1: \documentclass[prl,twocolumn,aps,showpacs]{revtex4}
2: \usepackage{graphicx}
3: \begin{document}
4:
5: \title{Spinor condensates with a
6: laser-induced quadratic Zeeman effect}
7: \author{L. Santos$^{(1)}$, M. Fattori$^{(2)}$,
8: J. Stuhler$^{(2)}$\footnote{Present address: TOPTICA Photonics AG,
9: Lochhamer Schlag 19, D-82166 Gr\"afelfing, Germany.}
10: and T. Pfau$^{(2)}$} \affiliation{ \mbox{$^1$ Institut f\"ur
11: Theoretische Physik, Leibniz Universit\"at Hannover,
12: Appelstr. 2, D-30167 Hannover, Germany}\\
13: \mbox{$^2$ 5. Physikalisches Institut,
14: Universit\"at Stuttgart, Pfaffenwaldring 57 V, D-70550 Stuttgart, Germany} \\
15: }
16:
17: \begin{abstract}
18: We show that an effective quadratic Zeeman effect can be generated
19: in $^{52}$Cr by proper laser configurations, and in particular by the
20: dipole trap itself. The induced quadratic Zeeman effect
21: leads to a rich ground-state phase
22: diagram, can be used to induce topological defects by controllably
23: quenching across
24: transitions between phases of different symmetries, allows for the
25: observability of the Einstein-de Haas effect for relatively large
26: magnetic fields, and may be employed to create $S=1/2$ systems
27: with spinor dynamics. Similar ideas could be explored in other
28: atomic species opening an exciting new control tool in spinor systems.
29: \end{abstract}
30: \maketitle
31:
32: % Spinor gases. Ground state
33:
34: Spinor Bose-Einstein condensates (BEC) have
35: recently attracted a growing interest. A
36: spinor gas is formed by atoms in two or more internal states,
37: which can be simultaneously confined by optical dipole traps
38: \cite{Stenger98}. Spinor BECs present a rich variety of possible
39: ground states, including ferromagnetic and polar phases for spin-1
40: BECs \cite{Ho98,Ohmi98}, and an additional cyclic phase for the spin-2
41: case \cite{Ciobanu00,Koashi00}. Elegant topological
42: classifications of the possible spinor ground-states
43: have been recently proposed \cite{Makela,Barnett06}.
44: The spinor dynamics has been also actively
45: studied, in particular the coherent oscillations between the different
46: spinor components \cite{Dynamics}. In addition, a spinor gas has been
47: recently quenched across a transition between phases of different
48: symmetries, inducing topological defects \cite{StamperKurn,Zurek}.
49:
50: % Chromium
51:
52: The recent creation of a Chromium BEC \cite{Griesmaier05} opens
53: new interesting possibilities for the spinor physics. The ground
54: state of $^{52}$Cr is $^7$S$_3$, constituting the first accessible
55: example of a spin-3 BEC. The spin-3 BEC presents a novel rich
56: ground-state phase diagram at low magnetic fields
57: \cite{Diener06,Santos06,MakelaNew}. In particular, the existence of biaxial
58: spin-nematic phases \cite{Diener06} opens fascinating links
59: between the spin-3 BECs and the physics of liquid crystals.
60: In addition,
61: $^{52}$Cr has a large magnetic moment $\mu=6\mu_B$, where $\mu_B$ is
62: the Bohr magneton, i.e. six times larger than that of alkali atoms. The
63: corresponding large dipole-dipole interaction (DDI) can lead to
64: novel effects in the BEC physics \cite{Dipoles}. In particular,
65: dipolar effects were observed for the first time ever in quantum
66: gases in the expansion of a Chromium BEC \cite{Stuhler05}. The DDI
67: plays also a significant role in the spinor dynamics, since
68: it violates spin conservation, allowing for the transfer of spin
69: into center-of-mass angular momentum, i.e. the equivalent to the
70: Einstein-de Haas effect (EdH) \cite{Santos06,Kawaguchi06}.
71: Interestingly, the EdH and other dipolar effects
72: may be also observed in $^{87}$Rb spinor BECs
73: since, in spite of its low $\mu$, the DDI may be significant when
74: compared to the low energy scales associated with the spinor physics
75: \cite{RbEdH,UedaNew}.
76:
77: % Quadratic Zeeman effect
78:
79: In the presence of an external magnetic field, $B$, the different
80: Zeeman sublevels (with quantum number $m$)
81: of a spinor BEC acquire different shifts due to
82: the linear Zeeman effect (LZE), $\Delta E_{LZE}(m)=g \mu_B B m$,
83: with $g$ the Land\'e factor.
84: The LZE plays no role in the spinor dynamics of short-range
85: interacting BECs, because spin is never violated, and hence the
86: LZE may be gauged out. On the contrary, the dipole-induced EdH is
87: largely prevented by the LZE, even for rather low magnetic fields
88: \cite{Santos06,Kawaguchi06,UedaNew}. In addition to the LZE, and
89: due to the underlying hyperfine structure, a quadratic Zeeman
90: effect (QZE), $\Delta E_{QZE}(m)\propto B^2 m^2$ cannot be
91: neglected, since it indeed becomes very important for the
92: understanding of typical spinor BECs. The QZE is not present
93: in $^{52}$Cr due to the absence of hyperfine structure
94: but can be induced by a quasi-resonant light
95: field \cite{Cohen}. The big advantage of a light induced QZE
96: consists mainly in its tunability. In particular Gerbier et al.
97: have recently employed an off-resonant microwave field to
98: induce a QZE of the appropriate sign and resonantly control the
99: spinor dynamics in Rb atoms in optical lattices \cite{Bloch}.
100:
101: % This Letter
102:
103: In this Letter, we show that a light-induced QZE, tuned
104: independently from the magnetic field, opens promising ways of
105: control for $^{52}$Cr, and in general for other spinor gases. In
106: the first part of the Letter we analyze the rich ground-state
107: phase diagram introduced by the QZE. In particular
108: modifications of the induced QZE may quench across
109: transitions between phases with different symmetries,
110: and could lead to the observation of topological defects
111: \cite{StamperKurn,Zurek}. In the second part of the Letter,
112: we discuss how the effective QZE may be manipulated such that a
113: large EdH effect may be observed even in the presence of a relatively large,
114: and even fluctuating magnetic field. Moreover,
115: we show that the induced QZE can be used to engineer
116: $S=1/2$ systems with spinor dynamics, differing significantly from
117: standard binary BECs \cite{Hall98}, where spinor dynamics
118: is absent.
119:
120:
121: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
122:
123: %% The induced quadratic Zeeman effect
124:
125: The ground-state of $^{52}$Cr ($^{7}S_3$) can be
126: coupled by optical dipole transitions to different excited
127: P-states ($^{7}$P$_{2,3,4}$). For sufficiently detuned lasers from
128: these transitions, an $m$-dependent Stark shift can be
129: induced in the $^{7}S_3$ manifold. We show in the following that
130: such a shift mimics a QZE in $^{52}$Cr, despite the absence of
131: nuclear spin.
132: To illustrate this fact, we shall initially simplify the actual experimental
133: situation (which is discussed in detail below), and consider a
134: $\pi$-polarized laser (in the $z$-direction) on the particular
135: optical transition $^{7}$S$_3\leftrightarrow ^{7}$P$_{3}$.
136: For a sufficiently large laser detuning $\Delta$
137: from the transition frequency, the induced Stark shift for a given
138: state $m$ of the $^{7}$S$_3$ manifold is provided by $\Delta
139: E(m)\propto\hbar\Omega_0(m)^2/\Delta$, where the Rabi frequency is given by
140: $\Omega_0=(e E/\hbar) \langle^{7}P_{3},m|z|^{7}S_{3},m\rangle$
141: (due to selection rules only transitions to the same $m$ are
142: possible in the example considered), where $e$ is the electron
143: charge, and $E$ is the electric field of the laser. The
144: optical dipole moment of the transition is easily calculate from the
145: corresponding Clebsch-Gordan coefficients,
146: $\langle J=3,m;j=1,0|J=3,m\rangle =m/2\sqrt{3}$.
147: As a consequence a QZE $\Delta E(m)=\alpha m^2$ is induced,
148: where $\alpha$ is a function of the applied
149: intensity and the detuning, but independent of the magnetic field.
150: Similarly, for transitions to other $^{7}P$ states, with other
151: polarizations and detunings, general shifts of the form $\Delta
152: E(m)=\gamma m+\alpha m^2$ can be generated. In the following we
153: absorb $\gamma$ in the LZE.
154:
155: %% FIGURE 1
156: \begin{figure}[ht]
157: \begin{center}
158: \vspace*{-0.3cm} \hspace*{-3.0cm}
159: \includegraphics[width=6.5cm]{fig1.eps}
160: \end{center}
161: \vspace*{-0.3cm}
162: \caption{Phases as a function of $\tilde p$ and $\tilde\alpha$. See text for details.}
163: \label{fig:1}
164: \end{figure}
165:
166: % Experimental discussion
167:
168: Exact calculations taking into account the excited states
169: $z^{7}$P$_{2,3,4}$ and $y^{7}$P$_{2,3,4}$ show that $\alpha /2\mu_B
170: \approx \pm 2$ mG can be obtained with few mW of $\pi$ polarized
171: light with a wavelength of $430$ nm (red detuning
172: respect the $z^{7}P$ states) and $424$ nm (blue detuning), respectively.
173: However, a heating rate of $\approx 2$ $\mu$K/s due to off-resonance light
174: scattering would destroy the condensate in few tens of ms. To
175: increase the lifetime we can increase the detuning and the applied
176: light power. Interestingly, the dipole trap laser at $1064$ nm
177: normally used to condense $^{52}$Cr atoms \cite{Griesmaier05} for
178: $20$ W focused to $30 \mu$m induces a QZE $\alpha /2\mu_B \approx 7
179: $ mG for $\pi$ polarization and $\alpha /2\mu_B \approx -3$ mG for
180: a combination of $\sigma^+$ and $\sigma^-$ polarizations. This is
181: due to a discrepancy of $10\%$ between the values of the linewidths
182: of the excited states $y^{7}P_{2,3,4}$ \cite{Becker}. For such
183: dipole trap the lifetime of the condensate can be a few seconds.
184: Using combinations of different polarizations to change $\alpha$
185: can cause two photon Raman coupling between different sublevels
186: severely affecting the spinor ground state. The latter can be prevented by
187: combining different lasers with orthogonal polarizations and with
188: randomized relative phases.
189:
190:
191: % Hamiltonian
192:
193: In the following we consider an optically trapped Chromium BEC
194: with $N$ particles under the influence of the above mentioned QZE.
195: The corresponding Hamiltonian is $\hat H=\hat H_0+\hat V_{sr}+\hat
196: V_{dd}$. The single-particle part, $\hat H_0$, includes the
197: trapping energy and the LZE and QZE:
198: \begin{equation}
199: \hat H_0=\int d{\bf r} \sum_m \hat\psi_m^\dag \left
200: [\frac{-\hbar^2\nabla^2}{2M}+U_{trap}+pm +\alpha m^2\right ]
201: \hat\psi_m,
202: \end{equation}
203: where $\hat \psi_m^\dag$ ($\hat\psi_m$) is the creation
204: (annihilation) operator in the $m$ state, $M$ is the atomic mass,
205: $U_{trap}(\vec r)$ is the trapping potential, and $p=g\mu_B B+\gamma$, with
206: $g=2$ for $^{52}$Cr. The short-range interactions are given by
207: \cite{Ho98}
208: \begin{equation}
209: \hat V_{sr}=\frac{1}{2}\int d{\bf r} \sum_{S=0}^6 g_S \hat{\cal
210: P}_S({\bf r}),
211: \end{equation}
212: where $\hat{\cal P}_S$ is the projector on the total spin $S$
213: ($=0,2,4,6$), $g_S=4\pi\hbar^2a_S/M$, and $a_S$ is the $s$-wave
214: scattering length for a total spin $S$. The DDI $\hat V_{dd}$ is
215: given by
216: \begin{eqnarray}
217: \hat V_{dd}&=&\frac{c_d}{2}\int d{\bf r}\int d{\bf r}'
218: \frac{1}{|{\bf r}-{\bf r}'|^3}
219: \hat\psi_m^\dag ({\bf r})\hat\psi_{m'}^\dag ({\bf r}') \nonumber \\
220: &&\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\! \!\!\!\!\left [ {\bf
221: S}_{mn}\cdot{\bf S}_{m'n'}-3({\bf S}_{mn}\cdot{\bf e}) ({\bf
222: S}_{m'n'}\cdot{\bf e}) \right ] \hat\psi_n ({\bf
223: r})\hat\psi_n'({\bf r}'),
224: \end{eqnarray}
225: where ${\bf S}=(S_x,S_y,S_z)$, $S_{x,y,z}$ are the spin-3 matrices,
226: $c_d=\mu_0\mu_B^2g^2/4\pi$ ($=0.004g_6$ for $^{52}$Cr),
227: with $\mu_0$ the magnetic permeability of vacuum, and
228: ${\bf e}=({\bf r}-{\bf r}')/|{\bf r}-{\bf r}'|$.
229:
230: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
231:
232: % Ground-state discussion
233:
234: We first discuss the ground-state of the spin-3 BEC. We consider
235: mean-field (MF) approximation $\hat\psi_m({\bf
236: r})\simeq\sqrt{N}\psi_m({\bf r})$. In order to simplify the
237: analysis of the possible ground-state solutions we employ the
238: single-mode approximation (SMA): $\psi_m({\bf r})=\Phi({\bf
239: r})\psi_m$, with $\int d{\bf r} |\Phi({\bf r})|^2=1$, $\beta=\int
240: d{\bf r} |\Phi({\bf r})|^4$. Considering a magnetic field in the
241: $z$-direction, and following the procedure discussed in
242: Ref.~\cite{Santos06}, we obtain the expression for the energy per
243: particle $E=N\beta\epsilon/2$
244: \begin{eqnarray}
245: \epsilon&=&\tilde p \langle S_z \rangle + \tilde\alpha \langle
246: S_z^2\rangle +\tilde c_1 \langle S_z \rangle ^2+
247: \frac{4c_2}{7}|\Theta|^2 \nonumber \\
248: &+&c_3 \left ( \frac{3\langle S_z^2\rangle ^2}{2} -12\langle
249: S_z^2\rangle + \frac{|\langle S_+^2\rangle|^2}{2}+ 2|\langle S_+
250: S_z \rangle|^2 \right), \label{eps}
251: \end{eqnarray}
252: where $S_+=S_x+iS_y$, $\Theta=\sum_m (-1)^m
253: \psi_m\psi_{-m}/2$, $\tilde p=2p/N\beta$,
254: $\tilde\alpha=2\alpha/N\beta$, $\tilde c_1\simeq c_1-c_3/2$,
255: $c_1=(g_6-g_2)/18$, $c_2=g_0+(-55g_2+27g_4-5g_6)/33$,
256: $c_3=g_2/126-g_4/77+g_6/198$. For the case of $^{52}$Cr
257: \cite{Werner05} $c_0\simeq 0.65g_6$, $c_1\simeq 0.059g_6$,
258: $c_2\simeq g_0+0.374g_6$, and $c_3\simeq -0.002g_6$. The value of
259: $a_0$ is unknown. In the following we assume $g_0=g_6$. Compared
260: to the expression of Ref.~\cite{Santos06}, (\ref{eps}) contains an
261: extra term $\tilde\alpha\langle S_z^2\rangle$ corresponding to the
262: induced QZE. This new term leads to a rich physics of new phases.
263:
264: We have minimized by means of simulated annealing Eq.~(\ref{eps})
265: with respect to $\psi_m$, under the constraints
266: $\sum_m|\psi_m|^2=1$ and $\langle S_+\rangle =0$.
267: Fig.~\ref{fig:1} shows the corresponding phase diagram as a
268: function of the LZE ($\tilde p$) and the QZE ($\tilde\alpha$). For
269: sufficiently large values of $\tilde\alpha$ and $\tilde p$, different
270: ferromagnetic phases ($F_m$) are possible: $F_{-3}$ ($\tilde
271: p>5\tilde c_1+5\tilde\alpha+75c_3/2$), $F_{-2}$ ($3\tilde
272: c_1+3\tilde\alpha-27c_3/2<\tilde p< 5\tilde c_1+5\tilde\alpha+75 c_3 /2$),
273: $F_{-1}$ ($\tilde\alpha+\tilde c_1-21c_3/2-c_2/7<\tilde p<3\tilde
274: c_1+3\tilde\alpha-27c_3/2$), and $F_0$ ($\tilde p < \tilde\alpha+
275: \tilde c_1-21c_3/2-c_2/7$). For sufficiently negative $\tilde\alpha$, a polar
276: phase ($P$) transforms continuously into a ferromagnetic phase
277: ($F_{-3}$) (the transformation is complete at $\tilde p=6\tilde
278: c_1-2c_2/21\simeq 0.22g_6$). In addition to these phases, as shown in
279: Fig.~\ref{fig:1}, uniaxial ($U$) and biaxial ($B$) spin-nematic
280: phases are possible \cite{Diener06}, depending on the eigenvalues
281: $\lambda_{1,2,3}$ of the nematic tensor $\langle
282: S_iS_j-S_jS_i\rangle /2$ ($i,j=x,y,z$). The phase $U_1$
283: ($CY_{-3,-2}$ in Ref.~\cite{Santos06}) fulfills
284: $\lambda_1>\lambda_2=\lambda_3$, and transforms continuously into
285: $F_{-3}$. The phase $U_3$ ($CY_{-1,2}$ in Ref.~\cite{Santos06}) is
286: a discotic phase satisfying $\lambda_1=\lambda_2>\lambda_3$, and
287: transforms continuously into $F_{-1}$. Note that $U_1$ and $U_3$
288: become degenerated at $\tilde\alpha=0$ as pointed out in
289: Ref.~\cite{Santos06}. In addition, different biaxial phases (with
290: $\lambda_1\neq\lambda_2\neq\lambda_3$) occur, characterized by
291: $|\langle S_+^2\rangle |\neq 0$. $B_1^{(1)}$ is basically a
292: modification of $P$, $B_2^{(1)}$ is the $CY_{-3,-1,1,3}$ phase in
293: Ref.~\cite{Santos06}, $B_3^{(1)}$ is a modification of $F_0$. At
294: the boundaries between $F_{-3}$, $F_{-2}$ and $F_{-1}$ other
295: biaxial phases appear, $B_1^{(2)}$, $B_2^{(2)}$ and $B_3^{(2)}$,
296: which are respectively modifications of $F_{-3}$, $F_{-2}$ and
297: $F_{-1}$. Although a detailed analysis of the symmetries of the
298: different phases is beyond the scope of this Letter, we would like
299: to note, that following the classification scheme proposed in
300: Ref.~\cite{Barnett06}, the phases $P$, $U_1$ and $U_3$ have very
301: different symmetry properties, in particular $P$ transforms as an
302: hexagon, $U_1$ as a pyramid with pentagonal base, and $U_3$ as a
303: tetrahedron, and hence a controlled change in the induced QZE at
304: low magnetic fields may lead to very interesting quantum phase-transition
305: dynamics \cite{FootnoteZurek}, which is left for further
306: investigations.
307:
308: The induced QZE can play an important role in the observation of the EdH.
309: As discussed in Refs.~\cite{Kawaguchi06,Santos06}, the
310: DDI violates spin conservation. If the
311: Hamiltonian preserves a cylindrical symmetry around the dipole
312: direction, the conservation of the total angular momentum, leads
313: to the transfer of spin into center-of-mass angular momentum,
314: resembling the EdH. However,
315: Larmor precession prevents the EdH even for relatively
316: small $B$ (although other interesting dipolar effects
317: may occur even for a fixed magnetization \cite{UedaNew}).
318: We show in the following that the
319: QZE may be employed to induce under realistic conditions a
320: degeneracy between two neighboring states of the ground-state
321: manifold, allowing for a large EdH even for relatively
322: large magnetic fields.
323:
324: Following Ref.~\cite{Santos06} we analyze the spinor dynamics
325: within the MF approximation, but abandoning the SMA. The dynamics
326: of the different components is provided by seven coupled non local
327: non linear Schr\"odinger equations:
328: \begin{eqnarray}
329: i\hbar\frac{\partial}{\partial t}\psi_m({\bf r})&=& \left [
330: \frac{-\hbar^2\nabla^2}{2M}+U_{trap}+pm +\alpha m^2\right] \psi_m
331: \nonumber \\
332: %
333: &+& N \left [
334: c_0 n+ m(c_1 f_z+c_d {\cal A}_0) \right ] \psi_m \nonumber \\
335: %
336: &+&\frac{N}{2}
337: \left[c_1 f_-+ 2c_d {\cal A}_{-} \right ] S^+_{m,m-1} \psi_{m-1}\nonumber \\
338: %%
339: &+& \frac{N}{2}
340: \left[c_1f_+ + 2c_d {\cal A}_{+}\right ] S^-_{m,m+1}\psi_{m+1}\nonumber \\
341: %%
342: &+& (-1)^m \frac{2N c_2}{7}s_-\psi_{-m}^* \nonumber \\
343: &+& N c_3 \sum_{n} \sum_{i,j}O_{ij}(S^i S^j)_{mn}\psi_n,
344: \label{dynamic}
345: \end{eqnarray}
346: where $n(\bf r)$ is the total density, $f_i(\bf r)=\langle S_i
347: (\bf r)\rangle$, ${\cal A}_0=\sqrt{6\pi/5}
348: [\sqrt{8/3}\Gamma_{0,z}+\Gamma_{1,-}+\Gamma_{-1,+}]$, ${\cal
349: A}_\pm=\sqrt{6\pi/5}
350: [-\Gamma_{0,\pm}/\sqrt{6}\mp\Gamma_{\pm,z}+\Gamma_{\pm 2,\mp}]$,
351: $\Gamma_{m,i}=\int d{\bf r}' f_i({\bf r}') Y_{2m}({\bf r}-{\bf
352: r}')/|{\bf r}-{\bf r}'|^3$, $S^\pm_{m,m\mp
353: 1}=\sqrt{12-m(m\mp 1)}$, with $Y_{2m}$ spherical harmonics.
354:
355: We consider at $t=0$ all atoms in $m=-3$. Without DDI, spin
356: conservation restricts the system to $m=-3$ (scalar BEC). The DDI
357: allows for a transfer into $m=-2$, although Larmor precession
358: limits this transfer to very small magnetic fields ($B\simeq 0.1$
359: mG). In the following, we show that the QZE optimizes such
360: transfer for much larger $B$. For simplicity we consider a 2D BEC
361: \cite{footnote2D}, i.e. we assume in the $z$-direction a strong
362: harmonic potential of frequency $\omega_z$. Hence $\psi_m({\bf
363: r})=\phi_0(z)\psi_m({\bf \rho})$, where
364: $\phi_0(z)=\exp[-z^2/2l_z^2]/\pi^{1/4}\sqrt{l_z}$, with
365: $l_z=\sqrt{\hbar/m\omega_z}$. We consider an additional harmonic
366: confinement of frequency $\omega_{xy}$ on the $xy$-plane.
367: Fig.~\ref{fig:2} shows the population in $m=-3$, for the
368: particular case of $\omega_z=2$kHz, $\omega_{xy}=50$Hz, and
369: $N=2\times 10^4$ atoms. We consider the dipole direction in the
370: $y$-direction. If $p=0$ and $\alpha=0$ a clear transfer from
371: $m=-3$ to other modes is observed due to the coherent spin
372: relaxation induced by the DDI. However, if a magnetic field of
373: $20$mG is applied in the absence of QZE the transfer to other
374: modes is completely suppressed, and a scalar BEC in $m=-3$ is
375: recovered. The presence of the induced QZE alters the situation
376: significantly, since for $\alpha=p/5$, the $m=-3$ and $m=-2$
377: states become degenerated. In that case a significant population
378: is again transferred between $m=-3$ and $m=-2$. Note, however,
379: that due to technical reasons the magnetic field typically
380: fluctuates around a given value, and hence the degeneration is not
381: fulfilled at any time. We have taken this into account in our
382: simulations by randomly variating $B=B_0+\delta B$, for $\delta
383: B=1$ mG, such that $B_0$ and $\alpha$ satisfy the degeneration
384: condition. As shown in Fig.~\ref{fig:2} as long as the
385: degeneration is fulfilled in average, the induced QZE allows for a
386: large spin relaxation even for large and even fluctuating
387: magnetic fields. It is crucial, however,
388: to control the external magnetic fields to
389: avoid spurious polarization components in every laser, since e.g.
390: a residual magnetic field of $100\mu$G transversal to the quantization
391: axis would induce a coupling of $100 s^{-1}$ between the $m$ sublevels.
392: This (single-particle) effect could obscure the EdH.
393: However, the time-scale for the EdH, $\tau_{EdH}$, is
394: inversely proportional to the atomic density, contrary to the
395: spurious single-particle transfer time-scale, $\tau_{SP}$,
396: which is independent of it. Hence a sufficiently large
397: density can allow for $\tau_{EdH}<\tau_{SP}$, and hence a clear observation of the EdH.
398:
399: %% FIGURE 2
400: \begin{figure}[ht]
401: \begin{center}
402: %\vspace*{0.1cm}
403: \includegraphics[width=5.5cm,angle=0]{fig2.eps}
404: \end{center}
405: \vspace*{-0.2cm}
406: \caption{Evolution of the atomic fraction in $m=-3$, for
407: $\omega_z=2$kHz, $\omega_{xy}=50$Hz, $N=2\times 10^4$, with the
408: dipole along $y$, for $B=\alpha=0$ (dashed-dotted), $B=20$mG and
409: $\alpha=0$ (dotted), $B=20$mG and $\alpha=p/5$ (dashed), and
410: $\alpha=p/5$, and $B=20$mG with $\pm 1$mG of random fluctuation
411: (solid).} \label{fig:2}
412: \end{figure}
413:
414: As mentioned above, $\alpha=p/5$ induces a
415: degeneracy between the doublet $\{-3,-2\}$. The next Zeeman state,
416: $m=-1$, is separated by a gap $2p/5$ from the doublet, and
417: hence (even for low $B$) the combination of LZE and QZE shields
418: the doublet from the other Zeeman states, transforming the $S=3$
419: problem into an effective $S=1/2$ one. Note that contrary to
420: typical $S=1/2$ systems, in which there is no spinor dynamics
421: \cite{Hall98}, spin relaxation couples both levels. In this sense,
422: the induced QZE allows for a fundamentally new physical situation,
423: indeed the simplest spinor BEC system with spinor dynamics.
424: Moreover, an effective spin-$1/2$ system may be generated for any pair $\{
425: m,m+1\}$ if $\alpha=-p/(2m+1)$, i.e. six $S=1/2$-systems with
426: different collisional properties are possible. They will be
427: studied in detail elsewhere.
428:
429: Summarizing, a QZE can be induced in $^{52}$Cr by
430: proper laser configurations, in particular by the dipole trap
431: itself. The QZE can be controlled independently of the magnetic
432: field, leads to a rich variety of ground-state phases, can
433: be used to rapidly quench through quantum phase transitions,
434: allows for an observable EdH
435: effect for relatively large magnetic fields, and permits $S=1/2$
436: systems with spinor dynamics. Similar ideas could be explored in
437: other atomic species opening an exciting new control tool in
438: spinor systems.
439:
440: We would like to thank W. Zurek for enlightening discussions, and
441: the German Science Foundation (DFG) (SPP1116, SFB/TR 21 and
442: SFB407) for support.
443:
444: \begin{thebibliography}{99}
445:
446: \bibitem{Stenger98} J. Stenger {\it et al.}, Nature {\bf 396}, 345 (1998).
447:
448: \bibitem{Ho98} T.-L. Ho, Phys. Rev. Lett. {\bf 81}, 742 (1998).
449:
450: \bibitem{Ohmi98} T. Ohmi and K. Machida,
451: J. Phys. Soc. Jpn. {\bf 67}, 1822 (1998).
452:
453: \bibitem{Ciobanu00} C. V. Ciobanu, S.-K. Yip, and T.-L. Ho, Phys. Rev. A,
454: {\bf 61}, 033607 (2000).
455:
456: \bibitem{Koashi00} M. Koashi and M. Ueda, Phys. Rev. Lett. {\bf 84},
457: 1066 (2000); M. Ueda and M. Koashi, Phys. Rev. A {\bf 65}, 063602 (2002).
458:
459: \bibitem{Makela} Y. Zhang, H. M\"akel\"a, and K.-A. Suominen,
460: "Progress in Ferromagnetic Research",
461: Frank Columbus, editor (Nova Science Publishers, New York, 2004).
462:
463: \bibitem{Barnett06} R. Barnett, A. Turner, and E. Demler
464: Phys. Rev. Lett. {\bf 97}, 180412 (2006).
465:
466: \bibitem{Dynamics} M. D. Barret, J. A. Sauer, and M. S. Chapman,
467: Phys. Rev. Lett. {\bf 87}, 010404 (2001); H. Schmaljohann {\it et
468: al.}, Phys. Rev. Lett. {\bf 92}, 040402 (2004); M.-S. Chang {\it
469: et al.}, Phys. Rev. Lett. {\bf 92}, 140403 (2004); T. Kuwamoto
470: {\it et al.}, Phys. Rev. A {\bf 69}, 063604 (2004); M. H. Wheeler
471: {\it et al.}, Phys. Rev. Lett. {\bf 93}, 170402 (2004); J. M.
472: Higbie {\it et al.} Phys. Rev. Lett. {\bf 95}, 050401 (2005); A.
473: Widera {\it et al.}, Phys. Rev. Lett. {\bf 95}, 190405 (2005).
474:
475: \bibitem{StamperKurn} L. E. Sadler {\it et al.},
476: Nature {\bf 443}, 312 (2006).
477:
478: \bibitem{Zurek} W. H. Zurek, U. Dorner and P. Zoller,
479: Phys. Rev. Lett. {\bf 95}, 105701 (2005).
480:
481: \bibitem{Griesmaier05} A. Griesmaier {\it et al.}, Phys. Rev. Lett. {\bf 94},
482: 160401 (2005).
483:
484: \bibitem{Diener06} R. Diener and J. Ho, Phys. Rev. Lett. {\bf 96},
485: 190405 (2006).
486:
487: \bibitem{Santos06} L. Santos and T. Pfau, Phys. Rev. Lett. {\bf 96}, 190404
488: (2006).
489:
490: \bibitem{MakelaNew} H. M\"akel\"a, and K.-A. Suominen, cond-mat/0610071.
491:
492: \bibitem{Dipoles} S. Yi and L. You, Phys. Rev. A {\bf 61}, 041604 (2000);
493: K. G\'oral, K. Rz\c a\.zewski, and T. Pfau, Phys. Rev. A
494: {\bf 61}, 051601(R) (2000); L. Santos {\it et al.}, Phys. Rev.
495: Lett. {\bf 85}, 1791 (2000); L. Santos, G. V. Shlyapnikov, and M.
496: Lewenstein, Phys. Rev. Lett. {\bf 90}, 250401 (2003).
497:
498: \bibitem{Stuhler05} J. Stuhler {\it et al.} Phys. Rev. Lett. {\bf 95}, 150406 (2005).
499:
500: \bibitem{Kawaguchi06} Y. Kawaguchi, H. Saito and M. Ueda,
501: Phys. Rev. Lett. {\bf 96}, 080405 (2006).
502:
503: \bibitem{RbEdH} S. Yi and H. Pu, Phys. Rev. A {\bf 73}, 023602 (2006);
504: Y. Kawaguchi, H. Saito, and M. Ueda, Phys. Rev. Lett. {\bf 97}, 130404 (2006);
505: K. Gawryluk, M. Brewcyk, K. Bongs, and M. Gajda, cond-mat/0609061.
506:
507: \bibitem{UedaNew} Y. Kawaguchi, H. Saito and M. Ueda, cond-mat/0611131.
508:
509: \bibitem{Cohen} C. C. Tannoudji and J. Dupont-Roc, Phys. Rev. A {\bf 5},
510: 968 (1972).
511:
512: \bibitem{Bloch} F. Gerbier {\it et al.}, Phys. Rev. A {\bf 73},
513: 041602(R) (2006).
514:
515:
516: \bibitem{Hall98} D. S. Hall {\it et al.}, Phys. Rev. Lett. {\bf 81},
517: 1539 (1998).
518:
519: \bibitem{Becker} U. Becker, H. Bucka, and A. Schmidt,
520: Astron. Astrophys. {\bf 59}, 145 (1977).
521:
522: \bibitem{Werner05} J. Werner {\it et al.},
523: Phys. Rev. Lett. {\bf 94}, 183201 (2005).
524:
525: \bibitem{FootnoteZurek} W. Zurek, private communication.
526:
527: \bibitem{footnote2D} Our 2D results on population transfer
528: can be extrapolated to 3D.
529: However, the 2D case leads to interesting physics in itself.
530: If the dipole direction, $\hat d$, is on the $xy$-plane, the
531: cylindrical symmetry around $\hat d$ is broken, and
532: the transfer $-3$ $\rightarrow$ $-2$ does not lead to a rotation of
533: the $-2$ cloud (although non-rotating
534: patterns are observed). If $\hat d$ is along $z$, the
535: cylindrical symmetry is preserved, and the system rotates, but
536: the transfer is slowed
537: down, due to the averaging of the DDI \cite{Santos06}.
538:
539: \end{thebibliography}
540:
541: \end{document}