MIZUNO REPLICATION AND MATERIALS ONLY

  • I read that Storms heated a piece of Ni mesh to blue ie noteable thickness of NiO. He found it impossible to transfer Pd, even though he rubbed the crap out of it. As though the oxide layer acts like a lubricant, preventing the galling action. Also, judging by AlanG's awesome s.e.m., presumably the NiO layer also gets mixed up with the Pd too.


    Did he actually mean that? I read/interpreted that as the NiO layer getting removed by the procedure and the loss of material counterbalancing the deposition of Pd, making it look as if no Pd has been transferred.


    In his predictions further in the text (paper here) he writes that a sufficiently thick oxide layer on the NiO would make the material more active and suggests that other oxide-forming materials could be used as the substrate. If transferring Pd was impossible with NiO, why would he suggest this?


    I thought that to be consistent with what I observed in my tests, although I used different materials. The oxide layer formed by flame heating the substrate was significantly rougher than the starting surface and seemingly made burnishing the other material easier. Most of the oxide layer fell off in the process, which I'm assuming would cause a weight loss (without a microbalance it would be difficult to tell, however).

  • I'm considering doing most of the vacuum plumbing with fat KF-16 flanged piping and bellows hoses now.


    Chamber can bake to 350+C high temperatures. The KF-16 valves can go to 150 C, which I think is good enough. Obviously the piping between the valves can go higher, if it is even needed.


    My deuterium plumbing will have some smaller piping.


    - The valve next to the chamber allows me to disconnect the chamber from the rest of the vacuum system.

    - The valve in front of the turbopump allows me to send D2 into the chamber to scavenge oxygen, and to the actual experiment with D2.



    Do I really need the valve that has a quesion mark next to it? I don't think so...


    edit: I have a Pirani gauge and an ion gauge. I could put these just after the chamber disconnect valve (on the side that is further from the chamber).

  • I read that Storms heated a piece of Ni mesh to blue ie noteable thickness of NiO. He found it impossible to transfer Pd, even though he rubbed the crap out of it.


    Tricky that - the oxide layer formed by presumably intense heat is - potentially at least - physically different to that formed at low temperatures- it may for example be glassy in nature, rather than abrasive. But Ed is a careful and skilful experimenter, and has probably considered things like that.

  • Did he actually mean that? I read/interpreted that as the NiO layer getting removed by the procedure and the loss of material counterbalancing the deposition of Pd, making it look as if no Pd has been transferred.

    Yes, I am pretty sure he did.


    Here is the text I was referring to , found in the paper you linked to. "A second sample of Ni was flame heated to about 800° C to form a visible (blue color) oxide layer after which it was burnished in the same way as the first sample. Although the blue color was removed to produce a bright metallic surface, no detectable weight increase (±0.00005 g) was produced after 600 strokes. Apparently the nature of the Ni surface affects the amount of Pd transferred to the surface. Consequently, the condition of the surface is revealed as being another important variable. This condition needs to be explored."


    He quite clearly states, after the 'no weight gain' statement, that the surface of the Ni affects the amount of Pd transferred. He wouldn't say that if he meant that although there was no nett gain in weight, there was Pd transfer that offset almost perfectly, the loss of weight of NiO. Also, there is the very large number of 600 strokes, which, unless the NiO were making less effective, would be expected to transfer much more Pd, surely.


    Also, there is the point that there is almost never 'no NiO'. A layer forms instantly upon a clean surface being exposed to air. That is why it is almost impossible to solder Ni-containing metals- the oxide passivates them. I did indeed make the point that the NiO becoming involved in the Pd is likely to be relevant.


    I took Storms comment to mean that there are various parameters that govern the nature of the oxide, and there seems to need to be just the right amount present. It is a very effective thing, the oxide layer. i have some high purity Ni cathode that has been guillotined into 30mm squares. On the smooth sheared edges, it looks like they were cut 5 minutes ago, yet they have been in an open tub for about 3 years!


  • StevieH

    I thought the keyword here was "apparently" i.e. that it makes it look "as if"; but also that it might not necessarily be the case and this would need to be studied ("explored") more in detail. I'm not familiar enough with Storms' papers to know if he weights words this way, but other researchers seemingly do, so I might be reading too much into the text.


    In any case, the oxide formed on the surface of my Fe samples (after heating up to about 650-700°C in a reducing oxidizing (EDIT: blue-looking, "lean". Wiki) flame and quenching in water a few times) is definitely more of an abrasive nature, along what Alan Smith suggests above.

  • Tricky that - the oxide layer formed by presumably intense heat is - potentially at least - physically different to that formed at low temperatures- it may for example be glassy in nature, rather than abrasive.

    Yes indeed, that might imply a different nature. It may also depend on how it was heated. Storms used a flame. That may bring about some different form. From reading, I know that Ni can be resistance heated in air, to 1500+K, and it is covered with simple oxide crystals, or even mono crystals, which are quite easy to reduce in H2.


    The other interesting thing I noted was that contrary to poplar opinion, (including mine), it is far from safe to assume that in H2 at ~350C (taking that at a max bake-out type temperature), that any NiO is automatically reduced. (Not that it is ever safe to assume anything, of course). In the situation mentioned, heating to around 750K, or ~470C- higher than the TM reactor, only 65% reduction is found. That is from a 30mu thickness though, which I think is much thicker than a normal layer from just standing in air, but the point is that reduction is by no means automatically quantitative.


    So there is plenty of room for NiO, especially if intercalated in the Ni, to still be around and active after all the prep work in TM's reactor. That's why I was thinking that some kind of tracking of it, like AlanG with the CaCO3 could well be of value.

  • In any case, the oxide formed on the surface of my Fe samples (after heating up to about 650-700°C in a reducing flame and quenching in water a few times) is definitely more of an abrasive nature, along what Alan Smith suggests above.

    I interpreted what Alan said about 'glassy' to mean that it is slippery, which is why it doesn't transfer the Pd. He also uses the words 'rather than abrasive', which do tend to confirm that, I think.


    Anyway, if most of it comes off during burnishing when it is in this form, it does imply that it is not particularly coherent. Either way, it would be nice to get an idea of the actual amount of it which gets involved with the Pd during the rubbing process, on a naturally oxidised Ni like the one TM had his success with.

  • Someone else who tried to burnish found that it did not work. An SEM showed little or no Pd on the Ni. We are trying to determine why. I think an analysis of Mizuno's Pd rod would help.


    Earlier I wrote about the relative Mohs hardness of Ni and Pd, having found that worked (stamped) Pd could exceed the hardness of Ni. For that reason, I annealed my Pd chip, using the oxygen-rich part of an ordinary Propane torch.


    As far as detecting an oxide layer on the Ni, the attached image of the as-received mesh shows no Oxygen present in detectable quantity. However, surface analysis is tricky using SEM/EDS.The electron beam penetrates several microns into the target, and the x-ray fluorescence passes back out through the surface to the detector. This makes quantitative analysis of thin layers questionable.


    AlanG

  • I interpreted what Alan said about 'glassy' to mean that it is slippery, which is why it doesn't transfer the Pd. He also uses the words 'rather than abrasive', which do tend to confirm that, I think.


    I meant that of the two categories suggested by Alan Smith, the oxides formed from Fe/carbon steel as described in my case seem to be of the abrasive type rather than glassy and slippery as possibly observed by Storms from NiO.


    It's a rather thin layer. When I tried a similar procedure on a CuNi-Ni coin piece (I couldn't get it higher than 600°C, visually), it formed an apparently thicker blue-black layer similar to what Storms reports for his Ni sheet. Hard to tell if it's more slippery than the clean portions though. I'd like to try on a Ni sheet at some point.

  • Do I really need the valve that has a quesion mark next to it? I don't think so...


    The cold trap itself is problematic. If this is a conventional carbon/dry ice/alchohol cold trap then the D2 will not get past it. I think that what you need is a gas drier, zeolite molecular sieve is the way to go for that. Normally you would have a cold trap between reactor and QMS, in order to remove remnant D2 and let the helium go by - since a QMS would struggle to distinguish between the two.


  • As far as detecting an oxide layer on the Ni, the attached image of the as-received mesh shows no Oxygen present in detectable quantity. However, surface analysis is tricky using SEM/EDS.The electron beam penetrates several microns into the target, and the x-ray fluorescence passes back out through the surface to the detector. This makes quantitative analysis of thin layers questionable.


    AlanG


    I know this was for the common good Alan, but as I have been harping on about the oxide layer, on even what looks like clean Ni, I would like to say a personal thank you for trying to find it. I took your point about sem's penetration being too high for thin surface layers, and did some more reading. I came up with the paper in the link. The oxide layer defininitely is there, and its thickness is more like a few nanometers, if that. So not surprising that you couldn't see it.


    Of course, like the CaCO3 you found, we don't know if the oxide is relevant, or a benign spectator, but being aware in our situation is the best we can do. At least experiments can find how involved things are, as long as we are aware of them in the first place.


    Just one question re the EDS, for my own curiosity, do you know what the small peak is at about 8.3 keV?


    https://arxiv.org/pdf/1104.3481

  • Earlier I wrote about the relative Mohs hardness of Ni and Pd, having found that worked (stamped) Pd could exceed the hardness of Ni. For that reason, I annealed my Pd chip, using the oxygen-rich part of an ordinary Propane torch.


    As I mentioned, the person who thought he was dealing with Pd harder than Ni may have been wrong. He later discovered fragments of wire from the edges of the mesh, left in the try. That might account for the weight loss.

  • Here's another possible clue about the observed Calcite crystals , perhaps worthy of discussion:


    Calcite is known for it's unusual optical property of polarization-splitting of light, from UV down to 5 um in the deep infrared. The observed crystals are regular enough in morphology to possibly exhibit this effect. So IR from the heater could be split into two physically separate polarized IR beams. I don't know if that is part of the puzzle, but it's worth some thought.


    https://micro.magnet.fsu.edu/primer/java/prismsandbeamsplitters/polarizing


    Just one question re the EDS, for my own curiosity, do you know what the small peak is at about 8.3 keV?

    StevieH I will have another look at the sample, with the discriminator range adjusted to identify that peak. I think it might be from the Carbon sticky pad to which the mesh is attached. There is enough scattering of the incident electron beam that some of it reaches through the openings in the mesh, and for that reason I increased the threshold for C to avoid false detection in the sample.

  • do you know what the small peak is at about 8.3 keV?


    Here's the unfiltered EDS analysis, The peak at ~8.3 keV is actually a composite of three KBeta emission lines of Nickel. Note the Carbon K-Alpha line showing the backscatter from the carbon mount. The residual Oxygen peak just above that is right down in the noise of the detector. It might be real, but anything less than 1% is considered questionable in this class of EDS system.


  • The residual Oxygen peak just above that is right down in the noise of the detector. It might be real, but anything less than 1% is considered questionable in this class of EDS system.

    Point taken re the oxygen. As indicated in the paper I found, you need really specialised kit to find it in the case of Ni. It is often a mono layer, or there about. I don't think there is going to be anything conclusive about the oxygen, it may or may not be there under the reactor conditions, but if it is, it probably would have been there with TM's results anyway. So if it is relevant, it will be difficult to isolate, involvement wise.


    Thank you on everyone's behalf for that. It is all clues in the puzzle.

  • They do have a finite life, and are also fragile, being made from very thin thorium-oxide coated iridium wire. The system was showing a filament fault so we changed them - it had been moved over 200 miles after getting it, so vibration may have played a part. One looked ok, the other a big deformed but changing them is a delicate task not to be undertaken lightly. Happily there is a YT video showing how it is done- which was way more help than the instructions that came with the new parts.