et al.
et al.
et al.
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
REPORTS<br />
802<br />
on solid objects have remained elusive. Mechanic<strong>al</strong><br />
effects of photon recoil are routinely studied<br />
in atomic physics [(9) and references therein],<br />
and a RPSN observation an<strong>al</strong>ogous to ours has<br />
been made using a dilute gas of ultracold atoms<br />
(10). A promising route to studying RPSN in solid<br />
objects involves experiments that achieve high<br />
optomechanic<strong>al</strong> coupling to high-frequency, sm<strong>al</strong>l<br />
(nanom<strong>et</strong>er- to centim<strong>et</strong>er-sc<strong>al</strong>e) mechanic<strong>al</strong> resonators.<br />
Using such resonators, groups have initiated<br />
searches for RPSN (11, 12), observed classic<strong>al</strong><br />
an<strong>al</strong>ogs of RPSN (13), and predicted experiment<strong>al</strong><br />
signatures of RPSN (14–16). Back-action<br />
on a nanomechanic<strong>al</strong> resonator has <strong>al</strong>so been observed<br />
with the use of other measurement devices,<br />
such as single-electron transistors (17). Resonators<br />
have even been cooled with electromagn<strong>et</strong>ic<br />
radiation to near their motion<strong>al</strong> ground state, illustrating<br />
the capacity for dominant coherent optic<strong>al</strong><br />
forces (18–20). In these experiments, quantum<br />
back-action has thus far been limited to the sc<strong>al</strong>e<br />
of Zzp, whereas in this Report, we demonstrate<br />
a strong back-action heating effect from RPSN.<br />
Addition<strong>al</strong>ly, in near–ground-state cooling experiments,<br />
correlations b<strong>et</strong>ween shot noise and<br />
RPSN-driven mechanic<strong>al</strong> motion are an important<br />
component of the observed optic<strong>al</strong> spectra<br />
(21) and are responsible for the sideband asymm<strong>et</strong>ry<br />
observed in (22).<br />
Our optomechanic<strong>al</strong> system consists of a silicon<br />
nitride membrane resonator inside of a Fabry-<br />
Perot optic<strong>al</strong> cavity that is speci<strong>al</strong>ly designed to<br />
operate at cryogenic temperatures (Fig. 1C) (23).<br />
Thompson <strong>et</strong> <strong>al</strong>. have shown that membrane motion<br />
can be coupled to a cavity through a dispersive<br />
interaction, where the cavity resonance<br />
frequency shifts as the membrane moves <strong>al</strong>ong<br />
the optic<strong>al</strong> standing wave (24). This interaction<br />
imprints phase and amplitude modulation on<br />
transmitted laser light, <strong>al</strong>lowing for readout of the<br />
membrane motion. In conjunction, the laser applies<br />
an optic<strong>al</strong> gradient force to the membrane,<br />
pushing it toward higher optic<strong>al</strong> intensity. Our<br />
membrane is a highly tensioned square plate with<br />
a 0.5-mm side length, 40-nm thickness, and an<br />
effective mass of ~7 ng. We operate in a helium<br />
flow cryostat with the resonator at a base temperature<br />
of 4.9 K, where intrinsic mechanic<strong>al</strong> linewidths,<br />
G0/2p, are typic<strong>al</strong>ly less than 1 Hz. For<br />
the (2,2) mode oscillating at wm/2p =1.55MHz,<br />
we achieve a maximum single-photon optomechanic<strong>al</strong><br />
coupling rate g/2p =16Hz.<br />
We use two laser beams derived from the<br />
same 1064-nm source, both coupled to the same<br />
spati<strong>al</strong> mode of the cavity, but with orthogon<strong>al</strong><br />
polarizations (Fig. 1C) (13, 14). The h<strong>al</strong>f-planar,<br />
5.1-mm-long cavity has a full linewidth k/2p ~<br />
1 MHz, which varies slightly with the membrane<br />
position. The high-intensity “sign<strong>al</strong>” beam is actively<br />
stabilized to the optic<strong>al</strong> resonance. This<br />
beam provides the RPSN, and its transmitted intensity<br />
fluctuations constitute a record, which is<br />
parti<strong>al</strong>ly obscured by optic<strong>al</strong> loss, of the optic<strong>al</strong><br />
force on the resonator. The corresponding sensitive<br />
position measurement is wholly imprinted<br />
in the unrecorded phase quadrature. Addition<strong>al</strong> phase<br />
noise from fluctuations in the cavity-laser d<strong>et</strong>uning<br />
precludes shot noise–limited phase-quadrature<br />
d<strong>et</strong>ection (23). The much weaker “m<strong>et</strong>er” beam<br />
is tuned to the red of the optic<strong>al</strong> resonance im-<br />
printing the resonator’s displacement spectrum on<br />
its transmitted intensity. Although its shot noise<br />
drive is much sm<strong>al</strong>ler, the m<strong>et</strong>er beam provides<br />
optic<strong>al</strong> Raman sideband cooling of the mechanic<strong>al</strong><br />
mode (25) to 1.7 mK. The optic<strong>al</strong> damping<br />
Fig. 1. (A) Canonic<strong>al</strong> picture of continuous position measurement. RPSN (black), therm<strong>al</strong> motion<br />
(brown), and zero point motion (orange) combine to give the expected measurement result (blue). The<br />
dashed line represents the effective displacement noise from the shot noise–limited imprecision of an<br />
optic<strong>al</strong>measurement.(B) Photocurrent spectra. The photocurrent spectr<strong>al</strong> densities S IS (w)/I S 2 (blue) and<br />
S IM (w)/I M 2 (red), as well as the noise floors, including d<strong>et</strong>ector noise and the dominant shot noise (gray), are<br />
shown. (C) Experiment<strong>al</strong> s<strong>et</strong>up. Beams are combined and separated with polarizing beam splitters (PBS)<br />
and d<strong>et</strong>ected directly on photod<strong>et</strong>ectors (PD). The ins<strong>et</strong> photograph shows an in situ image of the square<br />
membrane and optic<strong>al</strong> mode spot, with blue dashed lines indicating the nodes of the (2,2) mechanic<strong>al</strong><br />
mode. The ins<strong>et</strong> diagram at right shows laser-cavity d<strong>et</strong>unings.<br />
Fig. 2. Displacement spectrum<br />
measurements. Measured<br />
peak displacement<br />
spectr<strong>al</strong> density (circles), therm<strong>al</strong><br />
contribution (brown<br />
line), and expected RPSN<br />
contribution (black line)<br />
are shown. The blue curve<br />
represents the theor<strong>et</strong>ic<strong>al</strong><br />
prediction for the sum of<br />
therm<strong>al</strong> motion and RPSN,<br />
and the dashed red curves<br />
are bounds on theor<strong>et</strong>ic<strong>al</strong><br />
estimates, including systematic<br />
uncertainty in device<br />
param<strong>et</strong>ers and the<br />
classic<strong>al</strong> noise contribution.<br />
Device param<strong>et</strong>ers: g/2p =<br />
16.1 T 0.3 Hz, k/2p =<br />
0.89 MHz, ∆S/2p =2.0T<br />
0.5 kHz, ∆M/2p =0.7MHz,<br />
N M =7.0T 0.3 × 10 6 ,<br />
wm/2p = 1.551 MHz,<br />
G0/2p =0.47Hz,and<br />
Gm/2p =1.43kHz.(Ins<strong>et</strong>)<br />
Transmission spectra for R S =0.056(blue)andR S = 1.0 (orange), with corresponding points in the main<br />
plot highlighted in blue and orange.<br />
15 FEBRUARY 2013 VOL 339 SCIENCE www.sciencemag.org<br />
on February 14, 2013<br />
www.sciencemag.org<br />
Downloaded from