1: \documentclass[twocolumn,showpacs,aps,prl,floatfix]{revtex4}
2: %\documentclass[preprint,aps,prb]{revtex4}
3: %\usepackage{graphics}
4: \usepackage{graphicx}
5: \usepackage{subfigure}
6: \usepackage{epsfig}
7: \usepackage{psfrag}
8:
9: \makeatletter
10:
11: \usepackage{verbatim}
12: %\usepackage{citen}
13:
14: \makeatletter
15:
16: \input epsf
17: \def\prb{Phys. Rev. B}
18: \def\prl{Phys. Rev. Lett.}
19: \def\pla{Phys. Lett. A}
20: \def\pr{Phys. Rev.}
21: \def\be{\begin{equation}}
22: \def\ee{\end{equation}}
23: \def\ba{\begin{eqnarray}}
24: \def\ea{\end{eqnarray}}
25:
26: \makeatother
27:
28: \begin{document}
29:
30: \title{Spin-wave dispersion in half-doped La$_{3/2}$Sr$_{1/2}$ NiO$_4$}
31:
32: \author{D.~X.~Yao$^1$ and E.~W.~Carlson$^2$}
33: \affiliation{(1) Dept. of Physics, Boston University, Boston, MA 02215\\
34: (2) Dept. of Physics, Purdue University, West Lafayette, IN 47907}
35:
36:
37: \date{September 12, 2006}
38:
39: \begin{abstract}
40: Recent neutron scattering measurements
41: reveal spin and charge ordering in the half-doped nickelate,
42: La$_{3/2}$ Sr$_{1/2}$ NiO$_4$. Many of the features of the
43: magnetic excitations have been explained in terms of the
44: spin waves of diagonal stripes with weak single-ion
45: anisotropy. However, an optical mode dispersing away from the $(\pi,\pi)$
46: point was not captured by this theory.
47: We show here that this apparent optical mode is a natural consequence
48: of stripe twinning in a diagonal stripe pattern with a magnetic coupling structure which
49: is two-fold symmetric, {\em i.e.} one possessing the same spatial rotational symmetry
50: as the ground state.
51:
52: \end{abstract}
53: \pacs{75.30.Ds, 75.10.Hk, 71.27.+a}
54: \maketitle
55:
56:
57: Strongly correlated electronic systems often exhibit some evidence of charge,
58: spin, or orbital order, or some combination thereof.
59: The interplay between these degrees of freedom can lead
60: to a wide variety of novel phases.
61: %especially in the cuprates and
62: %nickelates.\cite{tranquada04a,hayden04,stripesreview}
63: The nickelates in particular show a wide range of doping
64: in which both charge and spin order coexist in stripe patterns.
65: The less-ordered stripe structures in some families of
66: cuprates have been widely studied for their possible connection
67: to high temperature superconductivity.
68:
69: Recent experiments on
70: doped La$_{2-x}$ Sr$_{x}$ NiO$_4$ have shown clear evidence of static
71: diagonal charge and spin stripe order.\cite{boothroyd03a, boothroyd05}
72: Spin wave theory has been successful at describing much of the
73: behavior in these spin-ordered systems.\cite{erica04,yaocarlson06b,tranquada,tranquada03,kruger03,freeman05a}
74: We consider here the recent experiments by Freeman {\em et al.} on
75: the spin dynamics of
76: half-doped La$_{3/2}$ Sr$_{1/2}$NiO$_4$ using inelastic
77: neutron scattering.\cite{freeman05a}
78: In this material, the spins are in a diagonal stripe phase,
79: where stripes run $45^o$ from the Ni-O bond direction,
80: and the charged domain walls are only 2 lattice constants apart.
81: The charge density modulation can either be considered as densely packed stripes, or as
82: a checkerboard, for domain walls centered on the Ni sites,
83: since in that case the two ways of describing the charge pattern
84: are indistinguishable at this filling.
85: However, if the domain walls are centered on oxygen sites ({\em i.e.} for bond-centered stripes),
86: the charge checkerboard description is not applicable.
87:
88: In this paper, we are interested in the extra magnetic mode
89: dispersing away from the antiferromagnetic wavevector $Q_{\rm AF} = (0.5,0.5)$
90: %$(\pi,\pi)$
91: above $50$meV in Fig.~3 of Ref.~\onlinecite{freeman05a}.
92: One explanation put forth by the authors is that diagonal discommensurations
93: in the spin order may be able to account for this extra scattering mode.
94: We show here that the mode could also be due to asymmetry in the
95: spin coupling constants, in which case the ``extra mode'' is really an
96: extension of the acoustic band, made visible due to stripe twinning.
97:
98:
99: %%%%%%%%%%%% FIGURE: DEFINE COUPLINGS %%%%%%%%%%%%%%%%
100: \begin{figure}[htb]
101: {\centering
102: \subfigure[ DS2]
103: {\resizebox*{0.45
104: \columnwidth}{!}{\includegraphics{ds2.lattice.eps}\label{ds2.lattice}}}
105: \hspace{.2in}
106: \subfigure[ DB2]
107: {\resizebox*{0.4\columnwidth}{!}{\includegraphics{db2.lattice.eps}\label{db2.lattice}}}
108: \par}
109: \caption{(Color online) Spin-charge ordering in a NiO$_2$ square
110: lattice. Arrows represent spins on Ni$^{2+}$ sites and circles represent
111: Ni$^{3+}$ holes.
112: (a) DS2: Diagonal, site-centered stripes of spacing $p=2$.
113: Note that in this configuration the charges distribution is
114: symetric under $90^o$ rotations, and is equivalent to a checkerboard pattern.
115: The spin configuration is symmetric under $180^o$ rotations, and breaks the symmetry of the charge checkerboard. The straight-line exchange coupling across the charge domain wall is $J_b$,
116: and $J_c$ and $J_d$ are diagonal couplings perpendicular and
117: parallel to the charge domain wall, respectively.
118: (b) DB2: Diagonal, bond-centered stripes of spacing $p=2$.
119: Here, both the charge and spin patterns are
120: $2$-fold rotationally symmetric.
121: \label{lattice}}
122: \end{figure}
123: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
124:
125:
126:
127: The observed ordering vector is $Q=(0.275,0.275)$, which is
128: close to a commensurate stripe value of $Q=(0.25,0.25)$.
129: The slight deviation from commensurability is believed to be due
130: to the discommensurations described above. Since we are interested
131: in describing relatively high energy effects, in what follows
132: we neglect the small incommensurability, and consider commensurate diagonal stripe
133: structures of spacing $p=2$.
134: We consider two patterns in this paper: site-centered stripes as shown in
135: Fig.~\ref{lattice}(a), and the corresponding bond-centered stripes shown in
136: Fig.~\ref{lattice}(b).
137: %DX: Yes, this is better.
138: %EC: I wanted to make this more general, to apply to both configurations.
139: %The system of DS2 is described by a Heisenberg model with weak out-of-plane anisotropy
140: %(single-ion anisotropy) % yao: This is only for DS2
141: % on a square lattice,
142: %\begin{equation}
143: %H= J_b \sum_{\left< i,j\right>} \mathbf{S}_{i} \cdot
144: %\mathbf{S}_{j}+ J_c \sum_{\left< i,j\right>} \mathbf{S}_{i} \cdot
145: %\mathbf{S}_{j}+ J_d \sum_{\left< i,j\right>} \mathbf{S}_{i} \cdot
146: %\mathbf{S}_{j}+ K_c\sum_{i} (S_i^z)^2
147: %\label{model}
148: %\end{equation}
149: %where $J_b$ is the next nearest exchange coupling across the charge
150: %domain wall, and $J_c$ and $J_d$ are the diagonal couplings perpendicular
151: %and parallel to the charge domain wall, respectively. These are shown
152: %in Fig.~\ref{lattice}.
153: %The strength of the out-of-plane anisotropy is controlled by $K_c$.
154: %While $J_b$ and $J_c$ are antiferromagnetic, $J_d$ is ferromagnetic.
155: %The Hamiltonian of DB2 is similar, but with the different exchange couplings
156: %(see Fig.~\ref{lattice}(b)). %yao
157:
158: To model these two systems, we use a suitably parametrized
159: Heisenberg model
160: % EC K_c
161: %with weak out-of-plane anisotropy
162: %(single-ion anisotropy)
163: on a square lattice,
164: %DX: Since we define J_{i,j} for per bond, we need a 1/2 factor. K_c should
165: %be per site. Then we need put fact ``2'' back.
166: \begin{equation}
167: %H= \frac{1}{2}\sum_{\left< i,j\right>} J_{i,j} \mathbf{S}_{i} \cdot \mathbf{S}_{j}
168: H= \frac{1}{2}\sum_{i,j} J_{i,j} \mathbf{S}_{i} \cdot \mathbf{S}_{j}
169: %+ K_c\sum_{i} (S_i^z)^2 %EC removed.
170: \label{model}
171: \end{equation}
172: where the indices $i$ and $j$ run over all sites, and
173: the couplings $J_{i,j}$ are illustrated in
174: %where the couplings $J_{i,j}$ are defined per bond and are illustrated in %DX: added ``per bond''
175: Fig.~\ref{lattice}.
176: For diagonal, site-centered stripes of spacing $p=2$ (DS2),
177: there is no need for nearest-neighbor coupling, and so we set $J_a = 0$.
178: The straight-line coupling $J_b$ across the domain wall is antiferromagnetic.
179: The diagonal coupling $J_c$ across the domain wall is also antiferromagnetic,
180: but the diagonal coupling $J_d$ parallel to the stripes we take to be ferromagnetic,
181: $J_d <0$, as explained below. In the diagonal bond-centered case (DB2), the nearest neighbor coupling $J_a >0$
182: is finite and antiferromagnetic. We also include the ferromagnetic coupling $J_b <0$
183: across the domain wall.
184: % EC K_c
185: Since we are interested in describing high energy effects, we neglect
186: the very weak single-ion anisotropy term, which splits the $2$-fold degenerate acoustic band
187: at low energy, with one mode remaining gapless at the IC point $Q_{\rm IC} = (0.25,0.25)$,
188: and the other mode developing a small gap.\cite{freeman05a}
189: %In both cases (site- and bond-centered stripes), the strength of the
190: %out-of-plane anisotropy is controlled by $K_c$
191: %for per spin. % DX: per spin
192:
193: %%%%%%%%% FIGURE: 3D PLOTS %%%%%%%%%%%%%%%
194: \begin{figure}[tb]
195: \psfrag{6}{\Large$6$}
196: \psfrag{4}{\Large$4$}
197: \psfrag{2}{\Large$2$}
198: \psfrag{0.25}{\Large$0.25$}
199: \psfrag{0.5}{\Large$0.5$}
200: \psfrag{0.75}{\Large$0.75$}
201: \psfrag{1}{\Large$1$}
202: \psfrag{0}{\Large$0$}
203: \psfrag{w}{\LARGE$\omega$}
204: \psfrag{kx}{\LARGE$\hspace{-.5in}{k_x / 2\pi}$}
205: \psfrag{ky}{\LARGE$\hspace{.15in} {k_y / 2\pi}$}
206: {\centering
207: \subfigure[]
208: {\resizebox*{0.48\columnwidth}{!}{\includegraphics{af.eps}\label{af}}}
209: \subfigure[]
210: {\resizebox*{0.48\columnwidth}{!}{\includegraphics{uniso.eps}\label{uniso}}}
211: \par}
212: \caption{Band structures of diagonal site-centered stripes (DS2).
213: (a) $J_c=J_d=0$;
214: %EC K_c (a) $J_c=J_d=K_c=0$;
215: (b) $J_c=J_b$, $J_d=-0.5J_b$.
216: % EC K_c (b) $J_c=J_b$, $J_d=-0.5J_b$ and $K_c=0$.
217: \label{plot3d}}
218: \end{figure}
219: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
220:
221: %**The extra scattering mode could be an optical mode, which is why we consider DB2.
222: % EC K_c
223: %When $K_c=0$, the site-centered stripes of Fig.~\ref{lattice}(a) admit only two degenerate spin-wave modes, and display only an acoustic band. When $K_c\ne 0$, the acoustic band
224: %is split at low energy.
225: %We also consider the bond-centered stripes of Fig.~\ref{lattice}(b).
226: %These stripes have a true optical branch as well, making them a good
227: %candidate for describing the observed high energy mode above $50$meV.
228: %However, as we will see below, the optical mode always has significant weight
229: %at the antiferromagnetic ordering vector $Q_{\rm AF} = (0.5,0.5)$, which is not seen in the data.
230:
231: For the case of site-centered stripes
232: in the absence of the diagonal couplings, {\em i.e.} for $J_c=J_d=0$,
233: the spin system reduces to two interpenetrating antiferromagnets,
234: with two separate N\'eel vectors but identical N\'eel ordering temperatures. Any weak
235: diagonal coupling is sufficient to establish a unique relative direction between
236: the two N\'eel vectors, and the ground state becomes the
237: stripe structure shown in Fig.~\ref{lattice}(a).
238: The number of reciprocal lattice vectors is also decreased by a factor of $2$
239: in the presence of the diagonal couplings $J_c \ne 0$ or $J_d \ne 0$,
240: as can be seen in the bandstructure of Fig.~\ref{plot3d}.
241: In either case, independent of the value of $J_c$ and $J_d$,
242: although the antiferromagnetic point $Q_{\rm AF}$ is a magnetic reciprocal lattice vector
243: and therefore must have a spin wave cone dispersing out of it, there is no
244: net antiferromagnetism in the system, so that weight is forbidden at zero frequency
245: at $Q_{\rm AF}$. The cone that emanates out of $Q_{\rm AF}$ gains finite weight
246: as energy is increased, but remains faint at low energies.
247:
248:
249: %%%%%%%%%%% FIGURE: DS2 %%%%%%%%%%%%%%%%
250: \begin{figure}[tb]
251: \psfrag{w}{$\omega$}
252: \psfrag{kx}{$k_x/2\pi$}
253: \psfrag{0.25}{$0.25$}
254: \psfrag{0.5}{$0.5$}
255: \psfrag{0.75}{$0.75$}
256: \psfrag{0}{$0$}
257: \psfrag{1}{$1$}
258: \psfrag{2}{$2$}
259: \psfrag{3}{$3$}
260: \psfrag{4}{$4$}
261: \psfrag{6}{$6$}
262: \psfrag{8}{$8$}
263: \psfrag{10}{$10$}
264: \psfrag{a}{(a)}
265: \psfrag{b}{(b)}
266: \resizebox*{1.0\columnwidth}{!}{\includegraphics{dispersion.eps}}
267: \caption{DS2: Diagonal, site-centered stripes of spacing $p=2$.
268: The plots are for twinned stripes,
269: summing contributions along $(k_x,k_x)$ and $(k_x,-k_x)$.
270: (a)$J_c=J_b$ and $J_d=-0.5J_b$;
271: % EC K_c (a)$J_c=J_b$, $J_d=-0.5J_b$ and $K_c=0$;
272: (b)$J_c=2J_b$ and $J_d=-0.5J_b$.
273: % EC K_c (b)$J_c=2J_b$, $J_d=-0.5J_b$ and $K_c=0$.
274: \label{fig:DS2}}
275: \end{figure}
276: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
277:
278:
279: Another key feature of
280: nonzero couplings $J_c$ and $J_d$ for site-centered stripes
281: is the symmetry of the spin wave structure,
282: as shown in Fig.~\ref{plot3d}. In the limit where $J_c=J_d=0$,
283: the spin wave dispersion is symmetric under $90^o$ rotations,
284: as shown in Fig.~\ref{plot3d}(a).
285: However, when either $J_c$ or $J_d$ or both are nonzero, the symmetry
286: is broken, and the spin wave structure is now only symmetric under $180^o$ rotations,
287: as shown in Fig.~\ref{plot3d}(b).
288: This means that for any nonzero $J_c$ or $J_d$,
289: the spin wave velocity of the acoustic mode dispersing out of $Q_{\rm AF}=(0.5,0.5)$
290: is different parallel and perpendicular to the stripe direction.
291: %the acoustic mode has different spin wave velocities dispersing out
292: %of the $(0.5,0.5)$ point.
293: In the presence of stripe twins, the two velocities will appear as two branches
294: in plots of $\omega$ {\em vs.} $\vec{k}$, as shown in Fig.~\ref{fig:DS2}.
295:
296:
297:
298: The magnon dispersion from Eq.~(\ref{model}) can be solved analytically,
299: and for DS2 we find that
300: %DX: Kc->2Kc
301: \begin{equation}
302: \omega(k_x,k_y)=2\sqrt{A^2-B^2},
303: % EC K_c \omega(k_x,k_y)=2\sqrt{A^2-B^2 + K_c (A\pm B) },
304: \end{equation}
305: where
306: \begin{eqnarray}
307: A&=&2J_b+J_c-J_d+J_d\cos(k_x-k_y) \nonumber \\
308: B&=&J_b\cos(2k_x)+J_b\cos(2k_y)+J_c\cos(k_x+k_y).
309: \end{eqnarray}
310: %EC K_c
311: %We have set $K_c = 0$ in the above equations.
312: %When $K_c \ne 0$, the single-ion anisotropy term
313: %splits the two modes of the acoustic band at low energy,
314: %with one mode remaining gapless at the IC point $(0.25,0.25)$,
315: %and the other mode developing a gap of size %DX:Kc->2Kc
316: %$2\sqrt{2(2J_b+J_c)K_c}$. Fig.~\ref{disp0} shows the dispersions along
317: %$(k_x, k_x)$ and $(k_x, -k_x)$ with nonzero $K_c$.
318: There are two different spin wave velocities for the cones emanating
319: from the IC peak $Q_{\rm IC} = (0.25,0.25)$ and symmetry-related points,
320: one corresponding to spin wave velocities
321: perpendicular to the direction of the domain walls ({\em i.e.} perpendicular
322: to the stripes), and the other parallel to the direction of the domain walls,
323: \begin{eqnarray}
324: v_{\perp}&=&4(2J_b+J_c) \nonumber \\
325: v_{\parallel}&=&4\sqrt{(2J_b+J_c)(2J_b-J_d)}~. %yao: easier to understand
326: \end{eqnarray}
327:
328:
329: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
330: \begin{figure}[htb]
331: \resizebox*{1\columnwidth}{!}{\includegraphics{ds2_cut.eps}}
332: \caption{(Color online) Constant energy cuts with windows of $0.1J_bS$ for
333: twinned diagonal, site-centered stripes of spacing
334: $p=2$ at $J_c=2 J_b$ and $J_d=-0.5J_b$.
335: % EC K_c $d=2$ at $J_c=2 J_b$, $J_d=-0.5J_b$ and $K_c=0$.
336: The energy $E$ is in units of $J_b S$.
337: \label{cut.ds2}}
338: \end{figure}
339: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
340:
341:
342: In Fig.~\ref{fig:DS2}, we show the expected dispersions and
343: scattering intensities for DS2.
344: %diagonal, {\em site-centered} stripes of spacing $p=2$ (DS2). %DX ``.'' is put behind (..).
345: Plots are shown for twinned stripes, summing the contributions
346: parallel and perpendicular to the stripe direction, {\em i.e.} along
347: the $(k_x,-k_x)$ and $(k_x,k_x)$ directions, respectively.
348: %EC We had J_b and J_c here as of 9/11/06!
349: Because $J_c$ and $J_d$ are nonzero, there is an apparent optical mode.
350: Note that it is not a true optical mode, since
351: in this configuration there are only two spins per unit cell,
352: leading to only one (twofold degenerate) band.
353: %an acoustic band with no optical modes present.
354: As in the bond-centered case, weight is forbidden at low energy at
355: the antiferromagnetic point $Q_{\rm AF}$, so that the spin-wave
356: cone emanating from this magnetic reciprocal lattice vector
357: has vanishing weight as $\omega \rightarrow 0$.
358: In Fig.~\ref{fig:DS2}(a), we have set $J_c = J_b$ and $J_d = -0.5J_b$.
359: % EC K_c
360: %(Here we use $K_c=0$ since the single ion anisotropy term has
361: %negligible effect for the apparent optical branch at the relatively high energies of interest.)
362: In Fig.~\ref{fig:DS2}(b), we use $J_c = 2J_b$ with $J_d = -0.5J_b$.
363: Notice that in panel (a) of the figure,
364: the apparent optical mode is flat.
365:
366:
367: We have chosen the coupling constants with the following in mind:
368: The ``acoustic'' branch peaks at
369: $\omega(3 \pi/4, 3 \pi/4) = 4 J_b + 2 J_c$.
370: The apparent ``optical'' mode peaks at
371: $\omega(\pi/2, 3 \pi/2) = 4 \sqrt{(2J_b-J_d)(J_c-J_d)}$.
372: The data indicate that the apparent optical mode is higher in energy than
373: the top of the ``acoustic'' part:
374: $\omega(\pi/2, 3 \pi/2) > \omega(3 \pi/4, 3 \pi/4)$, which implies that
375: \begin{equation}
376: J_d \le (1/2)(2 + J_c - \sqrt{2}\sqrt{4 + J_c^2})
377: %~~{\rm Constraint~}1
378: \label{eqn:constraint1}
379: \end{equation}
380: when $J_b = 1$.
381: However, the extra mode above $50$meV is not too high in energy,
382: so parameters need to be chosen so as to satisfy this constraint,
383: but remain close to the equality.
384:
385: We also require the apparent ``optical'' branch
386: to be concave, since there is no evidence of a dip in the extra mode.
387: This requirement gives
388: \begin{equation}
389: {\partial^2 \over \partial k_x^2} \omega (k_x, -k_x) =
390: {4 (J_d - 2)(J_c - 2 -2J_d) \over \sqrt{-(J_c - J_d)(J_d -2)}}\le0~,
391: \end{equation}
392: resulting in the constraint that
393: \begin{equation}
394: J_d \le {1 \over 2}J_c - 1~.
395: \label{eqn:constraint2}
396: \end{equation}
397:
398:
399: %%%%%%%%%%%% FIGURE: DB2 %%%%%%%%%%%%%%%%
400: \begin{figure}[t]
401: %{\centering
402: % \subfigure[$J_b = -0.1J_a$]
403: % {\resizebox*{0.45
404: % \columnwidth}{!}{\includegraphics{db2_1.eps}}}
405: % \subfigure[ $J_b = -0.5J_a$]
406: % {\resizebox*{0.45\columnwidth}{!}{\includegraphics{db2_2.eps}}}
407: % \par}
408: \psfrag{w}{$\omega$}
409: \psfrag{kx}{$k_x/2\pi$}
410: \psfrag{0.25}{$0.25$}
411: \psfrag{0.5}{$0.5$}
412: \psfrag{0.75}{$0.75$}
413: \psfrag{0}{$0$}
414: \psfrag{1}{$1$}
415: \psfrag{2}{$2$}
416: \psfrag{3}{$3$}
417: \psfrag{4}{$4$}
418: \psfrag{6}{$6$}
419: \psfrag{8}{$8$}
420: \psfrag{10}{$10$}
421: \psfrag{a}{(a)}
422: \psfrag{b}{(b)}
423: \resizebox*{1.0\columnwidth}{!}{\includegraphics{db2.eps}}
424: \caption{DB2: Diagonal, bond-centered stripes of spacing $p=2$.
425: (See Fig.~\ref{lattice}(b).) The plots are for twinned stripes,
426: summing contributions along $(k_x,k_x)$ and $(k_x,-k_x)$.
427: (a) Very weak coupling across the domain wall, with $J_b = -0.1 J_a$.
428: (b) Weak coupling across the domain wall, with $J_b = -0.5J_a$.
429: \label{fig:DB2}}
430: \end{figure}
431: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
432:
433: As long as the second constraint, Eqn.~\ref{eqn:constraint2},
434: is satisfied in the range $0\le J_c \le 2 J_b$, then the first constraint,
435: Eqn.~\ref{eqn:constraint1},
436: is also satisfied.
437: %The data appears to satisfy Constraint 2, and yet remain pretty close to the
438: %limit $J_d = {1 \over 2}J_c - 1$.
439: %Fig.~\ref{fig:DS2}(a) has a flat apparent optical band because it is at this limit.
440: %Fig.~\ref{fig:DS2}(b) has a concave optical branch, rather than a flat one.
441: To describe the data, then, we find that we need a sizeable $J_c$, on the order of
442: $J_b$.
443: %$J_c \approx (1$--$2)J_b$.
444: This makes $J_c$ significantly larger than that reported at the lower doping $x=1/3$,
445: where diagonal site-centered stripes of spacing $p=3$ (DS3) were used
446: to explain the data successfully.\cite{boothroyd05}
447: Although the fits in Ref.~\onlinecite{boothroyd05} were
448: good for $J_c =0$ and were not significantly improved
449: by letting $J_c$ increase to $J_c \approx 0.5 J_b$, the data were still well described
450: using a nonzero $J_c$.
451: We find that the data at doping $x=1/3$ are also well described by taking
452: $J_c$ to be as large as $J_b$ or even $2 J_b$
453: as in our Fig.~\ref{fig:DS2}.
454:
455:
456:
457: We also find that we need $J_d<0$, {\em i.e.} the diagonal coupling
458: parallel to the stripes needs to be ferromagnetic, in order to
459: describe the data. If one considers the diagonal spin couplings $J_c$ and $J_d$ to be derived from, {\em e.g.},
460: a perturbative treatment of a single-band Hubbard or three-band Emery model on a square Ni-O lattice,
461: we expect $J_c = J_d$. Given that the spin ground state breaks the $4$-fold rotational symmetry
462: of the square lattice to only $2$-fold rotational symmetry, any finite spin-lattice coupling
463: results in the two diagonal directions being inequivalent, and leads to $J_c \ne J_d$.
464: For weak rotational symmetry breaking of the square lattice in the perturbative regime,
465: one expects $J_c = J_o + \epsilon$ with $J_d = J_o - \epsilon$ where $\epsilon$ is small compared to $J_o$,
466: so that the anisotropy between the two diagonal coupling directions is weak as well.
467: (This preserves the spin ground state of Fig.~\ref{lattice}(a).)
468: We find, however, that this regime of the coupling constants leads to the apparent optical mode
469: being too low in energy to capture the data.
470: This may indicate that the materials are far from the perturbative limit of the
471: single-band Hubbard or three-band Emery model.
472:
473:
474:
475: % In the half-doped case, we can also evaluate the
476: % couplings and $K_c$ by fitting the dispersion and anisotropic gap. If we want the same
477: % energy scale with $J_b$ only model in the $(k_x, k_x)$ direction, we can use
478: %$w(k_x,k_x)=2(2J_b+J_c) [1-cos^2(2k_x)]^{1/2}$,
479: %where we have set $K_c = 0$ since we are interested in the high energy behavior.
480: %Compared with Eq.~(2) in
481: % Ref.~\cite{freeman05a}, we can get a relation: $4J' =2J_b+J_c$. For
482: % $J_c=2J_b$, there are $J_b=J^{\prime}=5.8$meV, $J_c=2Jb=11.6$meV,
483: % $J_d=-0.5J_b=-2.9$meV and $K_c=2K_c'=0.1$meV. If $J_c=J_b$, then $J_b=4/3
484: % J'= 7.73$meV, $J_c=J_b=7.73$meV, $J_d=-0.5J_b=-3.87$meV and
485: % $K_c=2K'_c=0.1$meV. Here $J'_b$ and $K'_c$ are values when $J_c=J_d=0$.
486:
487: %yao:
488: Fig.~\ref{cut.ds2} shows constant energy cuts for
489: DS2 corresponding to the parameters in Fig.~\ref{fig:DS2}(b).
490: %diagonal site-centered stripes of spacing $p=2$ (DS2).
491: An important feature of this configuration is that
492: although the antiferromagnetic point $Q_{\rm AF} = (0.5,0.5)$ is a magnetic reciprocal lattice vector,
493: zero-frequency weight is forbidden there by symmetry,
494: since the stripes have no net N\'eel vector at $Q_{\rm AF}$.
495: Combined with the fact that there is no optical band,
496: the DS2 configuration {\em can never have spectral weight at
497: the antiferromagnetic point} $Q_{\rm AF}$, even at finite frequency.
498: Notice that as energy is increased in Fig.~\ref{cut.ds2}, a
499: faint spin wave cone emerges from $Q_{\rm AF}$ in a
500: light ring of scattering, but none of the constant energy
501: plots show any weight at $Q_{\rm AF}$.
502: This is consistent with the constant energy plots
503: for La$_{3/2}$Sr$_{1/2}$ NiO$_4$
504: shown in Ref.~\onlinecite{freeman05a}.
505: By contrast, the corresponding bond-centered configuration
506: (DB2) shown in Fig.~\ref{lattice}(b) has an optical band which
507: displays a saddlepoint at $Q_{\rm AF}$,
508: and rather large scattering intensity at finite frequency at $Q_{\rm AF}$ as
509: a result. (See Fig.~\ref{fig:DB2} of this paper, as well as Fig.~8 of our previous paper,
510: Ref.~\onlinecite{yaocarlson06b}.)
511:
512: While diagonal, site-centered stripes of spacing $p=2$ (DS2) are able to
513: account for the behavior of the apparent optical mode observed to disperse away from
514: $Q_{\rm AF}$ in Fig.~3 of Freeman {\em et al.}\cite{freeman05a},
515: there are two other high energy features which this model has not captured.
516: One is the asymmetry in intensity observed above $30$meV
517: in the spin wave cones emanating from the main IC peaks.
518: The other is a mode in the $31-39$meV range propagating away from
519: $(h,k)$ structural reciprocal lattice points.
520: These (as well as the apparent optical mode) have been attributed to discommensurations in the magnetic order.\cite{freeman05a}
521:
522:
523: In Fig.~\ref{fig:DB2}, we show the expected dispersions and intensities
524: for the diagonal, {\em bond-centered} stripes (DB2) shown in Fig.~\ref{lattice}(b).
525: This configuration has a true optical mode.
526: Fig.~\ref{fig:DB2}(a) shows weak coupling across the charged domain walls,
527: with $J_b = -0.1 J_a$, and Fig.~\ref{fig:DB2}(b) has somewhat stronger coupling across
528: the domain walls, with $J_b = -0.5 J_a$. Note that in the bond-centered case,
529: couplings across the domain walls are ferromagnetic.
530: Results are shown for twinned stripe patterns, summing the contribution
531: parallel and perpendicular to the stripe direction, {\em i.e.} along
532: the $(k_x,-k_x)$ and $(k_x,k_x)$ directions, respectively.
533: Although there is a reciprocal lattice vector at $Q_{\rm AF}$,
534: weight is forbidden there at zero frequency, since the N\'eel vector
535: switches sign across the domain walls.
536: We have reported the
537: analytic form of the spin wave dispersion in this case
538: in a previous publication.\cite{erica04}
539:
540: Because this spin configuration
541: is only $180^o$ symmetric, the spin-wave velocity emanating from $Q_{\rm AF}$
542: is different parallel and perpendicular to the stripe direction.
543: However, the branch emanating from $Q_{\rm AF}$ in the direction parallel to stripes
544: has so little weight as to be effectively invisible in the plots.
545: This configuration displays a true optical mode because
546: there are four spins in the unit cell. The optical mode has a {\em saddlepoint} at
547: $Q_{\rm AF}$ and finite energy, with increased weight at the saddlepoint.
548: For weak coupling across the domain walls ($|J_b|<|J_a|$), the
549: optical mode always displays significant weight at $Q_{\rm AF}$ at
550: finite frequency. This is not supported by the data\cite{freeman05a},
551: which at the energies measured
552: display no scattering at $Q_{\rm AF}$ and finite frequency.
553: This likely indicates that the domain walls are not bond-centered in this material,
554: but are probably site-centered.
555:
556:
557:
558: % EC K_c
559: %%%%%%%%%%%% FIGURE %%%%%%%%%%%%%%%%
560: %\begin{figure}[htb]
561: %\psfrag{w}{$\omega$}
562: %\psfrag{kx}{$k_x/2\pi$}
563: %\psfrag{0.25}{$0.25$}
564: %\psfrag{0.5}{$0.5$}
565: %\psfrag{0.75}{$0.75$}
566: %\psfrag{0}{$0$}
567: %\psfrag{1}{$1$}
568: %\psfrag{2}{$2$}
569: %\psfrag{3}{$3$}
570: %\psfrag{4}{$4$}
571: %\psfrag{6}{$6$}
572: %\psfrag{8}{$8$}
573: %\psfrag{10}{$10$}
574: %\psfrag{a}{(a)}
575: %\psfrag{b}{(b)}
576: %\resizebox*{1.0\columnwidth}{!}{\includegraphics{dispersion0.eps}}
577: %\caption{Spin wave dispersions for diagonal site-centered
578: % stripe with stripe spacing $d=2$ $J_c=J_b$, $J_d=-0.5J_b$ and
579: % $K_c=0.01J_b$: (a)along ($k_x$, $k_x$) direction; (b) along ($k_x$,
580: % -$k_x$) direction. %DX: Kc=2Kc
581: %\label{disp0}}
582: %\end{figure}
583: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
584:
585:
586:
587: In conclusion, we have used linear spin-wave theory to
588: describe the magnetic excitations recently observed
589: in neutron scattering\cite{freeman05a} on
590: La$_{3/2}$Sr$_{1/2}$ NiO$_4$.
591: Many features of the data were captured in the spin wave analysis of Ref.~\onlinecite{freeman05a}.
592: Other features, including an apparent optical mode dispersing away
593: from $Q_{\rm AF}$ above $50$meV,
594: were attributed to discommensurations in the spin order.
595: We show here that the apparent optical mode
596: may also be captured in linear spin wave theory by using a spin coupling
597: configuration that preserves the symmetry of the spin ground state.
598: Namely, we have shown that diagonal, site-centered stripes of
599: spacing $p=2$ capture this mode when the pattern of couplings is $2$-fold symmetric.
600: This is because the $2$-fold symmetric coupling pattern gives rise to
601: two different spin-wave velocities ({\em i.e.} $v_{\parallel} \ne v_{\perp}$)
602: emanating from the antiferromagnetic point $Q_{\rm AF}=(0.5,0.5)$.
603: For twinned samples, the two velocities are simultaneously visible,
604: and the higher velocity mode $v_{\parallel}$ parallel to the stripes is responsible for the ``extra'' scattering
605: above $50$meV.
606: Furthermore, this configuration is forbidden to display scattering at
607: the antiferromagnetic point $Q_{\rm AF}$, whereas
608: bond-centered stripes have a true optical mode with significant weight at $Q_{\rm AF}$,
609: which is not supported by the data. We therefore conclude that the magnetic excitations observed
610: in Ref.~\onlinecite{freeman05a} are consistent with site-centered stripes,
611: but not with bond-centered stripes..
612:
613: %%%%%%%% \section*{Acknowledgements} %%%%%%%%%%%%
614: It is a pleasure to thank D.~K.~Campbell and
615: A.~T.~Boothroyd for helpful discussions.
616: This work was supported by Boston University (DXY), and by the Purdue Research Foundation (EWC).
617: EWC is a Cottrell Scholar of Research Corporation.
618:
619: %\bibliography{bigbib}
620: \bibliographystyle{forprb}
621:
622: \begin{thebibliography}{8}
623: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
624: \expandafter\ifx\csname bibnamefont\endcsname\relax
625: \def\bibnamefont#1{#1}\fi
626: \expandafter\ifx\csname bibfnamefont\endcsname\relax
627: \def\bibfnamefont#1{#1}\fi
628: \expandafter\ifx\csname citenamefont\endcsname\relax
629: \def\citenamefont#1{#1}\fi
630: \expandafter\ifx\csname url\endcsname\relax
631: \def\url#1{\texttt{#1}}\fi
632: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
633: \providecommand{\bibinfo}[2]{#2}
634: \providecommand{\eprint}[2][]{\url{#2}}
635:
636: \bibitem[{\citenamefont{Boothroyd
637: et~al.}(2003{\natexlab{a}})\citenamefont{Boothroyd, Prabhakaran, Freeman,
638: Lister, Enderle, Hiess, and Kulda}}]{boothroyd03a}
639: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Boothroyd}},
640: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Prabhakaran}},
641: \bibinfo{author}{\bibfnamefont{P.~G.} \bibnamefont{Freeman}},
642: \bibinfo{author}{\bibfnamefont{S.~J.~S.} \bibnamefont{Lister}},
643: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Enderle}},
644: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Hiess}}, \bibnamefont{and}
645: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Kulda}},
646: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{67}},
647: \bibinfo{pages}{100407(R)} (\bibinfo{year}{2003}{\natexlab{a}}).
648:
649: \bibitem[{\citenamefont{Woo et~al.}(2005)\citenamefont{Woo, Boothroyd,
650: Nakajima, Perring, Frost, Freeman, Prabhakaran, Yamada, and
651: Tranquada}}]{boothroyd05}
652: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Woo}},
653: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Boothroyd}},
654: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Nakajima}},
655: \bibinfo{author}{\bibfnamefont{T.~G.} \bibnamefont{Perring}},
656: \bibinfo{author}{\bibfnamefont{C.} \bibnamefont{Frost}},
657: \bibinfo{author}{\bibfnamefont{P.~G.} \bibnamefont{Freeman}},
658: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Prabhakaran}},
659: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Yamada}}, \bibnamefont{and}
660: \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{Tranquada}},
661: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{72}},
662: \bibinfo{pages}{064437} (\bibinfo{year}{2005}).
663:
664: \bibitem[{\citenamefont{Carlson et~al.}(2004)\citenamefont{Carlson, Yao, and
665: Campbell}}]{erica04}
666: \bibinfo{author}{\bibfnamefont{E.~W.} \bibnamefont{Carlson}},
667: \bibinfo{author}{\bibfnamefont{D.~X.} \bibnamefont{Yao}}, \bibnamefont{and}
668: \bibinfo{author}{\bibfnamefont{D.~K.} \bibnamefont{Campbell}},
669: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{70}},
670: \bibinfo{pages}{064505} (\bibinfo{year}{2004}).
671:
672: \bibitem[{\citenamefont{Yao et~al.}(2006)\citenamefont{Yao, Carlson, and
673: Campbell}}]{yaocarlson06b}
674: \bibinfo{author}{\bibfnamefont{D.~X.} \bibnamefont{Yao}},
675: \bibinfo{author}{\bibfnamefont{E.~W.} \bibnamefont{Carlson}},
676: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{D.~K.}
677: \bibnamefont{Campbell}}, \bibinfo{journal}{Phys. Rev. B}
678: \textbf{\bibinfo{volume}{73}}, \bibinfo{pages}{224525}
679: (\bibinfo{year}{2006}).
680:
681: \bibitem[{\citenamefont{Tranquada et~al.}(1992)\citenamefont{Tranquada,
682: Gehring, Shirane, Shamoto, and Sato}}]{tranquada}
683: \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{Tranquada}},
684: \bibinfo{author}{\bibfnamefont{P.~M.} \bibnamefont{Gehring}},
685: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Shirane}},
686: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Shamoto}}, \bibnamefont{and}
687: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Sato}},
688: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{46}},
689: \bibinfo{pages}{5561} (\bibinfo{year}{1992}).
690:
691: \bibitem[{\citenamefont{Bourges
692: et~al.}(2003{\natexlab{b}})\citenamefont{P. Bourges, Y. Sidis, M. Braden,
693: K. Nakajima and J. M. Tranquada4}}]{tranquada03}
694: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bourges}},
695: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Sidis}},
696: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Braden}},
697: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Nakajima}}, \bibnamefont{and}
698: \bibinfo{author}{\bibfnamefont{J.~M.}~\bibnamefont{Tranquada}},
699: \bibinfo{journal}{Phys. Rev. Lett. } \textbf{\bibinfo{volume}{90}},
700: \bibinfo{pages}{147202} (\bibinfo{year}{2003}{\natexlab{b}}).
701:
702: \bibitem[{\citenamefont{Kr$\ddot{u}$ger and Scheidl}(2003)}]{kruger03}
703: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Kr$\ddot{u}$ger}}
704: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Scheidl}},
705: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{67}},
706: \bibinfo{pages}{134512} (\bibinfo{year}{2003}).
707:
708: \bibitem[{\citenamefont{Freeman et~al.}(2005)\citenamefont{Freeman, Boothroyd,
709: Prabhakaran, Frost, Enderle, and Hiess}}]{freeman05a}
710: \bibinfo{author}{\bibfnamefont{P.~G.} \bibnamefont{Freeman}},
711: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Boothroyd}},
712: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Prabhakaran}},
713: \bibinfo{author}{\bibfnamefont{C.} \bibnamefont{Frost}},
714: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Enderle}}, \bibnamefont{and}
715: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Hiess}},
716: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{71}},
717: \bibinfo{pages}{174412} (\bibinfo{year}{2005}).
718:
719: \end{thebibliography}
720:
721:
722: \end{document}
723:
724:
725:
726:
727:
728:
729:
730:
731: