({\it iv}) The d.c. conductivity in undoped graphene has a universal value at low
436: doping and temperatures. In undoped samples, it rises with temperature, as
437: the number of thermally activated carriers increases.
438: ({\it v}) The optical conductivity in a magnetic field is modulated
439: due to Landau levels.
440:
441: N. M. R. P. and F. G. acknowledge
442: the Quantum Condensed Matter Visitor's Program at Boston University for
443: support. N. M. R. P. acknowledges Funda\c{c}\~ao para
444: a Ci\^encia e Tecnologia for a sabbatical grant.
445: A. H. C. N. was supported by the NSF grant DMR-0343790.
446: We thank D. N. Basov, A. Geim, G.-H. Gweon, A. Lanzara, and S. W. Tsai for
447: illuminating discussions.
448:
449: %%%%%%%%%%%%%%%%%%%%
450: % SECTION %
451: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
452:
453: %%%%%%%%%%%%%%%%%%%%
454: % bibliography %
455: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
456: \begin{thebibliography}{99}
457:
458: \bibitem{BCP88}
459: N. B. Brandt, S. M. Chudinov, and Ya. G. Ponomarev,
460: in {\it Modern Problems in Condensed Matter Sciences},
461: V. M. Agranovich and A. A. Maradudin, eds.
462: (North Holand, Amsterdam, 1988).
463:
464: \bibitem{nanotubes}
465: S. G. Rao {\it et al.}
466: Nature {\bf 425}, 36 (2004).
467:
468: \bibitem{buckyball}
469: M. Sawamura {\it et al.}, Nature {\bf 419}, 702 (2002).
470:
471:
472: \bibitem{Netal04}
473: K. S. Novoselov {\it et al.},
474: Science {\bf 306}, 666 (2004);
475: K. S. Novoselov {\it et al.},
476: cond-mat/0503533;
477: A. K. Geim {\it et al.}, unpublished.
478:
479: \bibitem{outros}
480: C.~Berger {\it et al.}, cond-mat/0410240;
481: Y.~Zhang {\it et al.}, cond-mat/0410314;
482: Y.~Zhang {\it et al.}, cond-mat/0410315.
483:
484:
485: %\bibitem{pauling}
486: %Linus Pauling,
487: %in {\it The Nature of the Chemical Bond},
488: %(Cornell University Press, Ithaca, 1972) pp. 234-235.
489:
490:
491:
492: \bibitem{DS00}
493: L. Davison, and D. Sherrington,
494: J. Phys. A {\bf 33}, 8615 (2000).
495:
496: \bibitem{RS03}
497: F. Ritort, and P. Sollich,
498: Adv. in Phys. {\bf 52}, 219 (2003).
499:
500:
501:
502: \bibitem{Hetal04b}
503: A. Hashimoto {\it et al.}, Nature {\bf 430}, 870 (2004).
504: %; T. Matsui, {\it et al.}, cond-mat/0405011;
505: %S. Banarjee, {\it et al.}, cond-mat/0504629.
506:
507: \bibitem{magnet}
508: P. Esquinazi {\it et al.},
509: Phys. Rev. Lett. {\bf 91}, 227201 (2003).
510: %; see also papers in
511: %{\it Carbon-Based Magnetism: an overview of metal free carbon-based
512: %compounds and materials}, T. Makarova and F. Palacio, eds. (Elsevier,
513: %Amsterdam, 2005).
514:
515: \bibitem{WS00}
516: K. Wakayabashi, and M. Sigrist,
517: Phys.Rev.Lett. {\bf 84}, 3390 (2000).
518: K. Wakayabashi,
519: Phys. Rev.B {\bf 64}, 125428 (2001).
520:
521: \bibitem{Vetal05}
522: M- A. H. Vozmediano, {\it et al}, cond-mat/0505557.
523:
524: \bibitem{GGV92}
525: J. Gonz\'alez {\it et al.},
526: %F. Guinea, and M. A. H. Vozmediano,
527: Phys. Rev. Lett. {\bf 69}, 172 (1992); {\it ibid},
528: Phys.Rev.B {\bf 63}, 134421 (2001).
529:
530: \bibitem{CEL96}
531: J.C. Charlier {\it et al.},
532: %T. W. Ebbesen, and Ph. Lambin,
533: Phys.Rev.B {\bf 53}, 11108 (1996).
534:
535: \bibitem{MA01}
536: H. Matsumura and T. Ando,
537: J. Phys. Soc. Japan {\bf 70}, 2657 (2001).
538:
539: \bibitem{DSL04}
540: E. J. Duplock {\it et al.},
541: %M. Scheffler, P. J. D. Lindan,
542: Phys. Rev. Lett. {\bf 92}, 225502 (2004).
543:
544: \bibitem{hirsch}
545: See, for instance, P. J. Hirschfeld, and W. A. Atkinson,
546: J. Low. Temp. Phys. {\bf 126}, 881 (2002).
547:
548: \bibitem{bert}
549: B. I. Halperin, Phys. Rev. B {\bf 25}, 2185 (1982).
550:
551: \bibitem{schakel}
552: For a single Dirac fermion, see, A. M. J. Schakel, Phys. Rev. D {\bf 43}, 1428 (1991).
553:
554: \bibitem{GGV96}
555: J. Gonz\'alez {\it et al.},
556: %F. Guinea, and M. A. H. Vozmediano,
557: Phys. Rev. Lett. {\bf 77}, 3589 (1996).
558:
559: \bibitem{fisher}
560: T. Senthil and M. P. A. Fisher,
561: Phys. Rev. B {\bf 60}, 6893 (1999).
562:
563: \bibitem{F86}
564: E. Fradkin, Phys. Rev. B {\bf 33}, 3257 (1986).
565:
566: \bibitem{L93}
567: P. A. Lee, Phys. Rev. Lett. {\bf 71}, 1887 (1993).
568:
569: \bibitem{next}
570: N. M. R. Peres {\it et al.}, unpublished.
571:
572:
573: \end{thebibliography}
574:
575: \end{document}
576: