MFMP: Automated experiment with Ni-LiAlH

  • Good morning. I have uploaded the final file set through the cooling of the reactor to ambient, and the corresponding gamma spectra files. The experiment appears to have been completely null (no LENR). Its repeatability against the calibration was very good. If there was 1W of excess heat, it would have been detected with some confidence.

    can ,

    If you get a chance to add a final plot through the cool-down, that would be excellent. I really appreciate your participation! Regarding the pressure loss ... It is hard to discern the difference between a leak and the behavior of a reversible hydride. The LiAlD4 is not reversible, but the LiD into which it decomposes, is, I believe, a reversible hydride.

    Next up... in about a week I should have another experiment underway. I intend to etch the carbonyl Ni powder in ultrasonic HCl, rinse in de-oxygenated water, and perhaps freeze fracture before drying heated in a vacuum. The goal is to expose chemically clean Ni surface area. It will be mixed 10% with LiAlH4.

  • Gas "pressure" in condensed matter is an interesting quantity. While the gas is monatomic and interstitial, it is more of an alloy and there is not any real gas pressure. When a vacancy in the lattice is filled with more than one atom of hydrogen, then there is "gas" and gas "pressure". The pressure is related to how tightly the orbitals of the molecular form of the gas have been compressed to fit into the vacancy. Historically this was related to explosions of steam boilers. The cast iron boiler would have large bubble-like vacancies in the metal. Some monatomic H would enter from the steam into the iron lattice and would dissolve into the metal. When the monatomic H found a physical vacancy in the iron, it would stay there and accumulate with other H atoms forming H2 molecules. The pressure inside these vacancies in the iron could become >100k PSI and would eventually cause so much internal stress that the boiler would just explode under the combined stress of the internal vacancies and the pressure from the applied steam.

    I think you may be using the term 'vacancy' incorrectly here. A vacancy in a lattice is an atom missing from the lattice that can be occupied by a range of things e.g. a similar size and valance metal ion. H ions normally occupy interstitial sites, not vacancies, which is why the loading factor approaches 1:1. H2 gas does not normally collect around a vacancy to cause hydrogen embrittlement. That is caused by H2 gas collecting in cracks, most usually grain boundaries.

  • Since this thread is very much on the topic of Ni/H experiments, I attach part of an email received today from Alexander Parkhomov, containing some useful information on his current approach to fuel treatment and 'ignition'. Bob and others may find the content of interest.

    Dear Alan,......... I use a nickel powder, made by carbonyl technology with a granule size of a few microns. I do not apply any additional treatment prior to loading into the reactor.

    Successful experiments were carried out both with a mixture of nickel with LiAlH4 and with pure nickel saturated with hydrogen gas. Pumping out air with a fore-vacuum pump, filling with hydrogen to a pressure close to atmospheric, holding at a temperature of 200-300 oC for several hours to purify nickel from the oxide, pumping out to remove the formed water, refilling with hydrogen, saturation of nickel with hydrogen at a temperature of 300-400 oC for several hours and after that a smooth increase temperature to 1200-1300 оС. After the release of excess heat, it may be necessary to reduce the power of the electric heating. The pressure in the reactor chamber is kept close to atmospheric pressure........Alexander Parkhomov

  • BobHiggins

    I would try to focus on the reversible Li-hydride reaction. When the hydride decomposes hydrogen gets released in an atomic form and likely also in a reactive, short lived excited or ionized state. There have been interesting results in past MFMP experiments while playing with that reaction, so this could be one reason for exploring it out.

    According to the literature LiH forms quickly around 550°C and decomposes completely above 950°C, at atmospheric pressure. The idea would be cycling temperatures through these two thresholds, taking into account that internal temperature will be somewhat higher than the reported tube temperature (I'm assuming).

    I would try to not use excessively high temperatures right away, for example the initial cycling range may be 500-800°C. After a certain amount of time, this could be gradually increased for example to 550-950°C, then perhaps 600-1200°C or above. The point here is maximizing at all times H uptake and release into/from the Li but also to not immediately sinter the Ni powder.

    At a lower pressure it will take a lower temperature to decompose the LiH to a large extent, so a possible idea could be staying well below 1 bar, perhaps even as low as 100-200 mbar or less when all the H is decomposed. This will also have the advantage of volatilizing the lithium earlier. Clearly, the previously mentioned temperature cycling ranges depend on the target maximum pressure. According to the graph below, for LiH to decompose at roughly 800°C the maximum pressure must be no higher than about 200 mbar.


    As a side note, Dufour et al some time ago have claimed small amounts of excess heat from a Fe-Na-H system. Their explanation here could be an incentive to try experimenting lower pressures:


    [...] For the reaction to occur, electrons must be present in the reacting medium, where hydrogen is adsorbed on the metal (transition metals, like iron, adsorb hydrogen).Electrons are available when the vapor of an alkaline metal like sodium or lithium is present in the reacting medium and if the temperature is sufficiently high.

  • barty

    Yes, Parkhomov has never really stopped, though he moves at a slow pace- mainly because of funding issues. He is working with a 'casual' group of helpers these days I think. I will ask him some follow-up questions to address your specific enquiries.

  • can ,

    Adding in the lower temperature cycling would be easy to add to the protocol. I will think about what you have sent. I could pick the pressure.


    Yes, regarding AP's preparation being very little. He does grind the LiAlH4 and Ni powder together in air in a mortar and pestle. But, it is pretty cavalier. The in-situ hydrogen processing is something MFMP has done in all their experiments with no success. I DO have some of AP's Ni powder. Perhaps the thing to do is to try using HIS Ni powder. Either the next experiment or the one after, I will do this.

  • BobHiggins

    I've thought of easy, non-intrusive changes to the protocol only. Did you have something else in mind when you asked for protocol suggestions?

    Parkhomov's Ni powder and LiAlH4 were used for the MFMP GlowStick GS4:…log/502-glowstick-gs4-run


    The fuel will be Ni and LiAlH4, 10:1 by mass. Both powders were supplied by Alexander Parkhomov.

    EDIT: by the way, do you use or plan using new alumina tubes for every new experiment, or do you reuse old ones?

  • can ,

    All suggestions are welcome. Changes to protocol are generally easy.

    I use a new alumina tube for each fueled run. It is not practical to remove the fuel powder. What I do is when I am epoxying the brass fitting adapter to the new tube (cut to 7.5" length), I cutoff the tube with the fuel, and epoxy a seal filling the open end of the tube and mark the experiment date on it. If, at some time in the future, it makes sense to analyze the fuel, the epoxy seal will have helped keep it fresh. I have to cut through the brass tube adapter of the previous experiment to salvage the 1/4" Swagelok nut. A new brass adapter is made for each experiment and I must use new ferrules. Making the brass adapter is only 30 minutes worth of work on the lathe with a 3/8" brass rod starting stock.

    Also, if it makes sense, I can run another calibration with the new empty tube before filling it with fuel.

  • BobHiggins

    The reasoning behind asking if you were going to use a new tube for every experiment was that it's conceivable that at a low enough pressure and high temperatures the lithium previously reacted with the tube could start "boiling off" from the inner surfaces and permeate the inner atmosphere to some extent, even if no elemental lithium is initially present in the fuel powder for the ongoing experiment. So, if the previously highlighted hypothesis by Dufour et al is true, a "used" tube may be more "active" than a new one.

    Actually I think this is how the electron-donor alkali substrate that I cited several pages ago from one of Piantelli's patents would be working. It would probably be more efficient when heated (it's shown as a separate component, so why not?). In general, Dufour's hypothesis where the alkali vapors would provide electrons to the hydrogen atoms adsorbed on the transition metal and supposedly ionize and/or excite them does not seem too different from what Piantelli also says in his patent(s) on the subject.

    If this is how it works, then a used tube could possibly be able to show excess heat by itself just with hydrogen gas, as its inner walls will also contain transition metal impurities from the previous experiment. It's admittedly an unlikely hypothesis on top of another, but how would one be able to discern in this case if a previously used tube showing an apparent gain even when "empty" is really just showing an artifact?

    I find worth noting that for all intents and purposes Leif Holmlid's Fe2O3-K catalysts are broadly speaking supposed to operate in a similar manner. When heated, the volatile potassium alkali content of the catalyst forms a sort of "cloud" that can excite the hydrogen atoms dissociated on the surface of the catalyst and in its close vicinity. When the hydrogen atoms desorb from the catalyst, for example with a flow of hydrogen through it, they easily can get involved in the reaction Holmlid observes (which also requires a transition metal surface).

    I realize that the hypothesis where the reaction requires alkali atoms to be present in the environment in the form of a low density gas is far from your liquid Li-Al layer hypothesis; actually I think such a layer would work against it, in addition of quickly smoothing away any high surface treatment the underlying metal may have initially come up with. You might indeed want a little bit of lithium there so that when the hydride decomposes it can provide excited atomic hydrogen gas (and eventually, also gaseous alkali atoms) directly close to the surface, but not so much that the transition metal surface gets completely covered and the small surface features eventually destroyed.

    In this context it might therefore be desirable to add lithium (or other alkali metal hydrides) somewhat away from the reaction area, in order to enhance the gettering-release of hydrogen during the temperature cycling described in a previous comment and in general allowing more gaseous alkali atoms to be present in the atmosphere at higher temperatures, but without directly (not immediately, at least) affecting/covering the catalytic powder you chose to use.

    This is my 2c.

  • Alan Smith

    The idea for the proposed protocol (still involving some form of Ni and Li as required) is to at least initially use much lower pressures, perhaps something like 100 mbar maximum. The main reason for this is reducing the decomposition temperature of LiH, but the side effect is that the Li will start evaporating earlier. At this pressure the boiling point of Li is 1064°C. My source is this page:…f_the_elements_(data_page) (CRC data).

    Preferably no obstacles would be present between its source and the fuel, like for example felt plugs as used in the recent "Cravens" experiment (lithium would react with them). On the other hand, since according to the previous explanation bringing in contact large amounts of liquid lithium with the fuel would not be desirable, something would have to be conceived to keep it confined in starting location before it starts evaporating.

    The recent email from Parkhomov you cited doesn't seem to describe something much different than what has been tried so far by other people, but there's nothing about fuel proportions and the mixing procedure, as well as how it's inserted into the reactor. I don't expect illuminating information on that aspect, but perhaps it could be worth asking him more details about it. I recall that ages ago it was speculated that he mixed Ni and LiAlH4 with some other impurities, whether inadvertently or not. I also remember that he himself initially also thought that using a lower amount of LiAlH4 relatively to the rest of the powder could have been beneficial.

    I do think that perhaps it could be an idea to mix a bit less Li/LiAlH4 together with the Ni, while somewhat increasing its total amount in the initial charge by doing as described in the previous comments.

    For what it's worth, the MFMP GS5.2 and GS5.3, which apparently showed odd anomalies during the temperature/pressure cyclings, used 80% Ni (1 gram), 12% LiAlH4 and 8% Nano Li, but they were all mixed together.

  • Part of my reasoning about the Li-Al-H coating the Ni particles comes from observation of the ash from the Lugano experiment, Parkhomov's experiment, and the !Bang experiment. All showed this wetting of the Li-Al-H to the surface area of the Ni powder under SEM analysis. Note also that Li will not wet to the Ni by itself - it must be a Li-Al alloy. By the time the Li-Al alloy is molten, any nano-Li would have long since been molten and would just alter the liquid Li-Al atomic ratio. The nano-Li is just an easier way to handle Li into the fuel than pure Li metal.

    Another thing is the use of metal cans for housing the fuel inside the alumina tube. There will be deposits on the alumina tube as the liquid Li-Al-H leaks out (I have heard it described as a foamy liquid) of the can. Initially, it was conjectured that this eliminated the possibility that the alumina was complicit in the reaction, but then we discovered that the liquid Li-Al-H leaked out of the can onto the alumina and actually had a few dissolved atomic % of Ni in the film. I can't conclude anything from this.

    Predicting just what will work and what won't is really difficult without a theory having some grain of correlation to existing results. Because of this, Cravens advises, "run the experiment". One of the things we need to get better at doing is running more frequent experiments, turning different knobs. I am hoping automation will get me performing more experiments.

  • BobHiggins

    I can offer an additional anecdotal point of view on the subject.

    me356, whose latest reactor is going to be tested/validated soon by MFMP members, has claimed a few times that the Al in LiAlH4 has a kind of quenching effect on the reaction, which can also work purely with Ni and Li. I can bring you references if needed.

    This information, coupled with the fact that as you're saying Al is necessary to obtain this coating on the Ni (also given that intuitively speaking, unlike Li it won't evaporate away with temperature) makes me inclined to think that the Li-Al coating has the opposite effect on the reaction than assumed so far.

    Indeed if what is important according to others is that alkali metal vapors can get in contact with the hydrogen atoms adsorbed on the transition metal, then you don't want any coating there.

  • Bob Higgins: THX a lot for Your time and effort. We seem finally to come to a point, where we can be more and more sure, that, at least in this approach, LENR simply does not exist. Good work, besides the results, but we need to start somewhere, and falsifying something is as valid as proving it. Keep on.

  • In reference to my previous comment, here are some selected me356 posts where he suggests that the Al in LiAlH4 throttles back the reaction, which apparently works more aggressively using only Li. He also suggests here and elsewhere that the boiling temperature of lithium must be reached for the reaction to start and that it can be useful to decrease pressure so that this happens. This seems to go against the idea that a coating has to be present on the particles. Merely doing this is not enough however; he's often stressed that that the powder has to be treated for high surface area (he writes to "improve hydrogenation") and that some sort of triggering/impulsive action must be present, but he's always left out some critical details in the description, something that many have lamented.

    Since he reports to be quite successful and that he has a line of private communication with some MFMP members (who, again, will test one of his reactors soon) I find puzzling that much of what he's suggested is not being applied in the experiments. On the other hand he seems to have often tried/applied what others have suggested to MFMP.…n/?postID=16705#post16705


    To achieve excess heat even with lower temperatures it is important to keep pressure as low as possible (less than 1 milibar of hydrogen) with no impurities, Clean nickel and chamber as much as possible with hydrogen and vacuum pump even at high temperatures before triggering the reaction. Boiling point of lithium must be achieved to start Cat. Remember that boiling point is function of the pressure.

    LiAlH4 is not necessary (only pure Li is fine), but it will rather make reaction more stable (not that agressive) because of Al.

    Remember that the reaction can go out of control in just 1 second, temperature can increase by hundreds °C suddenly.

    me356: Reactor parameters [part 1]

    me356: Reactor parameters [part 1]


    I recommend so that the pressure is around 1 bar absolute - without external control it is not too easy. Because of safety and reactor integrity it is good to keep pressure low. Then you have to decrease it before triggering excess heat.

    Normally the problem is, that LiAlH4 will give you too much hydrogen, on the other hand you have not too much Lithium for the reaction. So you have to wait very long time if you have no additional pressure control until the pressure is low enough - can take hours - weeks if it is a wrong ratio.

    To allow reversible reaction of LAH that will give you good "control" you need more Lithium - this is very dependent on the composition of LAH.

    Alternatively you will lack aluminium if the reaction is too strong if too much of pure lithium was needed to add (but I guess that this is acceptable for the first tests).

    I recommend to not use LAH at all for the first tests if you have an external pressure control.

    me356: Reactor parameters [part 2]


    To make it work, you will need Lithium in vicinity or in a direct contact with Nickel. If LiAlH4 is used only for supplying hydrogen, you can use whatever else, but the reaction will be just Ni-H. Yes, removing aluminium will make the reaction stronger.

    As a side note, Max Temple on E-Cat World has done a good job collecting many (but not all) of his most interesting posts, even providing referencing links (the initial version didn't, I found this out yesterday):…356-taught-us-max-temple/

    On the other hand some time ago for my own convenience I have organized several of his more recent question-answers (à la Rossi) in a more readable format. It's sort of "raw" and not exactly publication-worthy, but I can still provide it if needed (see attached document).

    I would have preferred to not resort to the "me356 says", but if information from conventional researchers (Piantelli included) is not a convincing enough argument against the liquid LiAl coating on the particles, perhaps what this person is suggesting (both directly and indirectly) is.

  • can ,

    Thanks for the interesting suggestion and collected tips from ME356. While some Al is desirable for gettering the oxygen, it is not clear, as you point out that the 1:1 Li-Al ratio is preferred. Parkhomov surmised that LiAlH4 was used from the Lugano analyses, but none of those measurements indicate a 1:1 ratio of Li-Al. Measuring the Li is hard and assaying the ratio is hard. So, the Lugano reactor may well have had a different Li-Al ratio than 1:1 and the 1:1 only really comes from Parkhomov. This is test-able. Perhaps the easiest way to change the Li-Al ratio is to include LiH in addition to the LiAlH4. I can set the pressure low to basically whatever I want using my programmable back pressure regulator. Without Al, the Li will not wet to the Ni (only an observation).

    It is also pretty hard to get to high enough temperatures to have Li vapor when the heat is being supplied with a heater coil. If, instead, the heat is being supplied as a plasma discharge, the temperatures can easily reach the Li boiling point (almost guaranteed to) while remaining below the Al boiling point (2700C). Additionally, plasma discharge tubes can have a hot center plasma even while the tube itself has its envelope cooled with water, relieving the problem of the apparatus melting. This begins to sound like the case of the dusty plasma of Suhas or even (I hate to say it) Rossi's QuarkX.

    I am working on a plasma apparatus now for a next round of experiments (with water cooling and flow calorimetry). In the mean time, I will see if I can acquire some LiH or other Li. I will have to check... I may have a small sample of the nano-encapsulated Li.