As a senior laser technician for a major university, and as a former field service engineer for a maker of solid state lasers, as well as a former laser show guy, I have twenty or more years experience aligning and modifying laser cavities. Often aligning them in places other then nice, stable, labs.
(For those who want to know more about that qualification, HENE, HECAD, ION, N2, Xenon, OPSL, CVL, OPA, OPO, YAG, MOPA, Three Wave, Nano, Femto and more, and 32 Joules of 532 in 700 Picoseconds)
I had a good laugh with this one. Setting down in New Zealand , in a deep cave, is a helium-neon ring laser gyro cavity that is a square ten meters on a side. Its mirrors are selected to generously help correct with diffraction losses and to correct for astigmatism as the beam folds. However it has to run with the beam paths in a vacuum to avoid diffractive effects from air currents, which would quickly stop lasing. It runs at a few hundred microwatts, so there goes power transfer with that one. It does a wonderful job measuring relativistic effects as the earth turns, and is a mighty fine micro-tremor detector too...
Long lasers are a holy grail, as generally, the longer the laser cavity is, the lower the divergence is, to a limit that arises on how small the beam diameter can get and also extract power from the gain medium, and the other lower limit on beam size is tied to diffraction (even in a vacuum) and scales based on wavelength. There is a vary simple equation based on mirror radius and spacing for a two meter cavity that spells all this out.
At some point in time you hit a physical limit on how shallow the radius of the cavity mirrors are. While I have personally used mirrors with a 10 meter radius, anything longer then that is essentially a flat. The ten meter radius in the simple case means a practical distance limit of just under four and a half meters. So as the distance from the laser medium to the mirrors change, the mirror radius has to change. Yes, in some cases, a flat mirror flat mirror pair is used, but they are nearly impossible to align and are known for very, very, low power extraction from the laser medium. The need for a changing mirror radius for maximal power extraction is a given.
Yes, you can have intracavity lenses to compensate, but as the lens focus moves it drastically would change the beam path, taking you right out of the active region. Because the sides of your moving, curved, lenses are inherently a strong prism. So now you need a six axis positioner to move the compensating lens to correct for its own distortion as you zoom focus. The mechanics of this task, let alone the automation, would be amazing.
There are some tricks with prisms and retro reflectors to fold the cavity back onto itself for "self-alignment" but none of them would even begin to fit on a phone, let alone work over more then say three meters,
With a one to three meter laser cavity, I'm usually having to have the strongly curved mirrors aligned to one milli-radian or better. To get there I have to make the laser cavity out of very low expansion, or expansion compensated materials. Usually I'm adjusting eighty pitch screws to get enough angular resolution to peak the alignment. Then, even if we're lasing, we have to transverse the beam path across the beam path in the lasing medium to find the maximal power path, which usually takes a human about thirty minutes, and a automated laser alignment system about two to ten. There are a few scientific and industrial lasers that "Auto-Align" and "Auto Peak" with large arrays of stepper motors. Generally for them to work, a Human pre-aligns the output mirror until they can achieve capture. I've never seen a system, outside of a laser production line in the factory, that can align without a human pre-aligning one end.
"Magical" techniques such as optical phase conjunction have a third order dependence on incident power, so there goes any self aligning nonlinear crystal resonators, as they would need hundreds of millijoules of coherent light just to start.. (Translation, NO STAR TREK style methodology , the scattered light from the dust in the air or the "probe " beam hitting a wall would be hazardous before it even started power transfer. ) The other side effect of optical phase conjunction is it would try to start lasing with any nearby polished metal object, such as a fork or metallic watchband.
The box on the phone would be even bigger then what you would need for uBeam's ultrasound. The vibration that occurs with a kitchen table or restaurant table would constantly force the system horribly out of alignment, and the energy to constantly servo the motor plus the fine adjust piezo would be enormous. Laying your fist palm down, gently, anywhere near the phone is going to constantly force a re-align. Did I mention that long lasers are ran on vibration dampers, often active dampers, or air tables, that decouple vibration. Very long classical lasers are mounted in tunnels or caves, or on huge concrete pads, usually on the ground floor, just to avoid any bending in the structure that would result in detuning. This is why the switch to fiber lasers have proven to be the needed power breakthrough for making effective laser weapons.
I'm also awaiting the magical silicon cell that would not saturate well below the intensity per unit area provided by such a tight beam. After all, I usually have three ceramic scattering attenuator disks on a four percent beam pickoff for a five watt laser, to keep the beam from basically shorting the detection cell out with saturation issues. The other two methods of laser beam power conversion typically used are a piezo detector for high energy pulsed beams, and a Peltier with a black adsorbing disk in reverse for continuous wave beams. The "Power meter" detectors, aka thermopiles, need huge heat passive heat sinks to function.
I have not even started on the amount of dust in the air in a non-lab environment.
ALL the lasers I have ever worked on have enclosed beam paths to negate power loss and beam distortion from dust, oxygen, and moisture in the air.
NO, not in this century, and the hardware requirements would be amazingly expensive, even if you used ND:YAG or CO2 lasing mediums, Direct diode lasing mediums would be out of the question due to diffractive effects from the small chip size.
I'm not even going to consider self mode locking, self q-switching, regulatory, eye safety, mode spacing, and interferometric effects.
Please don't cite long interferometers like LIGO etc as a counter argument, those are an entirely different hill of beans or beams, as they are not self lasing to begin with, and run in a vacuum.
If any one does not believe me, I'll point you to a link to free cavity modeling software, where you can add up to five optical elements and it solves the Jones matrices for stability.
Yeah, it would wink out with little power transfer if blocked completely, but until the beam is sufficiently blocked to have the losses exceed the gain in the system, something like a bare wrist or face might get burned by grazing incidence illumination.
A slight layer of dust on the intra-cavity optics will quickly stop lasing, and a fingerprint on any optic would be fatal to the process. No one is going to hand out USP or Spectroscopy grade acetone to clean the mirrors every five minutes.
NO, NO, and NO, and if it could work, the real time computing needed to track the mirrors would burn any power transfer to shreds.
For 13 Million Dollars I will "attempt" to build you a working long distance laser cavity in a clean room for your Angel demo, and then arrange to flee the country the next day..
Steve