1: %-----------------------------------
2: % Edited by DES 6 March **
3: %-----------------------------------
4: %\documentstyle[aps,prl,preprint]{revtex}
5: %\documentclass[aps,twocolumn,floatfix]{revtex4}
6: \documentclass[aps,twocolumn]{revtex4}
7:
8: \usepackage{graphicx}% Include figure files
9: \usepackage{dcolumn}% Align table columns on decimal point
10: \usepackage{bm}% bold math
11: \usepackage{amsmath}
12: \begin{document}
13:
14: \newcommand{\bk}{{\bf k}}
15: \newcommand{\bp}{{\bf p}}
16: \newcommand{\bv}{{\bf v}}
17: \newcommand{\bq}{{\bf q}}
18: \newcommand{\bQ}{{\bf Q}}
19: \newcommand{\br}{{\bf r}}
20: \newcommand{\bR}{{\bf R}}
21: \newcommand{\bB}{{\bf B}}
22: \newcommand{\bA}{{\bf A}}
23: \newcommand{\bK}{{\bf K}}
24: \newcommand{\vd}{{v_\Delta}}
25: \newcommand{\tr}{{\rm Tr}}
26: %\def\slash#1{\hbox{$#1$\kern-0.5em\raise0.3ex\hbox{/}}}
27: \newcommand{\kslash}{\not\!k}
28: \newcommand{\qslash}{\not\!q}
29: \newcommand{\pslash}{\not\!p}
30: \newcommand{\rslash}{\not\!r}
31: \newcommand{\bs}{{\bar\sigma}}
32:
33: \title{Magnetic field induced charge and spin instabilities in
34: cuprate superconductors}
35: %strongly
36: %fluctuating cuprate superconductors: QED$_3$ in a box.}
37:
38: \author{M. Franz\rlap,$^1$ D. E. Sheehy$^1$ and Z. Te\v{s}anovi\'c$^2$}
39: \affiliation{$^1$Department of Physics and Astronomy,
40: University of British Columbia, Vancouver, BC, Canada V6T 1Z1\\
41: $^2$Department of Physics and Astronomy,
42: Johns Hopkins University, Baltimore, MD 21218}
43:
44: \date{\today}
45:
46:
47: \begin{abstract}
48: A $d$-wave superconductor, subject to strong phase fluctuations, is known
49: to suffer an antiferromagnetic instability closely related to the chiral
50: symmetry
51: breaking in (2+1)-dimensional quantum electrodynamics (QED$_3$). Based
52: on this idea we formulate a ``QED$_3$ in a box'' theory of
53: {\em local} instabilities of a $d$-wave superconductor in the
54: vicinity of a single pinned vortex undergoing quantum fluctuations.
55: As a generic outcome we find an incommensurate 2D spin
56: density wave forming in the neighborhood of a vortex with a concomitant
57: ``checkerboard'' pattern in the local electronic density of states, in
58: agreement with recent neutron scattering and tunneling spectroscopy
59: measurements.
60: \end{abstract}
61:
62: \maketitle
63: %\pacs{74.60.-w,74.60.Ec,74.72.-h}
64:
65: Among the open questions in modern condensed matter physics, few have
66: inspired more theoretical effort than the emergence of a superconducting
67: state from the doped antiferromagnetic (AF) insulator~\cite{anderson1}.
68: %
69: Recently, using an ``inverted'' approach to the problem \cite{balents1,ft1},
70: it has been shown that AF order arises naturally when the superconducting
71: order in a $d$-wave superconductor ($d$SC) is destroyed by vortex-antivortex
72: fluctuations \cite{herbut1,ft2}.
73: %
74: As we shall discuss, the implications of these theories transcend the
75: possibility of providing a
76: route to understanding the destruction of superconductivity in
77: strongly-underdoped cuprates; indeed, they also apply to
78: the problem of {\em local} field-induced vortices within the superconducting
79: state.
80:
81:
82: Recent neutron scattering~\cite{lake1,lake2} and scanning tunneling
83: spectroscopy (STS)~\cite{davis1}
84: experiments have revealed the presence of local AF and charge
85: order in the vicinity of field-induced vortices.
86: %
87: Existing theoretical treatments~\cite{arovas1,demler1,ting1,zhang1,demler2}
88: of vortex-induced AF ordering rely on the proximity of the system to a
89: quantum critical point. Within such treatments, it is the suppression of the
90: SC order parameter near the vortex
91: cores that leads to the nucleation of islands of AF order.
92: %
93: Here we present an alternative scenario in
94: which the AF order is brought about by {\em local quantum
95: fluctuations} of a vortex around its equilibrium position.
96: In the present theory
97: there is no competition between the SC amplitude and AF order:
98: the latter arises purely from the presence of vortex fluctuations and is a
99: genuine low-energy phenomenon taking place on lengthscales much longer than
100: the core size.
101:
102: It is a well-known fact that the low superfluid density in cuprates
103: makes the SC order
104: vulnerable to phase fluctuations~\cite{emery1,corson1,ong1}.
105: %
106: This observation has inspired theories in which the pseudogap state is modeled
107: as a phase-disordered $d$-wave
108: superconductor~\cite{balents1,ft1,franz1,kwon1}, such that the
109: demise of superconductivity is brought about by the unbinding and
110: proliferation of the topological defects -- vortices -- in the phase of the
111: SC order parameter.
112: %
113: It has been pointed
114: out~\cite{ft1} that fluctuating vortices produce
115: a non-trivial Berry-phase interaction between the quasiparticles of the
116: underlying $d$SC.
117: %
118: This interaction is described
119: in terms of a massless non-compact gauge field $a_\mu$, minimally coupled to
120: the Dirac fermions representing the low-energy quasiparticle
121: excitations of the
122: system. Within the theory of Ref.~\cite{ft1} which maps the problem
123: onto (2+1)-dimensional quantum electrodynamics (QED$_3$), it is this
124: interaction that destroys the Fermi liquid nature of quasiparticles in the
125: pseudogap state and ultimately drives the AF instability~\cite{herbut1,ft2}.
126: Remarkably, both the `algebraic' Fermi liquid describing the symmetric
127: pseudogap phase and the
128: antiferromagnet emerge from the same QED$_3$ theory \cite{ft1}.
129:
130: Here, we use the philosophy and formalism developed in
131: Refs.~\cite{ft1,herbut1,ft2} to model quasiparticle excitations in the {\em
132: superconducting state} in the spatial region close to a single field-induced
133: vortex undergoing fluctuations around its equilibrium position.
134: We call this model ``QED$_3$ in a box''.
135: We note that there
136: exists direct experimental evidence that individual vortices indeed undergo
137: significant quantum fluctuations~\cite{hoogenboom1}. We find that, under
138: generic conditions, interactions generated by such fluctuating vortex
139: lead to {\em local} instability of the superconducting state which takes
140: form of a 2D incommensurate AF spin density wave (SDW) with a wave vector
141: tied to the positions of the nodes in the underlying $d$-wave gap.
142:
143: In order to motivate our model for a single vortex we first review the
144: treatment of the AF instability in QED$_3$ and reformulate
145: it in a way that will be more suitable for our present purposes. We start from
146: the Euclidean QED$_3$ action $S=\int d^3x{\cal L}_D$ with
147: %
148: \begin{equation}
149: {\cal L}_D\equiv \sum_{l=1}^{N}
150: \bar\Psi_l(x) \gamma_\mu (i\partial_\mu -a_\mu)\Psi_l(x) + {\cal L}_B[a(x)],
151: \label{l1}
152: \end{equation}
153: %
154: describing the low-energy fermionic excitations of a $d$-wave SC coupled
155: to fluctuating vortices represented by the gauge field $a_\mu$~\cite{ft1}.
156: Here, $\Psi_l(x)$ is a four component Dirac spinor representing the fermionic
157: excitations associated with a pair of antipodal
158: nodes, $x=(\tau,{\br})$ denotes the space-time coordinate, and
159: $\gamma_\mu$ ($\mu=0,1,2$) are the gamma matrices satisfying
160: $\{\gamma_\mu,\gamma_\nu\}=2\delta_{\mu\nu}$. The number $N$ of
161: fermion species is equal to $2$ for single-layer cuprates; $N=4,6,...$
162: for bilayer, trilayer and multilayer materials.
163: The Lagrangian ${\cal L}_B$ encodes the dynamics
164: of the gauge field $a_\mu$ and is given by
165: ${\cal L}_B[a]=\Pi_{\mu\nu}(q)a_\mu(q)a_\nu(-q)$ with
166: %
167: \begin{equation}
168: \Pi_{\mu\nu}(q)=\left(m_a+ \frac{N}{8}|q|\right)
169: \left(\delta_{\mu\nu}-\frac{q_\mu q_\nu}{q^2}\right).
170: \label{ber}
171: \end{equation}
172: %
173: The gauge field mass $m_a$ vanishes when vortices
174: are unbound (i.e., in the pseudogap regime or, in the present situation, near
175: a single fluctuating vortex)
176: and is finite in the superconducting state where vortices appear
177: only in tightly bound loops or pairs.
178:
179: In the standard treatment~\cite{herbut1,ft2} the AF order occurs via the
180: phenomenon of chiral symmetry breaking~\cite{pisarski1,appelquist1,nash1}
181: in the QED$_3$ Lagrangian (\ref{l1}). The instability is signaled by the
182: spontaneous
183: generation of fermion mass, $m_D$, which is interpreted in our context as
184: the onset of SDW gap~\cite{herbut1,ft2} for the original Bogoliubov
185: quasiparticles.
186: The most general, nonperturbative treatment of mass-generation in QED$_3$
187: obtains $m_D$
188: as a solution of a self-consistent Dyson-Schwinger equation. Here we shall
189: follow a slightly simpler route which leads to the same result and has the advantage
190: of being more easily generalizable to the present problem. In
191: Eq.~(\ref{l1}) we integrate out the gauge field to
192: obtain the following fermionic effective action:
193: %
194: \begin{eqnarray}
195: S_{\rm eff}&=&\int d^3x
196: \bar\Psi(x) \gamma_\mu i\partial_\mu\Psi(x) \label{l2} \\
197: &-& \int d^3x\int d^3y J_\mu(x)D_{\mu\nu}(x-y)J_\nu(y),
198: \nonumber
199: \end{eqnarray}
200: %
201: where $J_\mu(x)\equiv\bar\Psi(x)\gamma_\mu\Psi(x)$ is the fermion 3-current
202: and $D_{\mu\nu}(x)$ is the Fourier transform of the gauge boson propagator
203: $D_{\mu\nu}(q)=\Pi_{\mu\nu}^{-1}(q)$. Henceforth we shall focus on a
204: single pair of nodes and thus drop the nodal index $l$.
205: The integrand of the
206: interaction term may be written as $D_{\mu\nu}(x-y)\tr[\Psi(y)\bar\Psi(x)
207: \gamma_\mu\Psi(x)\bar\Psi(y)\gamma_\nu]$ where the trace is taken over the
208: spinor indices. This form suggests a Hartree-Fock (HF) approach in which we
209: decouple the 4-fermion interaction to obtain
210: $D_{\mu\nu}(x-y)\tr[\Psi(y)\bar\Psi(x)\gamma_\mu G_0(x,y)\gamma_\nu]$ with
211: $G_0(x,y)=\langle\Psi(x)\bar\Psi(y)\rangle$ and the average is taken
212: with respect to the HF effective action to be specified shortly. To make the
213: structure of the interaction term more transparent we
214: utilize the relative and center of
215: mass coordinates $r=x-y$ and $R=(x+y)/2$ to write it as
216: %
217: \begin{equation}
218: \int d^3R\,\int d^3r\,\tr\left[
219: \Psi(R_+)\bar\Psi(R_-)\gamma_\mu G_0(R,r)\gamma_\nu\right]D_{\mu\nu}(r),
220: \label{int1}
221: \end{equation}
222: %
223: where $R_\pm\equiv R\pm\frac{r}{2}$. In the {\em uniform} system the
224: Green's function is independent of $R$, $G_0(R,r)=G_0(r)$. Furthermore,
225: both $G_0(r)$ and $D_{\mu\nu}(r)$ are strongly peaked at $r\to 0$. The
226: dominant contribution to the interaction therefore comes from this region and
227: we may write (\ref{int1}) as~\cite{remark1}
228: %
229: \begin{equation}
230: \int d^3R\bar\Psi(R)\Psi(R)\int d^3r
231: \gamma_\mu G_0(r)\gamma_\nu D_{\mu\nu}(r).
232: \label{int2}
233: \end{equation}
234: %
235: We have dropped the trace since the interaction is proportional to the unit matrix
236: in the spinor space.
237:
238: Inspection of Eq.~(\ref{int2}) suggests the following HF effective action
239: and self-consistency condition:
240: %
241: \begin{subequations}
242: \begin{eqnarray}
243: S_{\rm HF}&=&\int d^3x
244: \bar\Psi(x)( \gamma_\mu i\partial_\mu-im_D)\Psi(x),
245: \label{l3}
246: \\
247: im_D&=&\frac{1}{4}\tr\int d^3r
248: \gamma_\mu G_0(r)\gamma_\nu D_{\mu\nu}(r).
249: \label{self}
250: \end{eqnarray}
251: \end{subequations}
252: %
253: The last integral is easily evaluated by going to momentum space and Eq.\
254: (\ref{self}) becomes $m_D=(8m_D/N\pi^2)\ln(\Lambda/m_D)$, where
255: $\Lambda$ is the high-momentum cutoff. This yields a nontrivial solution
256: %
257: \begin{equation}
258: m_D=\Lambda e^{-N\pi^2/8},
259: \label{mass}
260: \end{equation}
261: %
262: in agreement with the classic result of Pisarski~\cite{pisarski1}.
263: More sophisticated treatments
264: \cite{appelquist1,nash1} based on the Schwinger-Dyson equation give a finite critical value of $N_c$ above which
265: no mass is generated; however, for our purposes the level of approximation embodied by Eq.~(\ref{mass})
266: will be sufficient.
267:
268:
269: We have thus seen that, in a uniform system, fluctuating vortices lead to the formation
270: of SDW order.
271: The challenge we now face is twofold: (i) we must adapt the above treatment
272: to the case of a single
273: fluctuating vortex, and (ii) since we seek to study the commensuration effects
274: present in real materials, we must formulate the corresponding theory on the
275: lattice. To address (i) let us denote by $\ell_v$ the characteristic
276: length scale over which the vortex fluctuates around its classical
277: equilibrium position. Within this length scale, the Berry-phase interaction
278: between quasiparticles (and hence tendency towards SDW ordering) will be strong.
279: We model this by taking in this region the gauge field
280: to be {\em massless}. On the other hand at distances well beyond $\ell_v$,
281: quasiparticles feel no interaction and we model this by gauge field having a
282: large mass $m_a$. In particular we take,
283: %
284: \begin{equation}
285: m_a(\bR)=\Delta_0\left(\frac{|\bR|}{\ell_v}\right)^n,
286: \label{gaugemass}
287: \end{equation}
288: %
289: where $\Delta_0$ is an energy scale which we take to be
290: the maximum superconducting gap, $|\bR|$ is the
291: distance from the vortex equilibrium position and $n$ is a positive exponent.
292: (We use $n=2$ but our numerical calculations below are largely insensitive
293: to the exact value of $n$.)
294:
295: To address (ii), (i.e.,~to put the theory on the lattice) we recall that the
296: effective action (\ref{l1}) and its HF version Eq.~(\ref{l3}) descend from a
297: model of a lattice
298: $d$SC linearized near the nodes of the gap. We therefore consider the
299: corresponding lattice Hamiltonian enriched by the ``mass'' term present
300: in Eq.~(\ref{l3}), to represent the HF decoupled Berry phase interaction. Thus we have,
301: %
302: \begin{equation}
303: H_{\rm HF}=\sum_\sigma\sum_{\langle ij\rangle}
304: \Phi^\dagger_{i\sigma}{\cal H}_{ij}^\sigma \Phi_{j\sigma}.
305: \label{h1}
306: \end{equation}
307: %
308: Here $\Phi^\dagger_{i\sigma}=(c^\dag_{i\sigma},c_{i\bs})$, $c^\dag_{i\sigma}$
309: represents the electron creation operator at lattice site $i$,
310: spin index $\bs=-\sigma$, and
311: %
312: \begin{equation}
313: {\cal H}_{ij}^\sigma=\left(
314: \begin{array}{cc}
315: -t_{ij}+\delta_{ij}(m_{i\sigma}-\mu) & \Delta_{ij} \\
316: \Delta^*_{ij} & t_{ij}-\delta_{ij}(m_{i\bs}-\mu)
317: \end{array}
318: \right), \nonumber
319: %\label{h2}
320: \end{equation}
321: %
322: with $t_{ij}$ the tight binding hopping amplitude, $\Delta_{ij}$ the
323: SC gap, $\mu$ chemical potential, and $m_{i\sigma}$ the local spin
324: magnetization representing the mass gap $m_D$ in Eq.\ (\ref{l3}).
325:
326: We diagonalize $H_{\rm HF}$ by the generalized Bogoliubov
327: transformation $c_{i\sigma}=\sum_n[u_{n\sigma}(\br_i)\gamma_{n\sigma}
328: +\sigma v^*_{n\bs}(\br_i)\gamma^\dag_{n\bs}]$, where
329: $\chi_{n\sigma}(\br_i)\equiv [u_{n\sigma}(\br_i),\sigma v_{n\sigma}(\br_i)]^T$
330: satisfy
331: %
332: \begin{equation}
333: \sum_j{\cal H}_{ij}^\sigma \chi_{n\sigma}(\br_j)=\epsilon_{n\sigma}
334: \chi_{n\sigma}(\br_i).
335: \label{eigen1}
336: \end{equation}
337: %
338: In terms of the $\chi_{n\sigma}$,
339: the self-consistency condition (\ref{self}) can be written as
340: %
341: \begin{eqnarray}
342: m_{i\sigma}&=&\sum_{n\sigma j}\sigma f(\epsilon_{n\sigma})
343: %\sum_j
344: V_i(\br_j)u^*_{n\sigma}(\bR_i+\br_j)u_{n\sigma}(\bR_i-\br_j),
345: \nonumber
346: \\
347: V_i(\br)&\equiv&\frac{1}{4}\int_0^\infty d\tau e^{-\tau\epsilon_{n\sigma}}
348: \tr\left[\gamma_\mu D_{\mu\nu}(\tau,\br)\gamma_\nu\right].
349: \label{self2}
350: \end{eqnarray}
351: %
352: For $m_a\neq 0$ the above integral cannot be evaluated in closed form.
353: However, we find that it can be accurately approximated by a simple
354: interpolation formula
355: %
356: \begin{equation}
357: V_i(\br)\simeq{\cal V}_0
358: \frac{c_1}
359: {r(rm_a(\bR_i)+c_1)(r\epsilon_{n\sigma}+c_1)}
360: \label{pot2}
361: \end{equation}
362: %
363: where $r=|\br|$, $c_1=2/\pi$ and ${\cal V}_0=16/N\pi^2$ for the case
364: of an isotropic Dirac
365: cone ($t=\Delta$). In the physical case $t>\Delta$ the constant ${\cal V}_0$
366: will be modified somewhat and in what follows we treat it as an adjustable
367: parameter of the model measuring the strength of the interaction. It is
368: interesting to note that, as seen from Eq.~(\ref{pot2}),
369: in (2+1)D a gauge field mass does {\em not} lead to
370: exponentially decaying interactions on long length scales.
371:
372: To capture the effect of vortex fluctuations on the local superconducting order
373: we solve the eigenproblem~(\ref{eigen1}) numerically on a lattice of
374: $M\times M$ sites and iterate to self-consistency using Eq.~(\ref{self2}).
375: For simplicity we consider only nearest-neighbor hopping amplitudes
376: $t_{ij}=t$ and a
377: uniform $d$-wave gap $\Delta_{ij}=\pm\Delta_0$, with $+$ and
378: $-$ signs referring to vertical and horizontal bonds respectively.
379: We emphasize that the vortex, fluctuating around the equilibrium position at the center of our
380: lattice, enters through the
381: position dependent gauge-field mass $m_a(\bR)$ given by Eq.\
382: (\ref{gaugemass}), which in turn enters the potential $V_i(\br)$
383: given by Eq.\ (\ref{pot2}). In the spirit of our working philosophy
384: that the SDW
385: order arises from the vortex fluctuations (and therefore from the gauge field),
386: we neglect at this stage any effects of the superflow around the vortex or
387: suppression of $\Delta_{ij}$ in the core. Such effects are well understood
388: and will be included in a future publication. We also neglect the effects
389: of changes in the fermionic spectrum due to the onset of SDW on
390: the interaction mediated by Berry gauge field Eq.\ (\ref{pot2}).
391:
392: The diagonalizations are performed using standard {\sc LAPACK} routines, which
393: allow us to handle systems up to $40\times 40$ sites. Typically, 10-15
394: iterations are
395: needed to ensure self-consistency in $m_{i\sigma}$. We use both periodic
396: and free boundary conditions and find that they have negligible effect on
397: the results reported below. Our typical results are summarized in Figs.~\ref{fig1}
398: and \ref{fig2}, showing the spatial distributions of the spin
399: magnetization $M_i=\sum_\sigma\sigma\langle c^\dag_{i\sigma}c_{i\sigma}
400: \rangle$, staggered spin magnetization
401: $M^S_i=M_i(-1)^{x_i+y_i}$, local electron charge density $n_i=\sum_\sigma
402: \langle c^\dag_{i\sigma}c_{i\sigma}\rangle$, and energy integrated
403: local density of states (LDOS) $S_{E_1}^{E_2}(i)=\int_{E_1}^{E_2}\rho_i(E)dE$
404: where $\rho_i(E)$ is the LDOS at site $i$, as well as their respective Fourier
405: transforms (FTs).
406: %
407: \begin{figure}[t]
408: \includegraphics[width=8.5cm]{fig1.ps}
409: \caption{\label{fig1}
410: Magnetization $M_i$ and staggered magnetization $M^S_i$ induced in the
411: vicinity of the fluctuating vortex. Parameters used: $t=1$, $\Delta_0=0.7$,
412: $\mu=-1.6$ resulting in maximum gap of 2.3 and
413: average charge density $n=0.62$ electrons per site,
414: ${\cal V}_0=1.0$ and $\ell_v=12$; periodic boundary conditions.
415: }
416: \end{figure}
417: %
418:
419: Panel (a) in Fig. \ref{fig1} illustrates the ``2D'' incommensurate SDW pattern
420: emerging in the vicinity of a fluctuating vortex
421: with an $8\times 8$ unit cell containing islands of AF order separated
422: by anti-phase domain walls [apparent in panel (c)]. The FT displayed in panel
423: (b) reveals that this pattern can be thought of as a superposition of four
424: 1D SDWs with wave vectors $\bQ_{\rm SDW}=\pi(1\pm\delta_{\rm SDW},
425: 1\pm\delta_{\rm SDW})$, $\delta_{\rm SDW}=\frac{1}{4}$.
426: The size of $\delta_{\rm SDW}$ is doping dependent: it shrinks with
427: increasing $\mu$
428: and vanishes at half filling $(\mu=1)$, giving rise to perfectly commensurate
429: AF SDW. We also find that for $\mu<-1.7$ the SDW becomes very weak for
430: reasonable values of coupling ${\cal V}_0$: overdoped samples are less
431: susceptible to AF instability.
432: %
433: \begin{figure}[t]
434: \includegraphics[width=8.5cm]{fig2.ps}
435: \caption{\label{fig2}
436: Local charge density $n_i$ and integrated LDOS $S_{0.2}^{0.7}(i)$ for the same
437: parameters as Fig. \ref{fig1}.
438: }
439: \end{figure}
440: %
441:
442:
443: According to general symmetry arguments, a spatial
444: modulation in the spin density generates a modulation in the charge density,
445: $\delta n_i\propto M_i^2$. For our 2D SDW pattern this implies that the
446: corresponding CDW will have a unit cell with half the area, rotated by
447: $45^\circ$ relative to the unit cell of $M^S_i$. Indeed, panels
448: (a) and (b) of Fig. \ref{fig2} confirm this general
449: expectation, showing a ``checkerboard'' CDW at principal wavevectors
450: $\bQ_{\rm CDW}=\pi(\pm\delta_{\rm CDW},0),\pi(0,\pm\delta_{\rm CDW})$ with
451: $\delta_{\rm CDW}=\frac{1}{2}$. A similar checkerboard pattern arises in the
452: integrated LDOS and is displayed in panels (c) and (d).
453:
454: Our findings of a checkerboard pattern in LDOS are consistent with the recent
455: STS experiments performed on Bi$_2$Sr$_2$CaCu$_2$O$_{8+\delta}$ (BSCCO)
456: crystals~\cite{davis1}. Our prediction is that the period of the pattern should
457: increase with underdoping and that the effect should vanish in the overdoped
458: samples. Also, if the observed LDOS pattern is associated with electron
459: density modulation in a single CuO layer, we predict that the corresponding
460: neutron scattering peaks should be found at wavectors $\bQ_{\rm SDW}=
461: \pi(1\pm\frac{1}{4},1\pm\frac{1}{4})$. We note that neutron experiments
462: \cite{lake1,lake2} on
463: La$_{1.84}$Sr$_{0.16}$CuO$_4$ (LSCO) show peaks at different $k$-space
464: positions, $\pi(1\pm\frac{1}{4},1)$ and $\pi(1,1\pm\frac{1}{4})$. Although the
465: findings of STS and neutron experiments are generally cited as being mutually
466: consistent our analysis above indicates that this is not necessarily so:
467: for a genuine 2D SDW
468: illustrated in Fig.~\ref{fig1} determination of the corresponding CDW
469: must consider the interference terms which cause the apparent $45^\circ$
470: rotation of the latter. In the absence of neutron measurements on BSCCO we see
471: two possible resolutions of this difficulty. First, it may be that the
472: CDW pattern is truly 2D and neutron scattering on BSCCO would find a pattern
473: illustrated in Fig. \ref{fig1}(b). Second, it could be that STS sees an
474: incoherent superposition of two orthogonal 1D CDWs originating in two CuO
475: layers comprising the BSCCO bilayer. This would explain the $x-y$ anisotropy
476: reported in Ref.~\cite{davis1} and the neutron pattern would be consistent
477: with that observed in LSCO. Within our simple model such 1D solutions have
478: slightly higher energy than the 2D solutions reported above but it is possible
479: that in a more complete model (using e.g.~a more realistic band structure) the
480: situation will be reversed.
481:
482: To summarize, we have presented a ``QED$_3$ in a box'' theory for field
483: induced spin and charge
484: instabilities in cuprates, driven by local fluctuations of a pinned vortex.
485: Without any need to fine-tune parameters we find LDOS patterns in detailed
486: agreement with the tunneling data on BSCCO~\cite{davis1} and we relate
487: them in a plausible way to existing neutron experiments~\cite{lake1,lake2}.
488: Other models~\cite{arovas1,demler1,ting1,zhang1,demler2} rely on the local
489: suppression of the superconducting order parameter and would therefore predict similar
490: instabilities in the vicinity of impurities, grain boundaries, and sample
491: edges where thus far no such effects have been observed.
492:
493: This research was supported in part by NSERC (MF,DES) and
494: NSF Grant DMR00-94981 (ZT).
495:
496:
497: \begin{thebibliography}{10}
498: \bibitem{anderson1} P.W. Anderson, {\em The theory of
499: superconductivity in the high-T$_c$ cuprates},
500: Princeton University Press (1997).
501: \bibitem{balents1} L. Balents, M.P.A. Fisher and C. Nayak, \prb {\bf 60},
502: 1654 (1999).
503: \bibitem{ft1} M. Franz and Z. Te\v{s}anovi\'c, \prl {\bf 87}, 257003 (2001).
504: \bibitem{herbut1} I.F. Herbut, \prl {\bf 88}, 047006 (2002).
505: \bibitem{ft2} Z. Te\v{s}anovi\'c, O. Vafek and M. Franz, \prb {\bf 65},
506: 180511 (2002).
507: %\bibitem{ft3} M. Franz, Z. Te\v{s}anovi\'c and O. Vafek, in preparation.
508: \bibitem{lake1} B. Lake {\em et al.}, Science {\bf 291}, 1759 (2001).
509: \bibitem{lake2} B. Lake {\em et al.}, Nature {\bf 415}, 299 (2002).
510: \bibitem{davis1} J.E. Hoffman {\em et al.}, Science {\bf 295}, 466 (2002).
511: \bibitem{arovas1} D.P. Arovas, A.J. Berlinsky, C. Kallin, S.-C. Zhang,
512: \prl {\bf 79}, 2871 (1997).
513: \bibitem{demler1} E. Demler, S. Sachdev, Y. Zhang, \prl {\bf 87}, 067202
514: (2001).
515: \bibitem{ting1} J.-X. Zhu and C.S. Ting, \prl {\bf 87}, 147002 (2001).
516: \bibitem{zhang1} J.-P. Hu and S.-C. Zhang, cond-mat/0108237.
517: \bibitem{demler2} Y. Zhang, E. Demler, S. Sachdev, cond-mat/0112343.
518: \bibitem{emery1} V. J. Emery and S. A. Kivelson, Nature {\bf 374},
519: 434 (1995).
520: \bibitem{corson1} J. Corson {\em et al.}, Nature {\bf 398}, 221 (1999).
521: \bibitem{ong1} Z. A. Xu {\em et al.}, Nature {\bf 406}, 486 (2000).
522: \bibitem{franz1} M. Franz and A. J. Millis, \prb {\bf 58}, 14572 (1998).
523: \bibitem{kwon1} H. J. Kwon and A. T. Dorsey, \prb {\bf 59}, 6438 (1999).
524: \bibitem{hoogenboom1} B.W. Hoogenboom {\em et al.}, \prb {\bf 62}, 9179
525: (2000).
526: \bibitem{pisarski1} R.D. Pisarski, \prd {\bf 29}, 2423 (1984).
527: \bibitem{appelquist1} T. W. Appelquist, M. Bowick, D. Karabali and
528: L.C.R. Wijewardhana, \prd {\bf 33}, 3704 (1986).
529: \bibitem{nash1} D. Nash, \prl {\bf 62}, 3024 (1989).
530:
531: \bibitem{remark1} This approximation is equivalent to taking the
532: $k$-independent part of $\Sigma(k)$ in the conventional
533: treatment using the Dyson-Schwinger equation \cite{pisarski1}.
534:
535: \end{thebibliography}
536:
537: \end{document}
538:
539:
540:
541:
542:
543:
544:
545:
546:
547:
548:
549:
550:
551:
552:
553:
554:
555:
556:
557:
558: