Hello Orbital Friends! It’s Friday, so here’s what Orbital Arc was up to the last two weeks (since last Friday was the day after Thanksgiving, and I skipped an update).
First, the annealing furnace at Rice is still broken, so lab progress remains stalled. This is a source of utter frustration, so I spent a significant amount of time the last two weeks looking for other labs I might work in. While each place will probably require me to train on all their equipment specifically (restarting the training slog), having more options on where and how to work will likely help prevent future slowdowns like this one.
The search for other labs has been a modest success. University of Houston has a nanofab and a separate research center that includes a few fab-like capabilities. Texas A&M has a pretty nice nanofab. UT Austin has a nanofab and a really nice looking separate research center listed online, which includes on-site CMP (All other labs I’ve seen use off-site vendors for chemical mechanical planarization, which makes any process that includes a CMP step a multi-week turnaround endeavor where you mail super-fragile components to the vendor, they make those components smoother, then mail them back). This Microelectronic Research Center at UT Austin also has a magnetron sputter system with a 6 inch target – set up for tantalum, but it could run the molybdenum process I found in literature that gets me 5 microns in 30 minutes instead of several weeks.
The search has been a modest success. The engagement with these labs has been a failure so far.
University of Houston has not responded to emails (They don’t really have the right tools anyway, so I haven’t been bugging them).
Texas A&M has a sputter system with 4 inch targets, which could at least run the process parameters I found for the fast sputter process, but they won’t let me attempt it – like Rice, they don’t allow thick layers in their sputter system. Not because the systems can’t do it, but because doing it wrong could damage the machine, and if I were to damage the machine, then dozens of other projects from all the other users would be stalled for a long time until it could be fixed (which, as we have seen at Rice, could take months). That’s a reasonable, if frustrating, objection. I also asked if they had a furnace that could allow me to circumvent the waiting for repairs at Rice, but TAMU also does not have an annealing furnace that can handle the temperature, duration, and atmosphere control requirements of my annealing process. Two swings, two misses there.
UT Austin, particularly the Microelectronics Research Center, is a different kind of story. That lab would be the most capable option I have seen in the entire state of Texas, except, I can’t get ahold of anyone there. You can find their website, with a list of all their tools and projects, but it looks like a pretty old website. When you email the listed contact in charge of the facility, you get a bounceback – email not in service. When you call the listed contact number, you get an answering machine. The mailbox was not full, so I left a message, but it has been two days, and nobody called me back. When you call main UT Austin and ask them to transfer you to the MRC, they can do so, but then it rings forever, and eventually just automatically disconnects the call. So, I’m not even sure this place still exists; my current plan is to drive to Austin, show up at their door, and see if anyone or anything is inside.
That would be a bit extreme – Austin is a 3 hour drive, which is a lot, just to knock a door. But, I’ve got a few other pieces of business to do in Austin, and my wife and I have wanted to visit Texas wine country for a long time anyway, so we’ll go next week and make a road trip of it.
Other than trying to chase down an annealing furnace (Seriously, if it isn’t fixed by New Years, I am going to retrofit my kiln with a nitrogen line, pump in displacing nitrogen at high flow, and just DIY this furnace. All I need is 1000 C for an hour at a time with no oxygen. It should not be this hard.) I have been continuing to research the tiny details of thruster operation, specifically photodissociation and photoionization potential.
See, when you shoot ions out into space, you need to shoot electrons out with them, to neutralize the space charge. Since the electron-ion recombination is energetically favorable, it will release energy from the resulting molecule. That energy will probably be a photon, which could fly off in any direction upon release. And, some of those photons will fly back toward the spacecraft.
In the basic case, as many as 50% of those photons might be absorbed by the spacecraft, adding to the thermal load from operating the thruster. In the more complicated case, if the emitted photons come from the energy released when a particle moves from an ionized state to a ground state, that should be the same energy needed to move a particle from the ground state to the ionized state – in other words, the photon might come out, fly off, and ionize something else, which is bad because we only want ions to be made the way we intend and in the locations we intend. So, when I thought about this possibility I became a bit worried about secondary photoionization as a potential additional wear mechanism.
After a week or so of research, I think I was worried for nothing. Photochemistry is yet another new discipline for me, and not one that I have any intention of becoming deeply skilled at, but I’ve learned enough to know that most of the time the path from ion to neutral won’t happen in one big jump, so the photon emitted won’t have full ionization energy. When the electron recombines with the ion, the resulting molecule will be in an excited state. To de-excite, it can bump into something and transfer energy kinetically, decompose itself to constituent atoms that carry the excess energy off as thermal energy, or emit energy via photons. Only certain, specific wavelengths of photons are permitted from any given energy state, though – there are quantum levels of excitation that are preferred, and usually the photons emitted from a given atom correspond to the energy difference between one specific preferred state and another.
This quantum-level-difference-based emission is taken advantage of in lasers. We use some kind of flash pulse to inject a bunch of energy into a laser medium, and excite most of the molecules in the medium. These molecules may get varying amounts of energy, but the can only emit energy amounts corresponding to a change of energy level between two preferred states, so you go from a random mix of energy applied to a coherent selection of wavelengths emitted. There’s more to it than that – the passage of a preferred wavelength photon near an excited molecule can “stimulate” the emission of another preferred wavelength photon of the same wavelength, leading to a signal gain at a specific wavelength; this stimulation (that’s the “s” in the “laser” acronym) is what makes lasers so good at coherent emission of specific wavelength spectra.
The tricky thing with lasers, though, is that the spectra of emission is always lower energy than the energy injection system. You flash the laser medium with 194 nm light, and it emits less energetic 405 nm light (these numbers are made up, but illustrative). The output is ALWAYS lower energy than the input, because the emission of the energy isn’t all at once, it happens in steps between non-zero energy levels.
As long as that last sentence remains true, it should be very nearly impossible for recombination to release photons that can cause ionization upstream, at least as I understand it now. Problem averted.
Multiphoton ionization could still occur, but the photon intensity needed to make that happen regularly is so high that the probability for any given molecule to experience it will be vanishingly small. Turns out that photoionization in general is a super energy inefficient process – we struggle to get it to happen even when we want it to happen, and consider 0.1% power efficiency a huge success for high energy, short wavelength vacuum UV laser operation, and even that is still at energies well below typical ionizing thresholds.
And, that’s the last two weeks. Sadly, not much progress on hardware. On research, better, in that I learned enough about a thing to worry about it, and then learned enough about it to stop worrying about it as much, and now I can stop learning about it, and go back to finding universities and commercial fab labs to pester.