Laser Locking for Long-term Magneto-Optical Trap Stability Kevin W. Vogel Advisor: Georg Raithel I built and tested the dichroic atomic vapor laser locking system on the Advanced Physics Lab magneto-optical trap. My goal was to determine the locking system’s affect on the long-term stability of trapped atoms in the MOT. Presented 07/28/04
Outline Magneto-Optical Trap (MOT) Laser Locking Methods Dichroic Atomic Vapor Laser Locking MOT Improvements
Magneto-Optical Trap (MOT) Capture and cool Rubidium atoms to μK temps 6 orthogonal pairs of circularly polarized counter propagating laser beams Anti-Helmholtz magnetic field A MOT consists of three orthogonal pairs of circularly polarized counter propagating laser beams and an anti-Helmholtz magnetic field. Atoms are cooled by an exchange of momentum with absorbed photons. Atoms aquire momentum directly opposite to their direction of travel since the re-emission of photons is in random directions. (This re-emission is seen as the glowing atoms in the MOT.) The magnetic field places the atoms moving away from the center of the trap in resonance with the photons moving toward the trap, forcing them toward the center. At the center of the trap, atoms only interact with photons moving in the opposite direction as their motion. To optimize the absorption rate, the lasers are tuned 10MHz below the desired transition, Rb85 F=3 -> F=4 transition. To keep the atoms trapped, the laser frequency must be stabilized at this frequency. The atoms absorb laser photons with a frequency corresponding to the Rb85 F=3 -> F=4 transition The magnetic field enables the atoms to stay on resonance with the laser. Anti-Helmholtz magnetic field creates a quadrapole field with zero B-field at center. ▼B is the greatest at the center.
Diode Laser Frequency Stabilization Frequency changes due to temperature and diffraction grating position Tuned to transition frequency Locked with a feedback circuit Frequency ν 0 Volts The frequency of the laser will change due to temperature and diffraction grating position since these affect the cavity length. The laser is temperature controlled to minimize temperature effects. To lock the laser to a specific frequency, the laser is first tuned to the frequency and set at 0 volts. An electronic feedback circuit forces the voltage back to zero whenever fluctuations move it away. The steeper the slope, the smaller the range of frequencies possible while locked so overall the deviation in frequency is small. (A vertical line would ensure one frequency and a horizontal would allow any frequency.)
Laser Locking Methods Saturated Absorption Spectroscopy Narrow locked frequency range Easy to lose lock Lock time: 10 – 60 min. Dichroic Atomic Vapor Laser Locking (DAVLL) Difficult to lose lock Broader locked frequency range Lock time: ? 5 MHz 0 V 0 V 500 MHz Saturated absorption spectroscopy locking method is currently used in all setups in our lab. It has the advantage of a very steep slope, so the frequency range is small, but if fluctuations move the voltage too far off zero so as to leave the slope, the lock is lost. Therefore it is quite easy to lose lock, even by tapping the table or sudden noise. Dichroic atomic vapor laser locking uses a broader shallower locking slope. This allows the frequency to vary more widely than in the first method, but the wide range makes it very difficult to lose lock. Knocking on the table would temporarily allow the atoms to escape, but then the feedback circuit quickly corrects the voltage and the atoms reappear. My goal was to test this second method to determine the locking system’s affect on the long-term stability of trapped atoms in the MOT.
DAVLL Setup The locking system is set up so that it only uses a small portion of the laser’s power. The beam is feed through a Rubidium vapor cell within a 100 Gauss magnetic field. In my case I used ceramic permanent magnets to create the field since an electromagnet would have produced too many watts. The quarter-wave plate creates a phase difference of one-quarter cycle between the ordinary and extraordinary elements of light passing through. It is used in this case to ensure eual amounts of oppositely circularly polarized light reaches the two photodiodes. to MOT
DAVLL Lock Signal Transition shifted by Zeeman effect Laser output is linearly polarized The ECDL output is linearly polarized. Linear polarization can be thought of as equal amounts of right and left circularly polarized light. The transitions of the atoms in the magnetic field are split by the Zeeman Effect. Each polarization is absorbed by one of the Zeeman transitions. Each circular polarization is absorbed by a shifted transition
Improvements: Low Noise Circuit Temp Controlled Permanent Magnets Produces differential absorption signal with minimal electrical noise Temp Controlled Permanent Magnets Permanent magnet field strength is temperature dependent Keeps temp within ±0.003°C
Results 14 hours! mode hops The sudden drops are due to the laser changing modes. The laser may hop back into the correct mode. (Laser modes are the possible EM standing waves in the laser cavity.)
Other MOT Improvements Permanent heater to clean Rubidium cell Larger vacuum chamber cell to increase atom flow Magnetic coils for larger cell New laser grating and bracket Heating the Rb cell will remove the excess Rb off the cell windows where it would alter the lock point. The new chamber has a larger diameter for the atoms to flow through, thus increasing the number of atoms in the MOT. With a new larger cell, new magnetic coils must be made to produce the needed 200 G/cm magnetic field gradient at the center. Laser grating to increase laser power, enabling more atoms to be trapped.
Acknowledgments Georg Raithel, Ramon Torres-Isea, Spencer Olson, Rahul Mhaskar, Tara Cubel, Aaron Reinhard, Natalya Morrow, Rui Zhang, Brenton Knuffman, Alisa Walz-Flannigan, Jae-Hoon Choi, Eberhard Hansis, Alex Povilus NSF Physics Department