Open-air hot powder cell—costs and planning

As for possible ideas for the actual experimental setup, I found that in the study below the authors use a [presumably automated] syringe pump and a capillary tube to inject water+ethylbenzene inside a tubular reactor not exceedingly different in principle than what I had in mind. Gases like N₂, H₂, CO₂ are fed through different conduits. They report that the liquids get immediately vaporized upon entering the chamber (at 600 °C) and prior getting in contact with the catalyst sample. The main differences here from what I wanted to achieve are that the reactants flow over the catalyst sample (here in milligram-amounts) rather than being forced to diffuse through it, and heater location/arrangement. Also, the catalysts are synthesized ex-situ.

Li Z, Shanks BH. Stability and phase transitions of potassium-promoted iron oxide in various gas phase environments. Applied Catalysis A: General. 2009 Feb 15;354(1-2):50-6. https://doi.org/10.1016/j.apcata.2008.11.007

After some pondering, the gas admission system could be simpler than I assumed. Since contrarily to other experiment types the presence of slight amounts of water is not deemed harmful here and if anything sought after, a rudimentary external electrolysis setup should be fine. This will allow to introduce directly H₂ but also O₂ to keep the material oxidized and active if it gets reduced excessively (although this will be difficult to tell as inspection will only be possible from the top surface exposed to the atmosphere). Only slight amounts of gas would be needed for it to function and I don't expect significant pressures to be generated in the bottom empty region. External gas admission could be used together with the reactor tube electrodes.

Using 10 mm and 25 mm OD SS tubes as proposed by gerold.s, assuming 1 mm wall thickness for the outer tube and an active zone height of 30 mm, I get a volume of about 10 cm³, which should be about 15 g of catalyst material assuming a density of 1.5 g/cm³ (similar commercially available catalysts also have a bulk density of about this range).

I have only quickly looked at how glow plugs are generally made and the hot region does not usually look very large.

Since in this experiment the supposed (ultra-) dense hydrogen clusters would be generated as hydrogen atoms diffuse through the catalytic material, intuitively speaking it seems like it would be more useful to have a long narrow path rather than a short and thick one. The longer hydrogen atoms travel inside the catalyst's pore system (continuously getting adsorbed-desorbed from its internal "walls"), the higher the probability of them converting into a dense form inside of it—at least according to my idea.

However, a too long path might make more difficult for a current to be uniformly passed through the catalyst, since temperature gradients, local material distribution and possible hot spots will affect its conductivity. In the worst case scenario a short-circuit could occur which will damage the catalyst and the possibility of reliably passing a current through it afterwards.

From the above, three independently controllable power supplies/sources having different characteristics would be required to run the experiment as described.

To better show the idea, today I took the time to make a 3d model of the apparatus as presented in the previous diagram. Maybe it's still too fancy however, and there are still some issues to be solved in the design (mainly the insulating internal spacers, and the bottom part which would require machining and/or welding to the outer tube).

Perhaps a simpler design could omit the concentric tube electrodes (+insulators) and only use an internal heater with simple closed-end tubes, with admission of hydrogen and other gases through a long small tube from the top opening. This will allow to use cheaper and more inert metallic spacers made of metal.

I still believe that the electrode arrangement could be very important. On the other hand, provided that I manage to find ready-made closed-end tubes of appropriate size (or perhaps get the proper tooling to thread and seal them with short bolts), this arrangement might allow testing things out rapidly and cheaply. A long and narrow path for the admitted gases should be preferable to a short and wide one, but perhaps I'm overthinking it and ordinary readily available stainless steel containers will be just fine.

As a side note, since the catalyst becomes conducting at high temperatures, a probably familiar alternative arrangement should be possible as well, although the longer the tube will be the more difficult it will in turn be to pass a current through it.

After some practical considerations on constraints regarding tooling, ease of building/maintenance and thermocouple placement, to achieve what I initially planned I think I will have to give up on the idea of the internal heater and do something similar to this diagram:

The main questions now are:

• Will a glass fiber sheath be enough to electrically insulate the heating wire (intermittently fed with AC power) from the outer steel tube? Dubious.
• Either I can find pre-made thermocouple-type ceramic insulators of the ideal size or I have to find something else. Another problem is their resistance against a strongly alkaline environment.
• Thermocouple compatibility with the tube as arranged in the diagram, once a current start flowing through it?
• The catalytic material will preferably have to be compacted homogeneously before heat is applied, but the way the setup is arranged will likely make it difficult.
• Outer insulation not shown and not fully considered yet. If the tube orientation is set to be horizontal, machinable refractory bricks could be convenient to use.

Additional points for clarity:

• A direct current at low voltage (about 12V) is intended to be passed through the catalytic material (in green), which should become conductive at moderate to high temperatures. However, control hasn't been fully thought of yet. At the very least, since its resistance will decrease strongly with temperature, some sort of current limiting means will be necessary. This alone could be as simple as a resistor, but other issues exist.
• The washer-nut portion on the right is intended to provide limited protection against hot gases to the electrical wires located there.
• The hydraulic fitting on the left is intended to be connected to an external standard water electrolysis system.
• I haven't purchased anything yet. Mainly, I'm waiting for a couple papers to come out (one of which possibly of direct relevance); afterwards I'll decide if this is really worth attempting, given the possible safety issues (fire/burn hazards; explosions).

Vector (SVG) file of the diagram attached. I realize that it's not exactly a CAD drawing.

Alan Smith

As in this very small photo?

It looks like they would increase considerably the size of the heating wire.

As for compressing the catalytic material, I just thought that perhaps something along these lines could work, although it would make the assembly more complex and increase the number of parts. Only the inner portion is shown; imagine it inside a steel tube as in the previously posted diagram.

How conductive exactly is the catalytic material that I planned to use? At least in pure form, according to https://doi.org/10.1134/S1063783413050351, the ionic conductivity (EDIT: which is one component of the total conductivity; the other is electronic conductivity, but here it's not shown) is about 5×10^-8 S/cm at room temperature and 2.5×10^-2 S/cm at 900 K. The resistivity in Ohm.cm is the reciprocal (1/x) of these values. I've converted the graph provided in the paper (fig.1, visible in the preview from the link) in a more easily readable form (below, to the right), where the fast increase with temperature (up to a certain point, at least) is also clearer:

EDIT: in https://doi.org/10.1134/S1023193507090017 the total conductivity of the same compound at different concentrations of doping with another element is investigated. The line #1 on the graph on the left (from the paper) corresponds to the pure catalytic compound. Again, I tried converting the graph into a more easily readable form.

This study stops at lower temperatures, but the general characteristics of the curve up to roughly the same point is similar. The compound is almost an insulator at room temperature and fairly conductive at high ones: actively regulating the operation of such behavior could be almost as complex as with an arc discharge system.

gerold.s

After investing quite a deal of time in the past couple days considering different options, I'm slowly coming to the idea that an internal glow plug (enclosed inside a metal tube) as you initially suggested could in practice pose less difficulties—that is, unless other issues arise.

Main reasons I found so far:

• Standard low-cost cartridge heaters (e.g. from many Chinese sources) do not seem to be suitable for temperatures higher than 600°C (likely less than this, given their construction and materials used) or holding them continuously.
• Cost-effectively electrically Insulating a heating wire at high temperature from the steel outer tube is not as straightforward as it appeared initially (fiberglass wire sleeves are typically rated up to 500–600 °C; ceramic fish-spine beads are expensive per unit for the amount needed and increase the size of the heating element considerably).

However glow plugs in turn also have possible disadvantages:

• Hot zone with a limited length (often about 20–25 mm), which means that the catalyst over the planned length would not be homogeneously heated. This could be the largest problem in my case.
• Heating system overall more expensive than a SCR-driven AC heater.
• Needs electrically insulating the metal shell from the outer tube, as it's the negative terminal (this condition is not different than the original concept of the setup—polarity is the opposite though—but it means that a quick-and-dirty test might be less straightforward).
• Long-term durability? They are not usually intended to be used for a too long time continuously.

What are the measurements and electrical characteristics of the glow plugs you had in mind? I'm aware from a cursory search that they are designed for a nominal operating voltage of generally 10–24V (depending on the actual model—many different types and sizes exist) and that their DC resistance can vary depending on temperature from fractions of an Ohm to several Ohm. A beefy DC power supply will be needed to drive them, but it would be the same needed to conduct a current directly though the catalyst (in the full setup).

EDIT: so in practice, using actual glow plug measurements, it could be something along these lines, but it would be using less than half the amount of catalytic material compared to initial plans. Furthermore, previous issues related to the insulators, thermocouple placement and connection to the gas feed would still need to be solved.

EDIT: or possibly, something like this (diagram rotated to fit the 16:9 thumbnail), if input voltage is the same for both connections (e.g. 12V).

The electrical characteristics / response of the system could become complex (at low temperature most of the current would go through the glow plug, but at higher ones it would mostly be through the catalyst), but temperature-controlled operation would be simplified.

Wyttenbach

Spark plugs wouldn't be suitable here, though. Glow plugs just become incandescent; they are used in diesel engines.

magicsound

The setup is still far from being finalized, but thermocouple placement and its electrical insulation definitely will be an issue. Since temperature changes when passing a current directly through the catalytic material might be quick (if anything, due to its resistance decreasing rapidly with temperature and possible formation of hot spots due to material inhomogeneities), the one controlling the heating would need to be put close to the hot area.

EDIT: a related problem in any case is that typical cheap solid state relays intended for driving on/off (e.g. with a PID temperature controller) AC loads will not be suitable for a high-current DC load (glow plug + low-voltage direct current through the tubes), so the simplified diagram above will need something else than that.

Alan Smith

I should probably add that to the parts list. Possibly an actively cooled heat sink will increase reliability. (EDIT: Apparently a 100A version of the same SSR also exists. Serious cooling will be likely needed though).

Earlier today I tried finding the current draw characteristics of typical glow plugs. They can initially draw also above 25A (each), but as they get to operating temperature this should settle to the manufacturer's rated current value. The second graph below should be for 4 glow plugs plugs. Actual current draw will depend on model, type, conditions, etc.

http://www.dieselrxproducts.com/glow-plugs.php

https://www.picoauto.com/libra…e-guided-tests/glow-plugs

For example these NGK 11.5V glow plugs for Ford Transit diesel engines are rated for a 5A current draw. However resistance is listed at 0.9 Ohm; I'm guessing this is when cold: https://www.ngkntk.com/part-fi…SIT%202006.5/Y-548J/5785/

When considering the maximum current draw of the glow plug + load through the catalyst however the overall load could potentially be larger (or much larger if a short-circuit occurs) than 40A unless at the very least a ballast resistor is used.

gerold.s

I see, so a glow plug intended for VAG TDI 2.0 engines produced around 2005-2006. The NGK catalog I found has some measurements. It lists it as a 7V glow plug though.

My idea was primarily getting one as close as possible to 12V so that standard switching DC power supplies could be directly used, and to employ standard models made for popular cars or trucks which are available for cheaper prices.

The whole rationale behind potential parts choice for me so far has been minimizing costs by using off-the-shelf components and avoiding custom ones that need machining or welding. This would also decrease downtimes in case anything breaks, which is likely since there's no guarantee at all that the setup will indeed work as intended. So while purpose-machined Macor insulators will probably be the best choice from a practical standpoint, they would fall short on those other aspects.

I tried anyway imagining a possible setup without too many of those constraints, and therefore unlikely to be realized from my side. Engineering issues might also exist (the caps for example will likely end up having ports oriented randomly instead of neatly as in the diagram). The internal thermocouple could be located around the red spot near the tip of the glow plug, coming from the opposite side. I don't know what kind of commercially available power supply would be suitable for 40A at about 7 V DC. Current through the tubes would not strictly need to be in parallel with that to the glow plug and control would be more precise if both inputs are separated; this arrangement seems convenient however.

Alan Smith

I was thinking of CPU coolers with a fast fan and a good thermal compound, but I've seen that heat sinks made on purpose for these solid state relays also exist.

gerold.s

The central tube is required to pass (coaxially) a current through the catalyst in a hopefully homogeneous fashion, which is one of the the main differences from most other experiments. The catalyst material becomes an ionic conductor at sufficiently high (>500–550 C) temperatures as some graphs I posted a few comments earlier show. This highly temperature-dependent behavior is a reason why I needed the catalyst to be homogeneously heated and thought that the small hot spot provided by the incandescent glow plug might not be entirely suitable for this.

Over time (not clear exactly how much) a high potassium concentration could be reached near the surface of the negative electrode and at high temperatures this could quickly affect the ceramic tip of the of the glow plug, so a steel tube to keep the material away from it would probably be desired anyway. To keep the K concentration in the catalyst more or less homogeneous, according to the original idea electrode polarity would be periodically switched (at a slow rate) but this would complicate control further.

The idea of the central tube is also that it allows to quickly switch other thermocouples or heaters if necessary, without affecting the sintered catalyst (although in practice the process will likely be more involved than expected and affect the catalyst anyway).

If it was just a matter of heating some sintered powder at high temperature, it could have been done with many less complications.

For the diagrams I use a vector drawing program called Inkscape. https://inkscape.org/

Paradigmnoia

From what I've read recently, glow plugs in modern turbodiesel engines have a more active role in that they help regulating combustion chamber temperature for emission control reasons, and I also previously thought that they are not meant to be used for more than a few tens of seconds at a time. I was initially skeptical about glow plugs also for this reason.

Anyway I still haven't purchased anything and this discussion so far is still mostly at a brainstorming level. Until a few days ago I was almost ready to get parts for a quick-n-dirty setup with a heating wire, though.