320 likes | 341 Views
This research focuses on using lasers and optical cavities to search for new particles by shining light through walls. The experiment involves resonant enhancement and challenges in the design and implementation. It aims to reach frontier limits in the search for hidden sector photons and axions with high sensitivity. The REAPR design incorporates Fabry-Perot cavities, magnet detectors, and laser IO for resonantly enhanced photon regeneration. The experiment requires careful alignment, control of stray light, locking of the regeneration cavity, and thermal compensation. The project involves collaborations with various research institutions.
E N D
D.B. Tanner University of Florida With Guido Mueller, Pierre Sikivie, Karl van Bibber, Aaron Chou, Jason Steffan, William Wester and others Lasers and Optical Cavities to Search for New Particles • Shining light through walls • Resonant enhancement • Challenges to the experiment
REAPR • Resonantly Enhanced Alps Photon Regeneration • Alps == Any light particle, e.g., an axion, or hidden-sector photon, or … • Purely laboratory experiment. • Hidden sector photon search with two cavities; • axions with two cavities and dipole magnets. • Uses state-of-the-art (but available) technology. • Reaches gagg ~ 2 x 10-11 GeV-1 in the axion search. • Frontier: limit as I→ 0.
a x x Shining light through the wall rate GammeV, BFT, BMV, ALPS, LIPPS, OSQAR
Magnet Magnet Photon Detectors Matched Fabry-Perots Laser IO Resonantly-Enhanced Photon Regeneration Basic concept – use Fabry-Perot cavities in both magnets. whereF, F ’ are the finesses of the cavities. (Both can be 105) Hoogeveen and Ziegenhagen (1991); Sikivie, DT, and van Bibber (2007), Mueller et al (2009)
REAPR design • 1064 nm lasers • Generation cavity on left, regeneration on right. • Use heterodyne detection of light in regeneration cavity. • Central bench and beam block can be removed for alignment, testing
Cavities must be aligned on mirror image modes (as if inner mirrors and wall were not present). Stray light must be attenuated by huge factor. Photon regeneration cavity must be locked to production cavity without filling it with light at the laser wavelength. High power => heating => mirror deformation => modal changes. Need sensitive readout of weak regenerated signal. Finesse of 105. Production lasermust be locked to axion generation cavity. Magnet Magnet Detectors Laser IO Matched Fabry-Perots Challenges (ranked hard to easy)
1. Cavities must be aligned… • Start with a single long cavity consisting of the two end mirrors, CM1 and CM2 • Central bench is removed for this step • Fix CM1 and CM2 for desired alignment, as determined by quad photodiodes • Now put central bench in place and align FM1 and FM1 • Repeat as necessary.
2. Stray light control • Wall must be opaque! • Cannot use ULE glass for bench; metal instead. • Baffle all ghost beams. • LIGO has had to address this seriously, because stray light introduces phase noise in photodetector. • With no line of sight, stray light must involve reflections, and we can vibrate components and look for a signal in detection bench from this. • Use fiber to bring some laser light to detection cavity, for tests.
3. Locking regeneration cavity • Use two lasers. • Laser 1 injects power into generation cavity • Laser 2 is offset locked to Laser 1 • Offset frequency W = integer * FSR of the cavities • Regeneration cavity is PDH locked to Laser 2 • Laser 2 used to for heterodyne readout of signal in regeneration cavity
4. Heating 1 MW incident, 1 ppm loss, absorbed P = 1 W Heating, deformation, loss of mode • Thermal compensation • Add heaters to perimeter of mirror • Heat reduces thermal gradients • Used successfully in LIGO
Use Laser 2 as an LO for regenerated photons at the frequency of Laser 1. Signal is (in terms of the number of photons in each field) f is an arbitrary phase and W is the laser frequency difference Noise is shot noise: Phase is arbitrary and unknown, so add I and Q in quadrature; the shot-noise limited SNR is i.e, one photon at an SNR of 1. 5. Readout scheme
6. Finesse • Finesse ~ number of bounces a ray makes. • withT1the transmission of the input mirror and V the loss. • At 1064 nm, T ~ 5.5 ppm • Finesse ~ 1 million (ignoring scattering, which you cannot do) • At UF we’ve made many cavities with finesses > 10,000 and locked them to lasers.
7. Locking the cavities Ec ESB wm • Pound-Drever-Hall locking • Done all the time
The REAPR Collaboration Univ. of Florida: David Tanner (LIGO, ADMX) Guido Muller (LIGO, Chair of LISA Interferometer working group) Pierre Sikivie (ADMX, axion physics) FNAL: Aaron Chou (Wilson Fellow, co-spokesperson GammeV), William Wester (co-spokesperson GammeV), Jason Steffen (Brinson Fellow) Peter Mazur Ray Tomlin Al Baumbaugh Jim Volk Naval Postgraduate School: Karl van Bibber (co-spokesperson ADMX) Univ. of Michigan: Dick Gustafson (LIGO)
Karl van Bibber’s “EE” argument The gain on the production side is simple: • The number of forward passes the light makes in the magnet is larger by a factor of F/p • Or, the cavity gain in power is F/p • The axion flux is larger by a factor of F/p
On the regeneration side, 1 pass through the magnet produces: P1 = E12 In the cavity, the light approaching a mirror is Pc = Ec2 After 1 round trip this partial ray has intensity Prt = R2*Ec2 This adds in phase to the regenerated wave E1 (add amplitudes!) Ec = R*Ec + E1 (1 - R)*Ec = E1 Ec = E1/T Pc = P1/T2 This light is transmitted through the mirror to the detector Pdet = P1/T ~ F*P1/p Karl van Bibber’s “EE” argument Take R + T = 1 for both mirrors
The axion field converts in the regeneration cavity to a signal field ES at Laser 1 frequency ω0. Mix this with Laser 2 (the LO) at a photodiode; the signal is proportional to the intensity Write this in terms of the number of photons in each field 5. Readout scheme
Noise is shot noise: Phase is arbitrary and unknown, so add I and Q in quadrature Shot-noise limited SNR is i.e, one photon at an SNR of 1. 5. Readout scheme II
Third generation ati ati ati
Cavity parameters • Gaussian beams • Strawman parameters • Items governing finesse • Items governing length
Resonant approach improves sensitivity to gaggby a factor of 300 or so. It can reach gagg ~ 2 x 10-11 GeV-1 in 90 days of live time. All the technology for such an experiment exists. TeV magnets. FNAL’s magnet capabilities enable both phases. Laser, cavity, instrument control, and readout adopt technology proven in LIGO and LISA. Conclusions