Thursday, April 28, 2016

So how do these LFC things work?

Derryck gave Tech Ops a great introduction to LFCs on Monday morning, here are a couple of slides from his presentation...
First of all - some laser basics
Then the parameters that define a LFC
& how those properties are established
The 4 primary components of the comb are shown in the pic below: the titanium sapphire laser (1), a diode laser stabilised to a rubidium transition (2), a photonic crystal fibre for super-continuum generation (3) & a Fabry-Perot filter cavity (4).
The main components of the LFC on the optical bench
So at the heart of the system is a titanium-sapphire femtosecond pulsed laser.  The Ti:sapphire crystal glows red/orange when excited by the laser light & the copper block that holds it needs to be water-cooled while the laser is running.  This main (Ti:sapphire) laser is pumped by a frequency-doubled, continuous wave, neodymium-doped yttrium orthovanadate (Nd:YVO4) laser that's powered at 6 W.  Frequency doubling of the Nd:YVO4 converts the invisible infrared light (at 1064 nm) to the characteristic green light at 532 nm that's needed to produce the red (800 nm) Ti:sapphire beam.  The ~30 cm round-trip path length of the Ti:sapphire laser cavity sets the repetition frequency (mode spacing of the comb) to just less than 1 GHz.  That 1.2 W beam is what's needed to make all the cool stuff happen, but there's quite a bit to be done to the light before we can fire it into our spectrograph.
A glimpse inside the Ti:sapphire laser, being pumped by a powerful Nd:YVO4 laser
There's a third laser in the system - a continuous wave diode laser that produces a ~40 mW beam of red light (around 780 nm).  This passes through a rubidium gas cell & one of the resulting transitions (which are superbly traceable atomic references) is then used to lock the Ti:sapphire modes.  The extremely well established wavelength of that line (780.24629 nm) also serves as a definitive reference point for our wavelength solution.  By measuring its exact position in our HRS spectra (to a minuscule fraction of a pixel), & then being able to count the comb teeth (whose separations in frequency space are exactly known due to the locked repetition frequency of the comb) will allow us to accurately map where each spectral feature is located in wavelength space.  This is The Holy Grail of wavelength calibration, which is absolutely critical for high resolution spectroscopy, & particularly for the precision radial velocity measurements needed to detect low mass exoplanets.  Of course all of this calibration work is Much easier said than done - but that's the principle anyway!
Power supply & controls for the diode laser that produces the rubidium line used to lock the comb spacing & establish the wavelength scale for the comb (& hence for the red arm of HRS!)
The next critical component on the LFC's optical bench is the 50 cm long loop of photonic crystal fibre (PCF).  The core of this fibre is only about 3 microns in diameter so the rubidium & Ti:sapphire beams need to be carefully steered & aligned in order to couple them properly into the fibre input (the reddish spot to the right of centre in the pic below).
Coupling the 2 laser beams into the photonic crystal fibre's input face
The cladding of the PCF has an elaborate micro-structure with air holes.  The holes reduce the effective refractive index of the cladding, which then improves the confinement of the modes & thus increases the intensity of the light within the fibre.  This high intensity is necessary to drive the non-linear process required to broaden the spectral range of the comb & to produce what's known as the super-continuum.  One can literally see the colour of the light change as it traverses the length of this extraordinary little fibre!
The photonic crystal fibre that broadens the spectral range of the comb by roughly a factor of ten
What looks like a continuous spectrum (similar to what one gets out of a spectrograph when you inject white light into it) is in fact made up of countless individual laser wavelengths - it's a rainbow of closely spaced laser spots!  It's these spots that have immaculately defined wavelengths that can be used to precisely calibrate our high resolution spectroscopic data.
The spectacularly beautiful super-continuum produced by the LFC
Making minute adjustments to the X/Y position of the fibre input using fine micrometers on the stage assembly changes the colour of the light emerging from the fibre exit!  While this may seem reasonable to a laser physicist, for most of us this is best explained by invoking the existence of pure magic...
Note the yellow & red spots on the hand, these are from the fibre output & the original Ti:sapphire beams
A tiny tweak of the fibre input turns the green beam from yellow, to orange, to red.  The profound novelty remains, even after a week of playing with this!
Minutely adjusting the fibre alignment varies the width of the super-continuum that's generated, changing the colour of the light emerging from the PCF
The last "sub-system" on the optical bench is the Fabry-Perot cavity.  This device is basically a tunable filter that's used to knock out all but every 15th line, so that the comb's not too dense to be useful.  A pair of partially reflective plane-parallel mirrors with complementary coatings are mounted on piezo-electric actuators that allow the spacing of the mirrors to be tightly controlled.  The mirrors first need to be optimally aligned using the precision micrometers that control the tip & tilt of the smaller mirror with respect to the larger, fixed one.  The light bounces between the two mirrors & constructive interference leads to certain wavelength ranges being preserved (transmitted) & others being eliminated through destructive interference.  
The Fabry-Perot cavity used to filter the super-continuum: the small mirror to the right of the centre of the image gets aligned in tip/tilt wrt the larger one on the right, & then the mirror separation is controlled with piezo-electric actuators until the required wavelength range is transmitted
It's this filtered version of the comb that then gets directed to the injection fibre (the blue cable visible near the back of the bench in the pic below).  That fibre is used to convey the comb light to the HRS high stability bench, where it gets introduced into the spectrograph in much the same way as the ThAr lamp light that's conventionally used for HRS wavelength calibration.
Aligned & locked LFC - with the lights on in the room
Each of these units need to be locked to preserve the comb.  The Ti:sapphire laser is mode-locked, & then there are 3 electronic locks: the diode laser to the sharpest of the rubidium transitions, the repetition rate (the frequency that sets the spacing of the comb teeth) & then the Fabry-Perot cavity.  With all of these locked & stable, we have a LFC that's good to go!
The scene's even more spectacular when the room lights are switched off!
More tomorrow about injecting the comb light into HRS!

No comments:

Post a Comment