1: % ****** Start of file apssamp.tex ******
2: %
3: % This file is part of the APS files in the REVTeX 4 distribution.
4: % Version 4.0 of REVTeX, August 2001
5: %
6: % Copyright (c) 2001 The American Physical Society.
7: %
8: % See the REVTeX 4 README file for restrictions and more information.
9: %
10: % TeX'ing this file requires that you have AMS-LaTeX 2.0 installed
11: % as well as the rest of the prerequisites for REVTeX 4.0
12: %
13: % See the REVTeX 4 README file
14: % It also requires running BibTeX. The commands are as follows:
15: %
16: % 1) latex apssamp.tex
17: % 2) bibtex apssamp
18: % 3) latex apssamp.tex
19: % 4) latex apssamp.tex
20: %
21: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
22: %\documentclass[preprint,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
23:
24: % Some other (several out of many) possibilities
25: %\documentclass[preprint,aps]{revtex4}
26: %\documentclass[preprint,aps,draft]{revtex4}
27: %\documentclass[prb]{revtex4}% Physical Review B
28:
29: \usepackage{graphicx}% Include figure files
30: \usepackage{dcolumn}% Align table columns on decimal point
31: \usepackage{bm}% bold math
32:
33: %\nofiles
34:
35: \begin{document}
36:
37: %\preprint{APS/123-QED}
38:
39: \title{Magnetically-controlled velocity selection in a cold atom
40: sample using stimulated Raman transitions}
41:
42: \author{Matthew L. Terraciano}
43: \author{Spencer E. Olson}%
44: \author{Mark Bashkansky}%
45: \author{Zachary Dutton}%
46: \author{Fredrik K. Fatemi}%
47:
48: \affiliation{Naval Research Laboratory, 4555 Overlook Ave. S.W.,
49: Washington, DC 20375}
50:
51: \date{\today}% It is always \today, today,
52: % but any date may be explicitly specified
53:
54: \begin{abstract}
55: We observe velocity-selective two-photon resonances in a cold atom
56: cloud in the presence of a magnetic field. We use these resonances
57: to demonstrate a simple magnetometer with sub-mG resolution. The
58: technique is particularly useful for zeroing the magnetic field and
59: does not require any additional laser frequencies than are already
60: used for standard magneto-optical traps. We verify the effects
61: using Faraday rotation spectroscopy.
62: \end{abstract}
63:
64: \pacs{42.50.Vk, 32.60.+i}% PACS, the Physics and Astronomy
65: % Classification Scheme.
66: %\keywords{Suggested keywords}%Use showkeys class option if keyword
67: %display desired
68: \maketitle
69:
70: \section{\label{sec:level1}Introduction}
71:
72:
73: Stimulated Raman transitions that couple atomic ground states with
74: counterpropagating laser beams are resonant only within a narrow
75: velocity band. This atomic velocity selection~\cite{Kasevich} has
76: proven to be a useful tool for a variety of experiments, including
77: subrecoil Raman cooling~\cite{boyer:043405}, atom
78: interferometry~\cite{McQuirk}, and atom velocimetry~\cite{Chabe}.
79: Stray magnetic fields can adversely affect this process by shifting
80: the magnetic sublevels, thereby perturbing the participating
81: velocity bands~\cite{boyer:043405, Chabe, Garreau}. Conversely, when
82: transitions occur between different magnetic sublevels of a single
83: hyperfine level, velocity selectivity can provide an excellent
84: measure of stray or applied magnetic fields.
85:
86: Elimination of stray fields to sub-milliGauss levels is particularly
87: important for sub-recoil cooling processes~\cite{boyer:043405,
88: Vuletic}. Typically, these fields are nulled by Helmholtz coils
89: along each cartesian direction. Correct compensation currents can
90: roughly be estimated by visual indicators such as atom expansion in
91: an optical molasses, but these cues are strongly dependent on
92: optical alignment. Stray fields can be directly measured using, for
93: example, Faraday spectroscopy, which provides picoTesla
94: sensitivity~\cite{Isayama, Jessen, Kaiser}, but requires additional
95: laser frequencies and time-resolved polarimetry. Measurement of
96: vector magnetic fields with magneto-resistive probes has been used
97: for active compensation of both DC and AC fields, but needs several
98: sensors placed externally to the vacuum chamber~\cite{Garreau}.
99:
100: In this paper, we describe a simple imaging technique for measuring
101: magnetic fields with sub-milliGauss resolution using a sample of
102: cold atoms from a point trap. The technique relies on
103: velocity-selective two-photon resonances \cite{Kasevich} (VSTPR) in
104: a magnetic field, where the two-photon resonance occurs between
105: different magnetic sublevels within a single hyperfine level. When
106: applied to atoms cooled in an alkali-vapor MOT, no additional laser
107: frequencies are required, because the VSTPR pulse can be derived
108: from the repumping laser beams. The two main requirements are 1)
109: VSTPR beams along a horizontal axis and 2) a CCD camera whose
110: optical axis is orthogonal to the propagation direction of the VSTPR
111: beams.
112:
113: \section{\label{sec:level2}Background}
114:
115: \begin{figure}
116: \includegraphics[width=\columnwidth]{fig1.jpg}
117: \caption{\label{fig:EnergyLevelDiagram} Relevant energy levels for
118: the Raman transitions. Magnetic sublevels are split by the Larmor
119: precession frequency, $\omega_L$. The polarization configuration is
120: lin~$\perp$~lin. The energy scale is exaggerated for clarity.}
121: \end{figure}
122:
123: In this section, we briefly describe the process.
124: Figure~\ref{fig:EnergyLevelDiagram} shows the relevant energy
125: levels. An atom moving with velocity \textbf{v} along the $x$-axis
126: is exposed to a light field composed of two counterpropagating laser
127: beams with wave vectors $\mathbf{k_1}$ and $\mathbf{k_2}$, where
128: $\mathbf{k_1}\simeq-\mathbf{k_2}\simeq\mathbf{k}$ along the
129: $x$-axis. The polarizations of the two beams are lin~$\perp$~lin.
130: For an arbitrary B-field, this polarization configuration couples
131: $m_f - m_i = \Delta{m} = 0, {\pm}1, {\pm}2$ magnetic sublevels of a
132: single hyperfine level, where we choose our quantization axis along
133: the magnetic field. Absorption of a photon from one beam and
134: emission into the other results in a linear momentum change of
135: $\pm\hbar(\mathbf{k_1} - \mathbf{k_2}) \approx \pm2\hbar\mathbf{k} =
136: \pm2Mv_r\mathbf{\hat{x}}$, where $v_r$ is the recoil velocity and
137: $M$ is the mass. The one-photon detuning, $\Delta$ is chosen to be
138: much larger than the hyperfine splittings of the upper state.
139:
140: The two-photon detuning is defined here as $\delta = \omega_1 -
141: \omega_2 - (\Delta{m})\omega_L$, where $\omega_L$ is the Zeeman
142: splitting. In a small magnetic field, $\hbar\omega_L=g_F{\mu}_B{B}$,
143: where $g_F$ is the gyromagnetic ratio, and $\mu_B$ is the Bohr
144: magneton. For $^{85}$Rb, $g_F{\mu}_B/\hbar$ = 466.74
145: kHz/Gauss~\cite{gF}. Two-photon resonance occurs for atoms whose
146: velocity satisfies
147: \begin{equation}
148: \delta = \delta_{LS} + \delta_D + 4\delta_{r},
149: \end{equation}
150:
151: \noindent where the two-photon Doppler shift, $\delta_D =
152: 2\mathbf{k}{\cdot}\mathbf{v}$, and the recoil frequency $\delta_{r}
153: = {\hbar}k^2/2M$. The relative light shift, $\delta_{LS}$ is a weak
154: function of the participating magnetic sublevels, which are both in
155: the same hyperfine ground state. Apart from $\delta_{LS}$, and with
156: $\omega_1 = \omega_2$, the resonance condition is satisfied for
157: atoms with $\mathbf{v} = \mathbf{v_0}\pm{v_r\mathbf{\hat{x}}}$ such
158: that $2\mathbf{k}{\cdot}\mathbf{v_0} = (\Delta{m})\omega_L$.
159:
160: Throughout the duration of the VSTPR pulse, resonant atoms oscillate
161: between the two momentum states separated by $2{\hbar}\mathbf{k}$.
162: Because the atoms are initially confined in a point trap, an image
163: of the atom cloud after expansion is a spatial map of the average
164: velocity distribution, which has been perturbed by the VSTPR pulse.
165: A freely expanding cloud has approximately a smooth Gaussian
166: velocity spectrum along $x$ (the VSTPR beam direction), but the
167: momentum oscillations that occur for the resonant atoms alter this
168: average velocity distribution. Images taken along a camera direction
169: orthogonal to $x$ record these narrow perturbations.
170:
171: \section{\label{sec:level3}Experiment Setup}
172:
173: \begin{figure}
174: \includegraphics{fig2.jpg}
175: \caption{\label{fig:chamber} Experimental setup for observing VSTPR
176: in a magnetic field.}
177: \end{figure}
178:
179: The layout of our apparatus is shown in Fig.~\ref{fig:chamber}. The
180: experiment begins with a vapor cell MOT containing 10$^7$ $^{85}$Rb
181: atoms. The MOT diameter is $\approx$500~$\mu$m and the temperature
182: is $\approx$200~$\mu$K. The cooling beams (detuned -18 MHz from the
183: $F$=3 - $F'$=4 transition) are derived from an extended cavity diode
184: laser (New Focus Vortex, model 6013) that seeds a 120 mW laser diode
185: (Sharp GH0781JA2C). 60 mW is directed to one port of a 2 x 3
186: polarization maintaining (PM) fiber splitter. Each of the three
187: output fibers carries 12 mW. The outputs are collimated using 100-mm
188: focal length, 50mm-diameter achromats, giving 1/e$^2$ beam diameters
189: of 24 mm. The three beams propagate along orthogonal directions and
190: are retroreflected; one pair is vertical and the other two are in a
191: horizontal ($x$-$y$) plane. The repump light, connecting $F=2
192: \rightarrow F'=3$, is derived from an independent Vortex laser that
193: also seeds a diode. In normal MOT operation, 15 mW of repump light
194: is coupled into the other port of the 2x3 coupler.
195:
196: Our VSTPR beam is spatially filtered by PM fiber, and is collimated
197: by a 60mm gradient-index singlet lens ($1/e^2$ beam waist $\omega_0$
198: = 7.5 mm). We use up to 20 mW laser power. It is retroreflected in
199: a lin $\perp$ lin configuration. For B-field control, we use three
200: orthogonal pairs of Helmholtz coils. The magnetic field at the atom
201: cloud has components $B_i = \alpha_i(I_i - I_{0i})$ where $\alpha_i$
202: are the slopes $dB_i/dI_i$, $I_i$ are the applied currents, and
203: $I_{0i}$ are the currents required for compensation along each
204: Cartesian direction. The VSTPR beam travels horizontally along the
205: axis of the $x$-directed coil pair. For all results in this paper,
206: we have used a second beam derived from the repump laser as our
207: VSTPR beam ($\Delta=3$GHz).
208:
209: At time T = 0, the atoms are released from the MOT by extinguishing
210: all laser beams and the MOT coils. The bias magnetic coils remain
211: on. We do not perform any molasses cooling, because the large
212: velocity spread of the hotter sample of atoms provides greater range
213: over which velocity selection can occur. At time $T_r \approx 15$
214: ms, the VSTPR pulse is switched on for 5 ms and at $T_i = 40$ ms,
215: the MOT cooling and repump beams are switched on to image the
216: expanded cloud onto the CCD camera.
217:
218: \section{\label{sec:level4}Results}
219:
220: \begin{figure}
221: \includegraphics[width=8.4cm]{fig3.jpg}% Here is how to import EPS art
222: \caption{\label{fig:stripe_image} a) Left: Raw images of the
223: expanded atom cloud at different Helmholtz currents after exposure
224: to the VSTPR beams. Right: Corresponding cross-sectional profiles
225: after background subtraction. Fits are shown as solid lines. b)
226: Larmor precession frequency extracted from the stripe separation as
227: a function of current in the $z$-directed coils for a nonzero $B_y$
228: field (circles) and for $B_x=B_y=0$ (diamonds). Fits shown as solid
229: lines.}
230: \end{figure}
231:
232: %We tune the magnetic field by changing the current in the
233: %$z$-directed Helmholtz bias coils.
234: We control the magnetic field by changing the current in the
235: Helmholtz bias coils. We first demonstrate the effect by changing
236: the current in the $z$-directed Helmholtz coils.
237: Fig.~\ref{fig:stripe_image}a shows typical images recorded by the
238: CCD camera for four different current settings. The VSTPR pulse
239: creates perturbations that appear as vertical stripes in the
240: expanded cloud. Also shown in Fig.~\ref{fig:stripe_image}a are the
241: cross sections after subtracting the images obtained with no VSTPR
242: pulse. The spatial location of each peak corresponds to the average
243: velocity class satisfying $2{\mathbf{k}\cdot}\mathbf{v_0} =
244: \pm(\Delta{m})\omega_L$ that participated in the VSTPR process. In
245: general, $\omega_L$ is a function of position due to spatially
246: varying magnetic fields. These gradients could tip, bend, or blur
247: the resonant stripe features. Under normal operating conditions,
248: however, we do not observe these effects. With the B-field
249: primarily along the $z$-axis and using lin~$\perp$~lin
250: polarizations, the VSTPR pulse mainly connects $\Delta{m} = 1$
251: transitions. The cross-sectional profile of each vertical stripe can
252: be estimated by several functional forms. For simplicity, we fit
253: each peak to two Gaussian profiles such that the total area is 0,
254: but we describe a more exact fitting function below.
255:
256: In addition to the peaks corresponding to velocity classes at
257: $\pm{v_0}$, there is a peak at v = 0 due to $\Delta{m}=0$
258: transitions that arises from the longitudinal magnetic field
259: component. This is a useful marker for balancing the overall
260: velocity distribution. We rarely observe features corresponding to
261: ${\Delta}m = {\pm}2$ transitions. Detailed calculations of the
262: relative transition strengths will be published elsewhere, but in
263: general, the strengths of the ${\Delta}m = {\pm}2$ transitions are
264: roughly two orders of magnitude smaller than those for ${\Delta}m =
265: 0, {\pm}1$ for $\Delta = 3$~GHz. It is important to note that the
266: appearance of narrow features in the expanded cloud is not the
267: result of cooling, because there is no dissipative force. For
268: smaller $\Delta$, on the order of the hyperfine splittings of the D2
269: manifold, spontaneous scattering events can lead to
270: magnetically-induced laser cooling~\cite{Metcalf}.
271:
272: The $z$-directed Helmholtz bias coils are each made of 24 gauge wire
273: wound on an 8" vacuum flange, 1" long, and centered 5.8 cm from the
274: MOT, producing a field of approximately 1.5 G/A. Our fits to
275: Fig.~\ref{fig:stripe_image} show that after 35 ms of falling time,
276: the separations of the stripes for the 4 current settings 260 mA,
277: 330 mA, 400 mA, and 450 mA are 3182(4) $\mu$m, 4538(4), $\mu$m,
278: 5902(4) $\mu$m, and 6873(4) $\mu$m. The listed errors are
279: statistical. Our leading systematic source of multiplicative error
280: is the pixel calibration on the camera of $\approx$0.2\%. In
281: section~\ref{sec:level6}, we briefly describe a second calibration
282: procedure that removes pixel calibration errors. The 4 $\mu$m error
283: in stripe separation corresponds to an error of ~300 $\mu$G. Note
284: that the error in magnetic field will be explicitly dependent on the
285: atom species through $g_F$ and the VSTPR wavelength.
286:
287: The scalar magnetic field measurements for several current settings
288: are shown in Fig.~\ref{fig:stripe_image}b. Because the stripe
289: separation is proportional to ${(B_x^2 + B_y^2 + B_z^2)^{1/2}}$, a
290: plot of $\omega_L$ versus the current in the $z$-directed bias coil
291: traces a hyperbola, the minimum of which determines the field
292: component perpendicular to $z$ and the compensation current
293: $I_{0z}$. We show two cases, one with nonzero $B_y$, and one for the
294: case in which $B_x, B_y$ have been zeroed. From a fit to these
295: plots, we extract $\alpha_z = 1.524(2)$ G/A and $I_{0z}$ = 243.1(2)
296: mA, which corresponds to a compensation level of 300~$\mu$G. In
297: practice, it is simple to zero the magnetic field by viewing
298: real-time images of the expanded cloud and adjusting the currents
299: along each axis for minimum stripe separation. We note that when
300: the total magnetic field is close to zero, the stripes begin to
301: overlap and are no longer resolved. Compensation is achieved when
302: the overlap is maximized, resulting in a single narrow feature. In
303: our experience, this real-time adjustment of the stripe separation
304: results in compensation to milliGauss levels without any data
305: analysis.
306:
307: The visibility of the stripe features is dependent on a few factors.
308: First, since the image on the CCD camera is a convolution of the
309: initial MOT size with the velocity distribution, the contrast
310: increases for trapped samples with smaller physical dimensions.
311: Optimally, the imaging should be performed after the cloud has
312: expanded enough that two velocity classes separated by
313: $2{\hbar}\mathbf{k}/M$ can be resolved. If the initial MOT has a
314: radius R, this means that the imaging should be performed after a
315: time $R/v_{rec}$ from the release of the atoms from the trap. In
316: practice, the features are easily observed with imaging times
317: significantly less because the effect does not require that the
318: recoil velocities be resolved, only that perturbations to the
319: average velocity distribution can be observed.
320:
321: \begin{figure}
322: \includegraphics[width=8.4cm]{fig4.jpg}% Here is how to import EPS art
323: \caption{ a) Importance of timing of the VSTPR pulse. An atom
324: moving with ${\hbar}k$ has its momentum reversed at $T_r$. If this
325: pulse occurs at $T_i/2$, atoms initially having ${\pm}{\hbar}k$ are
326: overlapped at $T_i$. b) If the pulse arrives later or earlier,
327: contrast is reduced. c) Pictures of a single stripe using different
328: $T_r$, showing optimum contrast at $T_i/2$. d) Stripe cross-sections
329: for $T_r=0.5T_i$ and $T_r=0.25T_i$, along with fits using
330: Equation~\ref{eqn:Gaussian} (solid line).}
331: \label{fig:TimingSequence}
332: \end{figure}
333:
334: Other parameters that control the visibility of the stripes are the
335: duration and timing of the VSTPR pulse. Because this measurement is
336: time-averaged over the duration of the pulse, shorter pulses reduce
337: blurring effects due to time-varying fields. Furthermore, if
338: applied at $T_r = T_i/2$ they can maximize stripe contrast.
339: Figure~\ref{fig:TimingSequence}a shows the effect on stripe contrast
340: as a function of $T_r$ for atoms in zero field. An atom initially at
341: the origin will return to the origin if its momentum is reversed by
342: a $\pi$-pulse at $T_i/2$. For $\Delta{T}=T_r-T_i/2\neq0$, the atoms
343: are still deflected, but the atoms with initial momenta of
344: ${\pm}{\hbar}k$ no longer spatially overlap at $T_i$
345: (Fig.~\ref{fig:TimingSequence}b). In Fig.~\ref{fig:TimingSequence}c,
346: we show this effect experimentally using VSTPR pulse durations of
347: 200 ${\mu}$sec at different $T_r$. This simple geometrical picture
348: suggests an approximate functional form for the
349: background-subtracted stripe cross section under these conditions:
350:
351: \begin{eqnarray}
352: Y(x) & = & G(x_0-2v_r\Delta{T}) + G(x_0+2v_r\Delta{T}) \nonumber\\
353: & - & G(x_0-v_rT_i) - G(x_0+v_rT_i)
354: \label{eqn:Gaussian}
355: \end{eqnarray}
356:
357: \noindent where $G(x_0)$ is a Gaussian centered at $x_0$. In
358: Fig.~\ref{fig:TimingSequence}d, we show fits when $T_r=0.5T_i$ and
359: $T_r=0.25T_i$. For all other data presented in this manuscript, we
360: used pulse durations of 5 msec. Although AC magnetic fields were
361: not compensated, these longer pulses showed no measurable broadening
362: in our experiment. The $\pi$-pulse duration depends on the
363: particular magnetic sublevels involved, so for our spin-unpolarized
364: sample we generally choose a pulse duration that provides
365: consistently strong signals over a broad range of Raman pulse
366: intensities. For a typical VSTPR beam intensity of 10~mW/cm$^2$,
367: the resonant 2-photon Rabi frequency is ${\approx}2\pi\times$10~kHz.
368:
369: \section{\label{sec:level5}Comparison with Faraday Spectroscopy}
370:
371: \begin{figure}
372: \includegraphics[width=8.4cm]{fig5.jpg}% Here is how to import EPS art
373: \caption{\label{fig:faraday} a) Schematic of Faraday spectroscopy
374: setup. WP: Wollaston prism. b) Typical Faraday signal. c)
375: Comparison of magnetic field measurement with VSTPR (circles) and
376: with Faraday spectroscopy (squares) for similar conditions as
377: Fig.~\ref{fig:stripe_image}b. Solid line is a fit to the Faraday
378: data. Dotted line is a fit to the VSTPR data. Dashed line indicates
379: the minimum of the hyperbola as measured by Faraday spectroscopy.}
380: \end{figure}
381:
382: We have verified this VSTPR technique by using Faraday rotation
383: spectroscopy to measure the magnetic field~\cite{Isayama, Jessen,
384: Kaiser}. To perform these measurements, an additional pair of laser
385: beams is used along the $x$-axis (Fig.~\ref{fig:faraday}a). The
386: atoms are optically pumped into the $F=3, m_F = 3$ stretched state
387: (in the $x$-basis) by a 100 $\mu$s $\sigma^+$ pulse connecting $F=3
388: \rightarrow F' = 3$. This beam contains a small amount of repumper
389: light to keep the atoms in F=3. When this light is extinguished, the
390: atoms begin precessing freely. A linearly polarized probe beam with
391: $\simeq$100 $\mu$W and waist $\omega_0 = 500{\mu}$m passes through
392: the atom cloud to a simple polarimeter consisting of a Wollaston
393: prism that splits the probe beam into two orthogonal polarization
394: states that are detected by a balanced photodetector \cite{Hobbs}.
395: We make these measurements at the same time delay as the VSTPR
396: measurements. A typical Faraday signal is shown in Fig.
397: \ref{fig:faraday}b. The comparison of the VSTPR and Faraday
398: techniques is shown in Fig.~\ref{fig:faraday}c. Because we cannot
399: compare Faraday and stripe measurements near zero field where the
400: stripes are unresolved, we show in Fig.~\ref{fig:faraday}c the
401: stripe data with and without a transverse field along $y$. This
402: transverse field allows a stripe measurement at $I_{0z}$, which is
403: in good agreement with the Faraday measurement (dashed line in
404: Fig.~\ref{fig:faraday}c). From the Faraday measurements with no
405: transverse field, also shown in this figure, we derive $\alpha_z =
406: 1.547(3)$ G/A and $I_{0z} = 242.5(1)$ mA, both of which agree well
407: with the VSTPR technique.
408:
409: \section{\label{sec:level6}Calibration}
410:
411: To obtain correct values of the magnetic field, the stripe
412: separation must be carefully measured. Without calibration, the
413: technique is still useful for determining the $I_{0i}$ of each coil
414: by simply finding the minimum stripe separation, which is
415: independent of this source of systematic error. Because this is an
416: imaging technique, good estimates of spatial calibration can be made
417: simply by measuring the magnification on the CCD camera if
418: high-quality imaging lenses are used. Most lenses exhibit some
419: degree of aberrations that make accurate pixel calibration difficult
420: below the 0.5\% level. Even without this error, other slight
421: systematic errors, such as the exact functional form used to fit the
422: stripe cross section may remain. In this section, we describe a
423: technique for calibrating the Zeeman shifts indicated by the stripes
424: in a more direct manner that eliminates these systematic errors.
425:
426: \begin{figure}
427: \includegraphics[width=8.4cm]{fig6.jpg}% Here is how to import EPS art
428: \caption{\label{fig:calibration} a) Image of atom cloud (after
429: background subtraction) taken with several sidebands separated by
430: 100 kHz imposed on one VSTPR beam. b) Cross sectional profile.}
431: \end{figure}
432:
433: Instead of relying on an accurate pixel calibration, the splittings
434: of the stripes can be determined by making $\omega_1 = \omega_2 +
435: \delta_{12}$, where $\delta_{12}$ is a frequency shift imposed by an
436: RF source. The resonant velocity classes are now determined by
437: $2\mathbf{k}{\cdot}\mathbf{v} = ({\Delta}m)\omega_L + \delta_{12}$.
438: To demonstrate this idea, the retroreflecting mirror in
439: Fig.~\ref{fig:chamber} is replaced by a counterpropagating beam of
440: the same diameter and power. This counterpropagating beam is
441: derived from the original, so it is phase locked to $\omega_1$, but
442: its frequency is shifted by two acousto-optic modulators (AOM) to
443: achieve small $\delta_{12}~(<$1~MHz$)$. Additionally, we frequency
444: modulate one AOM so that its instantaneous frequency is
445: $\omega_{RF}$ + Asin($\omega_{m}t$), where $\omega_{RF}$ is the
446: drive frequency, $\omega_{m}$ is the modulation frequency, and A is
447: the maximum frequency deviation. This imparts frequency sidebands at
448: $n\omega_{m}$, where $n$ is an integer, so that multiple
449: $\delta_{12}$ are produced simultaneously. In
450: Fig.~\ref{fig:calibration}, we show an image and cross section taken
451: with $\omega_{m} = 2{\pi}\times100$~kHz. For $\delta_{12}=0$, the
452: range of measurable Zeeman shifts is limited to the Doppler width of
453: the atom cloud ($\simeq$1G). A nonzero $\delta_{12}$ overcomes this
454: limitation by shifting the stripe to an accessible velocity class.
455:
456: \section{\label{sec:level7}Conclusion}
457:
458: We have used velocity-selective resonances between magnetic
459: sublevels of a single hyperfine level in $^{85}$Rb to measure
460: magnetic fields in a cold atom cloud. The resonances are easily
461: observed with no additional laser frequencies than are required for
462: MOTs, and can be used to measure magnetic fields with sub-mG
463: resolution. Because of its simplicity, this technique should prove
464: especially useful for aiding magnetic field compensation, for which
465: purpose no calibration is required.
466:
467: This work was funded by the Defense Advanced Research Projects
468: Agency and the Office of Naval Research.
469:
470: %\bibliography{VSR_terraciano}% Produces the bibliography via BibTeX.
471:
472: \begin{thebibliography}{12}
473: \expandafter\ifx\csname
474: natexlab\endcsname\relax\def\natexlab#1{#1}\fi
475: \expandafter\ifx\csname bibnamefont\endcsname\relax
476: \def\bibnamefont#1{#1}\fi
477: \expandafter\ifx\csname bibfnamefont\endcsname\relax
478: \def\bibfnamefont#1{#1}\fi
479: \expandafter\ifx\csname citenamefont\endcsname\relax
480: \def\citenamefont#1{#1}\fi
481: \expandafter\ifx\csname url\endcsname\relax
482: \def\url#1{\texttt{#1}}\fi
483: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL
484: }\fi \providecommand{\bibinfo}[2]{#2}
485: \providecommand{\eprint}[2][]{\url{#2}}
486:
487: \bibitem[{\citenamefont{Kasevich et~al.}(1991)\citenamefont{Kasevich, Weiss,
488: Riis, Moler, Kasapi, and Chu}}]{Kasevich}
489: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kasevich}},
490: \bibinfo{author}{\bibfnamefont{D.~S.} \bibnamefont{Weiss}},
491: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Riis}},
492: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Moler}},
493: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Kasapi}}, \bibnamefont{and}
494: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Chu}},
495: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{66}},
496: \bibinfo{pages}{2297} (\bibinfo{year}{1991}).
497:
498: \bibitem[{\citenamefont{Boyer et~al.}(2004)\citenamefont{Boyer, Lising,
499: Rolston, and Phillips}}]{boyer:043405}
500: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Boyer}},
501: \bibinfo{author}{\bibfnamefont{L.~J.} \bibnamefont{Lising}},
502: \bibinfo{author}{\bibfnamefont{S.~L.} \bibnamefont{Rolston}},
503: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{W.~D.}
504: \bibnamefont{Phillips}}, \bibinfo{journal}{Phys. Rev. A}
505: \textbf{\bibinfo{volume}{70}}, \bibinfo{eid}{043405} (\bibinfo{year}{2004}).
506:
507: \bibitem[{\citenamefont{McGuirk et~al.}(2002)\citenamefont{McGuirk, Foster,
508: Fixler, Snadden, and Kasevich}}]{McQuirk}
509: \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{McGuirk}},
510: \bibinfo{author}{\bibfnamefont{G.~T.} \bibnamefont{Foster}},
511: \bibinfo{author}{\bibfnamefont{J.~B.} \bibnamefont{Fixler}},
512: \bibinfo{author}{\bibfnamefont{M.~J.} \bibnamefont{Snadden}},
513: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{M.~A.}
514: \bibnamefont{Kasevich}}, \bibinfo{journal}{Phys. Rev. A}
515: \textbf{\bibinfo{volume}{65}}, \bibinfo{pages}{033608}
516: (\bibinfo{year}{2002}).
517:
518: \bibitem[{\citenamefont{Chab\'{e} et~al.}(2007)\citenamefont{Chab\'{e},
519: Lignier, Szriftgiser, and Garreau}}]{Chabe}
520: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Chab\'{e}}},
521: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Lignier}},
522: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Szriftgiser}},
523: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~C.}
524: \bibnamefont{Garreau}}, \bibinfo{journal}{Opt. Commun.}
525: \textbf{\bibinfo{volume}{274}}, \bibinfo{pages}{254} (\bibinfo{year}{2007}).
526:
527: \bibitem[{\citenamefont{Ringot et~al.}(2001)\citenamefont{Ringot, Szriftgiser,
528: and Garreau}}]{Garreau}
529: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Ringot}},
530: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Szriftgiser}},
531: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~C.}
532: \bibnamefont{Garreau}}, \bibinfo{journal}{Phys. Rev. A}
533: \textbf{\bibinfo{volume}{65}}, \bibinfo{pages}{013403}
534: (\bibinfo{year}{2001}).
535:
536: \bibitem[{\citenamefont{Vuleti\ifmmode~\acute{c}\else \'{c}\fi{}
537: et~al.}(1998)\citenamefont{Vuleti\ifmmode~\acute{c}\else \'{c}\fi{}, Chin,
538: Kerman, and Chu}}]{Vuletic}
539: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Vuleti\ifmmode~\acute{c}\else
540: \'{c}\fi{}}}, \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Chin}},
541: \bibinfo{author}{\bibfnamefont{A.~J.} \bibnamefont{Kerman}},
542: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Chu}},
543: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{81}},
544: \bibinfo{pages}{5768} (\bibinfo{year}{1998}).
545:
546: \bibitem[{\citenamefont{Isayama et~al.}(1999)\citenamefont{Isayama, Takahashi,
547: Tanaka, Toyoda, Ishikawa, and Yabuzaki}}]{Isayama}
548: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Isayama}},
549: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Takahashi}},
550: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Tanaka}},
551: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Toyoda}},
552: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ishikawa}}, \bibnamefont{and}
553: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Yabuzaki}},
554: \bibinfo{journal}{Phys. Rev. A} \textbf{\bibinfo{volume}{59}},
555: \bibinfo{pages}{4836} (\bibinfo{year}{1999}).
556:
557: \bibitem[{\citenamefont{Smith et~al.}(2003)\citenamefont{Smith, Chaudhury, and
558: Jessen}}]{Jessen}
559: \bibinfo{author}{\bibfnamefont{G.~A.} \bibnamefont{Smith}},
560: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Chaudhury}},
561: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.~S.}
562: \bibnamefont{Jessen}}, \bibinfo{journal}{Journal of Optics B-Quantum and
563: Semiclassical Optics} \textbf{\bibinfo{volume}{5}}, \bibinfo{pages}{323}
564: (\bibinfo{year}{2003}).
565:
566: \bibitem[{\citenamefont{Labeyrie et~al.}(2001)\citenamefont{Labeyrie,
567: Miniatura, and Kaiser}}]{Kaiser}
568: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Labeyrie}},
569: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Miniatura}},
570: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Kaiser}},
571: \bibinfo{journal}{Phys. Rev. A} \textbf{\bibinfo{volume}{64}},
572: \bibinfo{pages}{033402} (\bibinfo{year}{2001}).
573:
574: \bibitem[{\citenamefont{Alexandrov et~al.}(2004)\citenamefont{Alexandrov,
575: Balabas, Vershovski, and Pazgalev}}]{gF}
576: \bibinfo{author}{\bibfnamefont{E.~B.} \bibnamefont{Alexandrov}},
577: \bibinfo{author}{\bibfnamefont{M.~V.} \bibnamefont{Balabas}},
578: \bibinfo{author}{\bibfnamefont{A.~K.} \bibnamefont{Vershovski}},
579: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.~S.}
580: \bibnamefont{Pazgalev}}, \bibinfo{journal}{Technical Physics}
581: \textbf{\bibinfo{volume}{49}}, \bibinfo{pages}{779} (\bibinfo{year}{2004}).
582:
583: \bibitem[{\citenamefont{Sheehy et~al.}(1990)\citenamefont{Sheehy, Shang,
584: van~der Straten, Hatamian, and Metcalf}}]{Metcalf}
585: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Sheehy}},
586: \bibinfo{author}{\bibfnamefont{S.-Q.} \bibnamefont{Shang}},
587: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{van~der Straten}},
588: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Hatamian}}, \bibnamefont{and}
589: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Metcalf}},
590: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{64}},
591: \bibinfo{pages}{858} (\bibinfo{year}{1990}).
592:
593: \bibitem[{\citenamefont{Hobbs}(1997)}]{Hobbs}
594: \bibinfo{author}{\bibfnamefont{P.~C.~D.} \bibnamefont{Hobbs}},
595: \bibinfo{journal}{Appl. Opt.} \textbf{\bibinfo{volume}{36}},
596: \bibinfo{pages}{903} (\bibinfo{year}{1997}).
597:
598: \end{thebibliography}
599:
600:
601: \end{document}
602: