1: %{\mbox{\boldmath ${\bf \sigma}$}}
2: %\documentclass[aps,prl,twocolumn,groupedaddress]{revtex4}
3: \documentclass[prl,aps,twocolumn,showpacs,superscriptaddress,%
4: floatfix]{revtex4}
5: \usepackage{graphicx}
6: \usepackage{bm}
7: \usepackage{amsmath}
8: \begin{document}
9: %Title
10: \title{Half-skyrmion picture of single hole doped high-T$_c$ cuprate}
11: \author{Takao Morinari}
12: \affiliation{Yukawa Institute for Theoretical Physics, Kyoto
13: University, Kyoto 606-8502, Japan}
14: \date{\today}
15: \begin{abstract}
16: % insert abstract here
17: The single hole doped CuO$_2$ plane
18: %$S=1/2$ antiferromagnetic Heisenberg model on the square lattice
19: is studied by the O(3) non-linear $\sigma$ model and its CP$^1$
20: representation.
21: It is argued that the spin configuration
22: around a Zhang-Rice singlet is described by a half-skyrmion.
23: The form of the excitation spectrum of the half-skyrmion is the same
24: as that in the $\pi$-flux phase.
25: \end{abstract}
26: % insert suggested PACS numbers in braces on next line
27: \pacs{75.25.+z,74.72.-h}
28: % insert suggested keywords - APS authors don't need to do this
29: %\keywords{}
30: %\maketitle must follow title, authors, abstract, \pacs, and \keywords
31: \maketitle
32:
33: % body of paper here - Use proper section commands
34: % References should be done using the \cite, \ref, and \label commands
35: In the phase diagram of high-temperature superconductors,
36: the most established phase is apparently the antiferromagnetic
37: long-range ordered phase, or the N{\' e}el ordered phase
38: of undoped parent compounds.
39: %AF Heisenberg model
40: Due to a large charge-transfer gap\cite{IFT}, a hole at each
41: copper site in the CuO$_2$ plane is localized to form a spin $S=1/2$
42: moment.
43: There is an antiferromagnetic superexchange interaction $J$
44: between the copper site spins and the system is described by the
45: antiferromagnetic Heisenberg model on the square lattice,
46: $H_{\rm AFH}=J\sum_{\langle i,j\rangle} {\bf S}_i \cdot {\bf S}_j$.
47: Antiferromagnetic spin wave dispersions observed by neutron scattering
48: experiments are in
49: quite good agreement with the spin wave theory with the value of $J$
50: determined by Raman scattering\cite{MASON}.
51: High-temperature superconductivity occurs upon moderate hole doping
52: on the N{\' e}el ordering phase.
53:
54:
55: For understanding the mechanism of high-temperature
56: superconductivity,
57: it is necessary to describe the doped holes.
58: However, for the purpose of figuring out the proper picture,
59: considering a moderately doped system does not appear
60: to be promising because such a system is quite complicated and
61: indeed there is no consensus on
62: the description of the system.
63: Contrastingly, a system with only one doped hole is much simpler
64: and there seems to be much hope to understand the physics
65: because it is closest to the well-established antiferromagnetic
66: long-range ordered phase.
67: Experimentally angle resolved photoemission spectroscopy (ARPES)
68: \cite{WELLS_ETAL,RONNING_ETAL,SHEN_RMP} on undoped
69: compounds provide valuable information on the single hole doped
70: system\cite{SHEN_RMP}.
71:
72:
73: As an effective theory of the antiferromagnetic Heisenberg model
74: $H_{\rm AFH}$, the O(3) non-linear $\sigma$ model (NL$\sigma$M) has
75: been studied extensively\cite{CHN}.
76: The model is derived from $H_{\rm AFH}$ by applying Haldane's
77: mapping\cite{HALDANE}:
78: \begin{equation}
79: S=\frac{\rho_s^0}{2} \int_0^{\beta} d\tau \int d^2 {\bf r}
80: \left[ \left( \nabla {\bf n} \right)^2
81: + \frac{1}{c_{\rm sw}^2}
82: \left( \frac{\partial {\bf n}}{\partial \tau} \right)^2
83: \right],
84: \label{eq_nlsm}
85: \end{equation}
86: where $\rho_s^0$ is the bare spin stiffness constant
87: on the length scale of the lattice constant,
88: which is taken to be unity,
89: and $c_{\rm sw}$ is the spin-wave velocity.
90: The integral with respect to the imaginary time $\tau$
91: is carried out up to the inverse temperature $\beta$.
92: (Hereafter we take the unit $\hbar = 1$.)
93: The unit vector ${\bf n}$ represents the directions of staggered spin
94: moments.
95: The low-energy and the long-length scale behaviors of $H_{\rm AFH}$
96: are well described by NL$\sigma$M as confirmed
97: by experiments \cite{MASON} and quantum Monte Calro
98: simulations\cite{QMC}.
99:
100:
101: There is a useful form for the NL$\sigma$M, which is called CP$^1$
102: model\cite{RAJARAMAN}.
103: Introducing complex fields $\zeta_{\sigma}(x)$ with $x=(c_{\rm
104: sw}\tau, {\bf r})$ through
105: ${\bf n}=\sum_{\sigma,\sigma=\uparrow,\downarrow}
106: \overline{\zeta}_{\sigma}
107: {\mbox{\boldmath ${\bf \sigma}$}}_{\sigma\sigma'}
108: \zeta_{\sigma'}$
109: with $\sigma^{\alpha}$($\alpha=x,y,z$) representing Pauli matrices,
110: we obtain $S=(\rho_s^0/2c_{\rm sw})\int d^3 x \left[ |\partial_{\mu}
111: \zeta_{\sigma}|^2+\left(\overline{\zeta}_{\sigma}
112: \partial_{\mu} \zeta_{\sigma} \right)^2 \right]$.
113: Performing a Stratonovich-Hubbard transformation for the second term
114: in the square brackets, we obtain the CP$^1$ model:
115: \begin{equation}
116: S=\frac{1}{g}\int d^3 x |\left( \partial_{\mu} -i \alpha_{\mu}
117: \right)\zeta_{\sigma}(x)|^2,
118: \label{eq_cp1}
119: \end{equation}
120: where $g=c_{\rm sw}/2\rho_s^0$ and
121: $\langle \alpha_{\mu}(x) \rangle=
122: \langle i\sum_{\sigma} \overline{\zeta}_{\sigma}(x)\partial_{\mu}
123: \zeta_{\sigma} (x) \rangle$
124: at the saddle point.
125: The field $\alpha_{\mu}$ is a U(1) gauge field associated with a
126: local gauge symmetry,
127: $\zeta_{\sigma}(x) \rightarrow \zeta_{\sigma}(x) \exp(i\chi(x))$.
128: Thus, the NL$\sigma$M is equivalent to the boson system with the U(1)
129: gauge field interaction.
130:
131:
132: The CP$^1$ model is also derived from
133: the Schwinger boson mean field theory (SBMFT)\cite{AA}.
134: In the theory, the spin is represented in terms of boson operators
135: $z_{j\sigma}^{\dagger}$ and $z_{j\sigma}$ as
136: ${\bf S}_j =\frac12 \sum_{\sigma,\sigma'}
137: z_{j\sigma}^{\dagger}
138: {\mbox{\boldmath ${\bf \sigma}$}}_{\sigma \sigma'}
139: z_{j\sigma'}$
140: with the constraint $\sum_{\sigma=\uparrow,\downarrow}
141: z_{j\sigma}^{\dagger} z_{j\sigma}=1$.
142: The SBMFT Hamiltonian is given by
143: $H_{\rm SB} = -(J/2)\sum_{\langle i,j \rangle}
144: \left[ \Delta^* \left( z_{i\uparrow}z_{j\downarrow}
145: - z_{i\downarrow}z_{j\uparrow} \right) + h.c. -|\Delta|^2
146: \right]+\sum_{j\sigma} \lambda
147: \left(z_{j\sigma}^{\dagger} z_{j\sigma}-1\right)$,
148: with $\Delta$ and $\lambda$ being the mean field parameters.
149: Here $\Delta = \langle z_{i\uparrow}z_{j\downarrow}
150: - z_{i\downarrow}z_{j\uparrow} \rangle$
151: describes the Schwinger boson
152: pairing\cite{READ_SACHDEV,CHUBUKOV,NG95}.
153: After Fourier transforming,
154: the Schwinger boson quasi-particle excitation spectrum is obtained by
155: a Bogoliubov transformation\cite{YOSHIOKA}:
156: $\omega_k = \sqrt{\lambda^2 - J^2|\Delta|^2
157: \left( \sin k_x \pm \sin k_y \right)^2}$,
158: where the plus sign is for $k_x k_y >0$ and
159: the minus sign is for $k_x k_y <0$.
160: In the N{\' e}el ordering phase, $\lambda=2J|\Delta|$ and
161: the Schwinger bosons are gapless at zone face centers
162: $(\pm \pi/2,\pm \pi/2)$.
163: The N{\' e}el ordering state is a consequence of Bose-Einstein
164: condensation
165: of the Schwinger bosons at those points\cite{YOSHIOKA,BEC}.
166: One can derive the CP$^1$ model after some
167: algebra\cite{READ_SACHDEV}.
168: The relation between the Schwinger boson fields and
169: $\overline{\zeta}_{\sigma}(x)$ and $\zeta_{\sigma}(x)$
170: in Eq.(\ref{eq_cp1}) is
171: $z_{j\sigma}=\zeta_{j\sigma}$ at one sublattice and
172: $\zeta_{j\sigma}=z_{j\sigma}^{\dagger}$ at the other sublattice.
173: One Schwinger boson carries spin $1/2$ and the unit gauge charge.
174: Therefore, pairs of the Schwinger bosons that describe the low-lying
175: excitations carry integer spins and the gauge charge two.
176:
177:
178:
179: Both of NL$\sigma$M and SBMFT are, so to speak, bosonic description of
180: the system.
181: Meanwhile, one can construct a fermionic theory by representing
182: the $S=1/2$ spins in terms of fermions.
183: Mean field analysis was carried out by Affleck and
184: Marston\cite{AFFLECK_MARSTON} and the $\pi$-flux phase was proposed.
185: The state is characterized by a flux $\pi$ penetrating each plaquette
186: with alternating directions.
187: The quasiparticle excitations are gapless at $(\pm \pi/2, \pm \pi/2)$.
188: %Although the ground state energy of this state is slightly higher than
189: %the N{\' e}el ordering state, introducing the non-zero staggered
190: %magnetization lowers the energy\cite{HSU}.
191: More advanced analysis based on an effective theory that
192: includes fluctuations around the mean field state through a U(1) gauge
193: field shows that a mass term is induced in the spectrum dynamically.
194: This phenomenon is associated with the N{\' e}el
195: ordering\cite{MARSTON,KIM_LEE}.
196:
197:
198: It has been shown experimentally that doped holes reside primarily
199: on oxygen sites.
200: %There is a strong hybridization between oxygen p-orbital states and
201: %copper d-orbital states.
202: Zhang and Rice suggested\cite{ZHANG_RICE} that there is
203: strong correlations of forming a singlet pair between oxygen
204: p-orbital holes and copper d-orbital holes.
205: The t-J model was proposed based on this picture.
206: Although the single hole problem has been studied mostly by this
207: model, the focus is mainly on frustration effects induced by hopping
208: of the Zhang-Rice singlet\cite{SHRAIMANN_SIGGIA,KANE_LEE_READ}.
209: However, not so much attention has been paid on the spin configuration
210: around the Zhang-Rice singlet.
211:
212:
213:
214: In this paper, it is argued that in the single hole doped system the
215: doped hole induces a half-skyrmion which is schematically shown in
216: Fig.~\ref{fig_hs}.
217: The half-skyrmion is a spin texture
218: characterized by a half of a topological charge.
219: \begin{figure}[bp]
220: \includegraphics[scale=0.8]{fig_hs.eps}
221: \caption{The half-skyrmion spin texture. Arrows indicate the
222: staggered moment and the filled circle at the center represents the
223: Zhang-Rice singlet formed site.}
224: \label{fig_hs}
225: \end{figure}
226:
227:
228:
229: % half-skyrmion solution
230: We start with considering the spin configuration around a static
231: Zhang-Rice singlet.
232: The moving case shall be considered later.
233: A Zhang-Rice singlet is composed of a d-orbital hole state and the
234: four oxygen hole states around the copper ion.
235: Constructing the Wannier functions, the operator for the
236: Zhang-Rice singlet is given by\cite{ZHANG_RICE}
237: $\frac{1}{\sqrt{2}}
238: \left( \phi_{j\uparrow} d_{j\downarrow} - \phi_{j\downarrow}
239: d_{j\uparrow} \right)$,
240: where $\phi_{j\sigma}$ is the symmetric
241: combination state of the four oxygen hole states around the copper ion
242: at the site $j$\cite{ZHANG_RICE}.
243: The d-orbital state of the Zhang-Rice singlet is given by
244: the superposition of the up-spin state and the down-spin state.
245: To be specific, we assume that
246: in the N{\' e}el ordered state before hole doping the ordered spin
247: moment at the site $j$ is in the direction of the positive $z$-axis.
248: Then, the d-orbital state with up-spin does not affect
249: the neighboring spins so much.
250: % and just makes contribution to
251: %the N{\' e}el ordered state.
252:
253:
254: In contrast, the down-spin state affects the neighboring spins to
255: change their directions.
256: From the analysis of the NL$\sigma$M below, we show that a spin
257: configuration, which is characterized by a non-zero topological
258: charge, is created around the down-spin state.
259: This spin configuration is called skyrmion\cite{RAJARAMAN}.
260: %(The up-spin state at the symmetric four oxygen states favors this
261: %skyrmion spin configuration because it gives rise to ferromagnetic
262: %coupling \cite{Aharony88} of the copper spin to the nearest
263: %neighbor copper spins.)
264: For the static case, the energy is given by
265: $
266: E=\frac{\rho_s^0}{2}\int d^2 {\bf r}
267: \left( \nabla {\bf n} \right)^2.
268: $
269: The field equation is $\nabla^2 {\bf n}-\left( {\bf n} \cdot \nabla^2
270: {\bf n} \right) {\bf n}=0$.
271: Solutions are devided into sectors characterized by a topological
272: charge : $
273: Q=\frac{1}{8\pi} \int d^2 {\bf r} \epsilon_{\mu \nu} {\bf n} \cdot
274: \left( \partial_{\mu} {\bf n} \times \partial_{\nu} {\bf n} \right)$,
275: where $\epsilon_{xy}=-\epsilon_{yx}=1$ and
276: $\epsilon_{xx}=\epsilon_{yy}=0$.
277: The energy in each sector has the lower bound:
278: $E\geq 4\pi \rho_s^0 |Q|$.
279: The equality is satisfied if and only if
280: $\partial_{\mu} {\bf n} = \pm \epsilon_{\mu \nu} \left( {\bf n} \times
281: \partial_{\nu} {\bf n} \right)$\cite{BELAVIN_POLYAKOV}.
282: The lowest enegy state is obtained by solving this equation with the
283: boundary conditions:
284: ${\bf n}\rightarrow +{\hat e}_z$ at infinity and
285: ${\bf n}= -{\hat e}_z$ at ${\bf r} = {\bf r}_j$.
286: %Here the ordered staggered magnetization is assumed to be in the
287: %positive $z$-direction.
288: In terms of a variable $w\equiv (n_x + i n_y)/(1-n_z)$, the equation
289: turns out to be the Chaucy-Riemann
290: equations\cite{BELAVIN_POLYAKOV}.
291: Therefore, $w$ is an analytic function of $z=x+iy$ or
292: $z^*=x-iy$.
293: The lowest energy state satisfying the boundary conditions is
294: $w_s=[(x-x_j)+i(y-y_j)]/\lambda$ and
295: $w_{as}=[(x-x_j)-i(y-y_j)]/\lambda$ with
296: $\lambda$ being a constant.
297: In the vector ${\bf n}$ representation, these solutions are
298: \begin{equation}
299: {\bf n}=\left(
300: \frac{2\lambda (x-x_j)}{|{\bf r}-{\bf r}_j|^2+\lambda^2},
301: \pm \frac{2\lambda (y-y_j)}{|{\bf r}-{\bf r}_j|^2+\lambda^2},
302: \frac{|{\bf r}-{\bf r}_j|^2-\lambda^2}{|{\bf r}-{\bf
303: r}_j|^2+\lambda^2}
304: \right),
305: \label{eq_sky}
306: \end{equation}
307: where the plus sign is for $\omega_s$ and the minus sign is for
308: $w_{as}$.
309:
310:
311: Now we consider the superposition of the uniform state
312: ${\bf n}=+{\hat e}_z$ for the up-spin state and the
313: skyrmion spin configuration (\ref{eq_sky})
314: created by the down-spin state.
315: The NL$\sigma$M analysis suggests that a spin
316: configuration with non-zero topological charge $Q$, where $0<|Q|<1$,
317: is formed with
318: %$\lambda \sim (\rm the~~lattice~~constant)$ and
319: $n_{\ell z}\geq 0$ for any site $\ell$.
320: The value of the topological charge $Q$ is determined
321: through the flux representation in terms of the
322: CP$^1$ gauge field $\alpha_{\mu}$:
323: $
324: Q=\frac{1}{2\pi} \int d^2 {\bf r} \left( \nabla \times
325: {\mbox{\boldmath ${\bf \alpha}$}} \right).
326: $
327: The spin configuration with $Q$
328: corresponds to the flux $2\pi Q$.
329: Due to Bose-Einstein condensation of the bosons $\zeta_{\sigma}(x)$,
330: the value of the flux $2\pi Q$ is not arbitrary.
331: Since low-lying excitations are pairs of these bosons, the flux
332: quantum is $\pi$ analogous to conventional BCS superconductors.
333: From the constraint on $Q$, i.e., $0<|Q|<1$, we conclude that $2\pi
334: |Q|=\pi$, that is, $|Q|=1/2$.
335: Therefore, the spin configuration is
336: the half-skyrmion, which is equivalent to the skyrmion with the
337: core, and its energy is $E_0=2\pi\rho_s^0$.
338: %(This energy does not include, of course, the core energy.)
339: This conclusion can be reached if we assume that the staggered
340: magnetization vanishes at the perimeter of the core.
341: The analysis of the NL$\sigma$M gives the half-skyrmion solution
342: as was shown by Saxena and Dandoloff\cite{SAXENA_DANDOLOFF}.
343: They considered the two-dimensional ferromagnetic Heisenberg model
344: in the context of the quantum Hall systems.
345: For the static case, the field equation is the same.
346: So we can apply their analysis to the antiferromagnetic case as well.
347:
348:
349: %The excitation energy $E_0=m c_{\rm sw}^2$ associated
350: %with the half-skyrmion formation is calculated as follows
351: %Since the Zhang-Rice singlet formed site does not contribute to the
352: %topological charge, $E_0$ is given by the summation of the
353: %energies over the region without this site.
354: %In such a region, we may use the relation between $Q$ and the energy.
355: %Therefore, we find $E_0=2\pi\rho_s^0$.
356:
357:
358: The moving half-skyrmion is
359: obtained from the static half-skyrmion by applying
360: a Lorentz transformation
361: because the action (\ref{eq_nlsm}) and (\ref{eq_cp1}) are Lorentz
362: invariant with $c_{\rm sw}$ the speed of ``light.''
363: From the calculation of the energy-momentum tensors, we find
364: $\epsilon_k^0 = \sqrt{c_{\rm sw}^2 (k_x^2+k_y^2)
365: + E_0^2}$.
366: For the $S=1/2$ antiferromagnet, the bare spin-wave velocity
367: is $c_{\rm sw}=\sqrt{2}J$ and the bare spin stiffness is
368: $\rho_s^0=J/4$.
369: Renormalization effect leads to
370: $c_{\rm sw} \rightarrow \sqrt{2}Z_c J$
371: and $\rho_s^0 \rightarrow Z_{\rho} J/4 (\equiv \rho_s)$\cite{CHN}.
372: Thus, the renormalized dispersion is given by
373: $\epsilon_k = J\sqrt{2Z_c^2 (k_x^2+k_y^2) + (\pi Z_{\rho}/2)^2}$.
374: Since the doped hole is fermion,
375: this relativistic dynamics of the half-skyrmion is described by
376: a Dirac fermion Lagrangian density:
377: \begin{equation}
378: {\cal L} =\sum_{\sigma} \overline{\psi}_{\sigma}
379: \left( \gamma_{\mu} \partial_{\mu} + m c_{\rm sw}^2\right)
380: \psi_{\sigma},
381: \label{eq_dirac}
382: \end{equation}
383: with $\overline{\psi} = \psi^{\dagger} \gamma_0$ and
384: $mc_{\rm sw}^2=\pi Z_{\rho}J/2$.
385: The $\gamma$ matrices satisfy
386: $\left\{ \gamma_{\mu}, \gamma_{\nu} \right\}=2 \delta_{\mu \nu}$
387: in the Euclidean space time.
388: Here $\sigma$ is an index for the sign of the topological charge.
389: The number of components of the Dirac fermion field is
390: four because there are a positive energy state and negative energy
391: state and the origin of the Dirac fermion dispersion is
392: at either ${\bf k}_1=(\pi/2, \pi/2)$ or
393: ${\bf k}_2=(-\pi/2,\pi/2)$.
394: The latter is implied from the fact that the Schwinger bosons are
395: gapless at those points in the symmetry broken N{\' e}el ordering
396: phase.
397: Note that the dynamics is associated with the
398: properties of the parent compound, namely, the Mott insulator.
399: In this sense, the fermion $\psi$ is different from
400: quasi-particles in the conventional Fermi liquid.
401:
402:
403:
404: The lattice version of the action is derived by discretizing
405: Eq.(\ref{eq_dirac}).
406: As is well-known in the lattice gauge field theory\cite{KOGUT83},
407: there is a fermion-doubling problem when one introduces a lattice
408: for a Dirac fermion.
409: This subtlity does not make any difficulty in our case because of the
410: antiferromagnetic correlations that naturally introduce two species
411: of Dirac fermions with half of the Brillouin zone (the magnetic
412: Brillouin zone).
413: Discretizing Eq.(\ref{eq_dirac}) and then performing Fourier
414: transformation, we obtain
415: $
416: {\cal L}=\sum_{\sigma} \overline{\psi}_{k\sigma}
417: \left(
418: \begin{array}{cc}
419: mc_{\rm sw}^2+\partial_{\tau} & \cos k_x + i \cos k_y \\
420: -\cos k_x + i\cos k_y & mc_{\rm sw}^2-\partial_{\tau}
421: \end{array}
422: \right)
423: \psi_{k\sigma}.
424: $
425: Here we have used the fact that the origin of the Dirac fermion
426: dispersion is at ${\bf k}_{1,2}$.
427: The dispersion relation is given by
428: \begin{equation}
429: \epsilon_k^{\pm} = \pm \sqrt{c_{\rm sw}^2
430: (\cos^2 k_x + \cos^2 k_y) + (mc_{\rm sw})^2}.
431: \label{eq_disp}
432: \end{equation}
433: The dispersion curve $\epsilon_k^- - mc_{\rm sw}^2$
434: is qualitatively in good agreement with the experimentally
435: observed dispersion by Wells {\it et al}\cite{WELLS_ETAL,LAUGHLIN97}.
436: In our theory, $Z_c$ and $Z_{\rho}$ are determined by the parameters
437: of the undoped system if one neglects spin wave effects on the
438: self-energy of the half-skyrmion.
439: For the $S=1/2$ antiferromagnetic Heisenberg model,
440: the values of $Z_c$ and $Z_{\rho}$ are estimated as
441: $Z_c = 1.17$ and $Z_{\rho} = 0.72$
442: from quantum Monte Calro simulations\cite{QMC}
443: and a series expansion analysis\cite{SINGH}.
444: Substituting these values into Eq.~(\ref{eq_disp}),
445: we find that the band width along $(0,0)$ to $(\pi,\pi)$
446: is $\sim 1.5J$ and the Dirac fermion mass is $m \sim 1.13J$.
447: Experimentally, the band width is estimated to
448: $2.2J$\cite{WELLS_ETAL}.
449: This $30\%$ discrepancy would be associated with the deviation of the
450: real system from the NL$\sigma$M description.
451: There is also self-energy corrections on the mass $m$
452: due to spin waves.
453: We have estimated it to $m/J = 1.13/(1-0.12 e_A^2)$
454: where $e_A$ is a coupling parameter between the half-skyrmion and
455: the antiferromagnetic spin waves.
456: The coupling parameter $e_A$ turns out to be a gauge charge in the
457: dual theory below.
458:
459:
460: %Now we briefly comment on the half-skyrmion life time.
461: %As mentioned above, NLSM is the effective theory around
462: %$(\pm \pi/2, \pm \pi/2)$.
463: %Since the spin wave dispersion is approximated as
464: %the relativistic dispersion around these points,
465: %the life time of the half-skyrmion is longest at them.
466: %Apart from those points, the spin wave dispersions deviate from
467: %the relativistic form.
468: %That deviation leads to shorter
469: %life times of the half skyrmion.
470: %The effect is expected to be largest around $(\pi,0)$ point.
471: %Analysis of the anomalous line shape of the ARPES
472: %spectrum\cite{SHEN_RMP} from this point of view
473: %is an issue in the future.
474:
475:
476: The effective action for the half-skyrmion can be derived by
477: duality mapping\cite{FISHER_LEE}.
478: In the CP$^1$ theory, the half-skyrmion solution residing at the
479: origin has the following form
480: $
481: z_{\sigma}(x)=\frac{\lambda u_{\sigma}+rv_{\sigma}
482: \exp(\pm i\theta)}{\sqrt{r^2+\lambda^2}},
483: $
484: where ${\bf u}^* \cdot {\bf v}=0$, and $|{\bf u}|=|{\bf v}|=1$.
485: (For the solution (\ref{eq_sky}) with setting ${\bf r}_j=0$,
486: these parameters are
487: $v_{\uparrow}=u_{\downarrow}=1$ and
488: $u_{\uparrow}=v_{\downarrow}=0$.)
489: If we take $|v_{\uparrow}|=|v_{\downarrow}|\neq 0$, then the
490: staggered moments lie in the plane with the same direction at
491: infinity.
492: In this boundary condition, $z_{\sigma}(x)\sim v_{\sigma}
493: \exp(\pm i\theta)$ at $r\gg \lambda$.
494: We write $z_{\sigma}(x)=\rho_{\sigma}^{1/2} \exp(i\phi(x))$,
495: where $\phi(x)=\phi_0(x) +\phi_v(x)$.
496: Here $\phi_0(x)$ is a non-singular function of $x$ and $\phi_v(x)$
497: describes the vortex (half-skyrmion) field:
498: $\phi_v(x) \sim q_v \tan^{-1} (y-y_v)/(x-x_v)$,
499: where $q_v$ is the sign of the topological charge and
500: $(x_v,y_v)$ denotes the position of the vortex.
501: (Note that $x_v$ and $y_v$ are functions of the imaginary time.)
502: In applying the duality mapping, a dual gauge field is introduced.
503: The coupling between the gauge field and the half-skyrmion
504: has the form of the minimal coupling.
505: Thus, we obtain the following QED$_3$ Lagrangian density in the
506: continuum,
507: \begin{equation}
508: {\cal L} = \sum_{\sigma} \overline{\psi}_{\sigma}
509: \left[
510: \gamma_{\mu} \left( \partial_{\mu}
511: - i q_{\sigma} A_{\mu} \right)
512: + m \right]
513: \psi_{\sigma} + \frac{1}{4e_A^2}
514: \left( \partial_{\mu} A_{\nu} - \partial_{\nu} A_{\mu}
515: \right)^2.
516: \label{eq_qed3}
517: \end{equation}
518: The gauge field $A_{\mu}$
519: originally comes from the spin current, which is
520: associated with the antiferromagnetic spin wave modes.
521: The gauge charge $e_A$ is a parameter of the theory.
522: Estimation of $e_A$ is beyond the analysis of the effective theories.
523: For analysis of the half-skyrmion self-energy,
524: we formulate the theory on the lattice according to a standard
525: procedure\cite{KOGUT83}.
526: The second order term with respect to
527: the one ``photon'' vertex part is evaluated numerically and the first
528: order term with respect to the two ``photon'' vertex part is computed
529: analytically.
530: While the former has negligible effect, the latter leads to
531: $\Sigma_k = -(i\sqrt{2}e_A^2/16)\sum_{\mu} \gamma_{\mu}
532: \sin \left( k_{\mu} a \right)
533: \int_{-\pi}^{\pi} \frac{d^2 \tilde{{\bf k}}}{(2\pi)^2}
534: \frac{1}{\sqrt{1-(\cos \tilde{k}_x + \cos \tilde{k}_y)^2/4}}$.
535: The mass renormalization due to this term is
536: $m \rightarrow m/(1-0.12 e_A^2)$.
537:
538: A similar action of Eq.(\ref{eq_qed3})
539: is obtained in the fermionic
540: representation of the $S=1/2$ antiferromagnetic Heisenberg model
541: \cite{MARSTON,KIM_LEE} based
542: on the $\pi$-flux phase\cite{AFFLECK_MARSTON} with dynamically induced
543: mass $m$, through the interaction with a gauge field.
544: (The condition of this dynamical mass generation\cite{APPELQUIST_ETAL}
545: is that the number of Dirac fermion species
546: is less than $32/\pi^2 \sim 3.2$.
547: This is the case for the QED$_3$ theory of the
548: antiferromagnet\cite{KIM_LEE,TESANOVIC}.)
549: From exact diagonalization studies it was shown \cite{GOODING}
550: that the bond spin currents which characterize the $\pi$-flux phase is
551: reproduced by a skyrmion-like spin texture\cite{SKYRMIONS}.
552: From the view point of the spin-charge separation\cite{ANDERSON},
553: Baskaran suggested\cite{BASKARAN} that a half-skyrmion can be seen
554: as deconfined spinons based on an analysis of an $n$-skyrmion
555: solution.
556:
557:
558: In summary, we have argued that, within the effective theory
559: approaches to the $S=1/2$ Heisenberg antiferromagnet, the doped hole
560: induces the half-skyrmion spin texture through the formation of the
561: Zhang-Rice singlet in the single hole doped system.
562: This picture is consistent with rapid suppression of the N{\' e}el
563: ordering by the doped holes because the half-skyrmion behaves as a
564: vortex in the condensate.
565: It would be interesting to extend the picture to the slightly doped
566: regime.
567:
568:
569: I would like to thank G.~Baskaran, X.~Dai,
570: M.P.A.~Fisher, L.~Balents, N.~Nakai,
571: K.~Shizuya, and A.~Tanaka for valuable discussions.
572: I also thank Z.~Te\u{s}anovi\'{c} for helpful comments on the
573: manuscript.
574: This work was supported in part by the Grant-in-Aid for the 21st
575: Century COE "Center for Diversity and Universality in Physics" from
576: the Ministry of Education, Culture, Sports, Science and Technology
577: (MEXT) of Japan.
578:
579: % Put \label in argument of \section for cross-referencing
580: %\section{\label{}}
581: \subsection{}
582: \subsubsection{}
583:
584: % If in two-column mode, this environment will change to single-column
585: % format so that long equations can be displayed. Use
586: % sparingly.
587: %\begin{widetext}
588: % put long equation here
589: %\end{widetext}
590:
591: % figures should be put into the text as floats.
592: % Use the graphics or graphicx packages (distributed with LaTeX2e)
593: % and the \includegraphics macro defined in those packages.
594: % See the LaTeX Graphics Companion by Michel Goosens, Sebastian Rahtz,
595: % and Frank Mittelbach for instance.
596: %
597: % Here is an example of the general form of a figure:
598: % Fill in the caption in the braces of the \caption{} command. Put the label
599: % that you will use with \ref{} command in the braces of the \label{} command.
600: % Use the figure* environment if the figure should span across the
601: % entire page. There is no need to do explicit centering.
602:
603: % \begin{figure}
604: % \includegraphics{}%
605: % \caption{\label{}}
606: % \end{figure}
607:
608: % Surround figure environment with turnpage environment for landscape
609: % figure
610: % \begin{turnpage}
611: % \begin{figure}
612: % \includegraphics{}%
613: % \caption{\label{}}
614: % \end{figure}
615: % \end{turnpage}
616:
617: % tables should appear as floats within the text
618: %
619: % Here is an example of the general form of a table:
620: % Fill in the caption in the braces of the \caption{} command. Put the label
621: % that you will use with \ref{} command in the braces of the \label{} command.
622: % Insert the column specifiers (l, r, c, d, etc.) in the empty braces of the
623: % \begin{tabular}{} command.
624: % The ruledtabular environment adds doubled rules to table and sets a
625: % reasonable default table settings.
626: % Use the table* environment to get a full-width table in two-column
627: % Add \usepackage{longtable} and the longtable (or longtable*}
628: % environment for nicely formatted long tables. Or use the the [H]
629: % placement option to break a long table (with less control than
630: % in longtable).
631: % \begin{table}%[H] add [H] placement to break table across pages
632: % \caption{\label{}}
633: % \begin{ruledtabular}
634: % \begin{tabular}{}
635: % Lines of table here ending with \\
636: % \end{tabular}
637: % \end{ruledtabular}
638: % \end{table}
639:
640: % Surround table environment with turnpage environment for landscape
641: % table
642: % \begin{turnpage}
643: % \begin{table}
644: % \caption{\label{}}
645: % \begin{ruledtabular}
646: % \begin{tabular}{}
647: % \end{tabular}
648: % \end{ruledtabular}
649: % \end{table}
650: % \end{turnpage}
651:
652: % Specify following sections are appendices. Use \appendix* if there
653: % only one appendix.
654: %\appendix
655: %\section{}
656:
657: % If you have acknowledgments, this puts in the proper section head.
658: %\begin{acknowledgments}
659: % put your acknowledgments here.
660: %\end{acknowledgments}
661:
662: % Create the reference section using BibTeX:
663: %\bibliography{basename of .bib file}
664: \begin{thebibliography}{99}
665: \bibitem{IFT} See, for example, M.~Imada, A.~Fujimori, and Y.~Tokura,
666: Rev.~Mod.~Phys. {\bf 70}, 1039 (1998).
667: \bibitem{MASON} See, for example, T.~E.~Mason, in {\it Handbook on the
668: Physics and Chemistry of Rare Earths}, Eds. K.~A.~Gschneidner Jr.,
669: L.~Eyring, and M.~B.~Maple (North Holland, 2000).
670: \bibitem{WELLS_ETAL}B.~O.~Wells {\it et al.}, Phys. Rev. Lett.
671: {\bf 74}, 964 (1995).
672: \bibitem{RONNING_ETAL}F. Ronning {\it et al.}, Science {\bf 282}, 2067
673: (1998).
674: \bibitem{SHEN_RMP}A.~Damascelli, Z.~Hussain, and Z.~X.~Shen,
675: Rev. Mod. Phys. {\bf 75}, 473 (2003).
676: \bibitem{CHN}S.~Chakravarty, B.~I.~Halperin, and D.~R.~Nelson,
677: Phys.~Rev.~B {\bf 39}, 2344 (1989).
678: \bibitem{HALDANE}F.~D.~M.~Haldane, Phys. Lett. {\bf 93A}, 464 (1983).
679: \bibitem{QMC}B.~B.~Beard {\it et al.}, Phys. Rev. Lett. {\bf 80}, 1742
680: (1998);J.-K.~Kim and M.~Troyer, Phys. Rev. Lett. {\bf 80}, 2705
681: (1998).
682: \bibitem{RAJARAMAN}R.~Rajaraman, {\it Solitons and Instantons}
683: (North-Holland, 1987).
684: \bibitem{AA}D.~P.~Arovas and A.~Auerbach,
685: Phys.~Rev.~B {\bf 38}, 316 (1988).
686: \bibitem{READ_SACHDEV}N.~Read and S.~Sachdev,
687: Phys.~Rev.~Lett. {\bf 62}, 1694 (1989);
688: Phys.~Rev.~B {\bf 42}, 4568 (1990).
689: \bibitem{CHUBUKOV}A.~Chubukov, Phys.~Rev.~B {\bf 44}, 12318 (1991).
690: \bibitem{NG95}T.-K.~Ng, Phys. Rev. B {\bf 52}, 9491 (1995);
691: T.-K.~Ng, Phys.~Rev.~Lett. {\bf 82}, 3504 (1999).
692: \bibitem{YOSHIOKA}D.~Yoshioka,
693: J.~Phys.~Soc.~Jpn. {\bf 58}, 3733 (1989).
694: \bibitem{BEC}J.~E.~Hirsh and S.~Tang,
695: Phys.~Rev.~B {\bf 39}, 2850 (1989);
696: M.~Raykin and A.~Auerbach, Phys.~Rev.~Lett. {\bf 70}, 3808(1993).
697: \bibitem{AFFLECK_MARSTON}I.~Affleck and J.~B.~Marston, Phys. Rev. B
698: {\bf 37}, 3774 (1988).
699: %\bibitem{HSU}T.~C.~Hsu, Phys. Rev. B {\bf 41}, 11379 (1990).
700: \bibitem{MARSTON}J.~B.~Marston, Phys.~Rev.~Lett. {\bf 64}, 1166
701: (1990).
702: \bibitem{KIM_LEE}D.~H.~Kim and P.~A.~Lee, Ann. Phys. {\bf 272}, 130
703: (1999).
704: \bibitem{ZHANG_RICE}F.~C.~Zhang and T.~M.~Rice, Phys. Rev. B {\bf 37},
705: 3759 (1988).
706: \bibitem{SHRAIMANN_SIGGIA}B.~I.~Shraiman and E.~D.~Siggia,
707: Phys.~Rev.~Lett. {\bf 61}, 467 (1988).
708: \bibitem{KANE_LEE_READ}C.~Kane, P.~A.~Lee, and N.~Read,
709: Phys.~Rev.~B {\bf 39}, 6880 (1989).
710: %\bibitem{Aharony88}A.~Aharony, R.~J.~Birgeneau, A.~Coniglio,
711: %M.~A.~Kastner, and H.~E.~Stanley, Phys.~Rev.~Lett. {\bf 60}, 1330
712: %(1988).
713: \bibitem{BELAVIN_POLYAKOV}A.~A.~Belavin and A.~M.~Polyakov,
714: JETP Lett. {\bf 22}, 245 (1975).
715: \bibitem{SAXENA_DANDOLOFF} A.~Saxena and R.~Dandoloff,
716: Phys.~Rev.~B {\bf 66}, 104414 (2002).
717: \bibitem{KOGUT83} See, for example, J.~B.~Kogut, Rev. Mod. Phys. {\bf
718: 55}, 775 (1983).
719: \bibitem{LAUGHLIN97}R.~B.~Laughlin, Phys. Rev. Lett. {\bf 79}, 1726
720: (1997).
721: \bibitem{SINGH}R.~R.~P.~Singh, Phys. Rev. B {\bf 39}, 9760 (1989).
722: \bibitem{FISHER_LEE}M.~P.~A.~Fisher and D.~H.~Lee,
723: Phys. Rev. B {\bf 39}, 2756 (1989).
724: \bibitem{APPELQUIST_ETAL}R.~D.~Pisarski, Phys.~Rev.~D {\bf 29}, 2423
725: (1984);T.~Appelquist, D.~Nash and
726: L.~C.~R.~Wijewardhana, Phys.~Rev.~Lett. {\bf 60}, 2575 (1988).
727: \bibitem{TESANOVIC}See also Z.~Te\u{s}anovi\'{c}, O.~Vafek, and
728: M.~Franz, Phys.~Rev.~B {\bf 65}, 180511(R), 2002.
729: \bibitem{GOODING}R.~J.~Gooding, Phys. Rev. Lett. {\bf 66}, 2266
730: (1991).
731: \bibitem{SKYRMIONS}
732: P.~B.~Wiegmann, Phys.~Rev.~Lett. {\bf 60}, 821 (1988);
733: J.~P.~Rodriguez, Phys.~Rev.~B {\bf 39}, 2906 (1989);
734: A.~Auerbach, B.~E.~Larson, and G.~N.~Murthy, Phys.~Rev.~B {\bf 43},
735: 11515 (1991); S.~Haas {\it et al}., Phys.~Rev.~Lett. {\bf 77}, 3021
736: (1996);E.~C.~Marino, Phys.~Rev.~B {\bf 61}, 1588 (2000).
737: \bibitem{ANDERSON}P.~W.~Anderson, Science {\bf 235}, 1196 (1987).
738: \bibitem{BASKARAN}G.~Baskaran, Phys. Rev. B {\bf 68}, 212409 (2003).
739: \end{thebibliography}
740: \end{document}
741: %
742: % ****** End of file template.aps ******
743: \bibitem{NG99}T.-K.~Ng, Phys. Rev. Lett. {\bf }, (1999).
744: \bibitem{RONNING_ETALUN}F.~Ronning, C.~Kim, Z.-X.~Shen, and
745: L.~L.~Miller, unpublished.
746: