Talk:Lithium–air battery

Wiki Education Foundation-supported course assignment
This article is or was the subject of a Wiki Education Foundation-supported course assignment. Further details are available on the course page. Student editor(s): Jshi94.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 02:44, 17 January 2022 (UTC)

Major updates
I drastically updated the article based upon a sandbox created by a group of graduate students for a course on Materials Chemistry at a large research university. The previous article had some poorly cited sections and some irrelevant information. The new sections cover any pertinent information that was removed from the past article. Na9234 (talk) 20:43, 13 December 2011 (UTC)

Theoretical and practical energy density difference
Could somebody explain in the article main reasons for difference between theoretical energy density of Lithium (11000 KW-h/Kg) and practical energy density of Li-air fuel cell (1000 KW-h/Kg).Is this only due to presence of electrolyte and weight of a case?But how much of electrolyte is there?!


 * The energy density of lithium only accounts for the mass of pure lithium, nothing else. A practical Li-air cell in its heaviest state (completely discharged) will have its cathode filled with lithium oxide (Li2O) at best, which is only 42% lithium by mass (more likely lithium peroxide, Li2O2, is the cathode product - only 27% lithium). Beyond this, the cell must have a separator, electrolyte, packaging, conducting support in the cathode (lithium oxide and lithium peroxide are believed to be insulators). All told, this extra material will significantly decrease the practical energy density below the theoretical density of lithium oxide and lithium peroxide. --Jrhardin87 (talk) 19:29, 20 July 2011 (UTC)

Expected lifetime?
How long are such batteries expected to last? anything better than the ~500 cycles/-2 to -20% per year of normal lithium ion batteries? 86.43.88.90 (talk) 11:46, 28 January 2010 (UTC)

favorite Li?
While the article clearly favours Li-O type battery, presumably based on the statements of the scientists, I wonder whether Ca-O might in fact be better overall: -The voltage is even higher -There is more Ca on this planet than Li (which competitively may be used for nuclear fusion) -The weight increase when used in mobile applications is less prominent (the oxygen is incorporated into the cell upon reacting with the Li, and adds to the battery weight) Thyl Engelhardt213.70.217.172 (talk) 06:38, 21 October 2010 (UTC)
 * Cite it and write it. Who's working on it?  What is published? We don't do original research here. Thre's nothing about them in the 2002 "Handbook of Batteries". (It's not like Li used in fusion power plants is going to be a major drain on the resource any time in our lifetimes...fusion has been "40 years away" for the last 50 years, and has produced nothing but thesis topics.) --Wtshymanski (talk) 13:19, 21 October 2010 (UTC)
 * This is the Li-Air battery page, not Ca-Air or metal-air. It doesn't make sense to discuss the advantages of CaO here. Should the Metal-air_electrochemical_cell orphan be promoted to Metal-air_battery (which currently redirects to Zn-air)? There are a number of metal-air battery technologies, and a general overview of the category would be a better place for a comparison of the different types of metal-air batteries. --Jrhardin87 (talk) 19:53, 20 July 2011 (UTC)
 * As indicated in the article, the theoretical specific energy of Ca-Air batteries is much lower than that for Li-Air. The slightly higher voltage is offset by a significant decrease in the capacity (the # of useable electrons per unit mass, i.e. Ah/kg), so that the total energy (which is voltage times capacity) is lower. So while the relative weight gain of a CaO battery on discharge is lower, this is only because the battery must be heavier to store the same amount of energy. --Jrhardin87 (talk) 19:58, 20 July 2011 (UTC)
 * Or not, provided that the shielding of the Li requires more -and hence heavier- material than shielding of the less reactive Ca. I'm just speculating, but I think it is worthwhile considering, if we ever step from science to engineering. Thyl Engelhardt213.70.217.172 (talk) 13:56, 29 February 2012 (UTC)

Can we make the article more accesable to the lay person plase?
This is a good article, but it's very technical in a lot of places. For the uninitiated, it's hard to understand. If someone could rewrite it in less technical language that a lay person could understand, that would be appreciated. —Preceding unsigned comment added by Allthenamesarealreadytaken (talk • contribs) 02:50, 13 May 2011 (UTC)
 * Please be more specific: list some or all of the passages you do not understand, or which you believe a layperson would not understand. Then we can focus on specific things that need improvement. One straightforward improvement is to add links on technical jargon. Then the reader can easily look up the definitions of unfamiliar terms. --Teratornis (talk) 17:36, 6 May 2012 (UTC)

Puzzle about reaction
I'm sure there's a good answer to my question, but it's not in the article and will puzzle many readers. It sounds like you're getting something for nothing, and also that there's contradictory information in the article. This may be just a misreading, but we need more information to understand this. The article talks about a 4Li + O2 → 2Li2O reaction, but also says no oxygen is stored in the battery. OK, none at the beginning, but does the reaction mean that oxygen in the form of Li2O *is* stored in the battery as it runs? In which case the battery would get heavier over time. Or, if this is somehow just a temporary reaction and the oxygen gets released again, then it sounds like we're getting something for nothing. The latter seems to contradict physics; the former seems to partially counteract the claim that the battery is so much vastly lighter than ordinary Li batteries if it in fact gets heavier as it runs (and does recharging the battery then essentially mean taking the oxygen out again?) Maybe both of these inferences are false; but then we need a more complete explanation of what's going on to be useful to readers.69.211.4.10 (talk) 15:10, 9 June 2011 (UTC)
 * It's just like a zinc-air battery; shiny metal turns to crud, and electricity is produced. The real trick would be turning the crud back into metal, but that's been very difficult even with zinc and so far has eluded researchers workign with lithium. --Wtshymanski (talk) 15:33, 9 June 2011 (UTC)
 * Yes, the lithium-air battery would have to gain weight as it discharges, much as iron gains weight as it rusts. Lithium-air batteries would still be lighter on average than batteries that carry their own oxidizers at all times, because even with all else being equal, the batteries in a vehicle would spend most of their time being only partially discharged, and this could be minimized with frequent recharging. But all else is not equal; Li-ion batteries use oxidizer chemicals that are heavier than pure oxygen.
 * Recharging the battery would produce oxygen gas, which would either discharge to the atmosphere, or could possibly be collected and used separately or sold as a commercial product. See Oxygen - the annual market for oxygen extracted from air is 100 million tonnes. Perhaps fleet users of battery-powered delivery vehicles would recharge them at central garages, producing enough oxygen to be worth collecting. Disclaimer: I have no idea whether collecting the oxygen from recharging Li-air batteries would be economical, but I'm sure someone would have thought to try. --Teratornis (talk) 18:13, 6 May 2012 (UTC)

Avoid expressions that date quickly
See WP:DATED. The article contains some references to time that violate this guideline (such as "currently"). --Teratornis (talk) 18:17, 6 May 2012 (UTC)

The energy density of gasoline ..
Article: "The energy density of gasoline is approximately 13 kW·h/kg, which corresponds to 1.7 kW·h/kg of energy provided to the wheels after losses" - A mere 13% of the energy being delivered to the wheels is an astounding inefficiency so, although gasoline is not the topic of the article, links and references to this datum and idea would be useful. 2.98.251.179 (talk) 15:57, 4 July 2013 (UTC)
 * The relation to the commonly known energy container gasoline (I would prefere diesel or cerosine) is very useful and spectacular. I use to do so for batteries or hydrogene storage. But even if there is a reference for it, 13% total efficiency was state of the art for cars ("to the wheels") in the 1930s or for present U.S.-American tanks. Probably the reference is obsolet or not competent for combustion engienes. For present German cars (BMW, VW, Audi, Mercedes, Porsche) is a total fuel efficiency of 40% the minimal standard. It is not mentioned what is included in the "losses" but for 13% the energy expense for construction, maintainance and wrecking of the car seem to be in (by theories of some ecologic activists). Calculated that way even solar cells, wind generators and nuclear power plants are energy sinks...--46.115.87.169 (talk) 07:41, 14 July 2013 (UTC)
 * It's also a rather odd comparison to make since Gasoline is a fuel that you can pour into your car when it runs out. Batteries have to be charged when you run out, much less convenient even if the efficiency of the battery is high. 82.32.172.40 (talk) 15:44, 5 November 2015 (UTC)
 * Not that odd, if you think about it. Lithium-type batteries are being looked at more and more to replace combustion engines, so comparing the relative energy densities between the two is a good and useful contrast. However, it's important to get good sources when making such comparisons, so as to avoid WP:BIAS. A ir ♠  C ombat  What'sup, dog? 16:03, 26 February 2016 (UTC)

Unclear wording in Challenges section
The last sentence of the second paragraph of the "Electrochemical" subheading of the "Challenges" section is not a sentence: "The mechanism of improvement, but may alter the structure of the oxide deposits.[31][32]" I would propose a change but I'm not entirely certain what the point of this sentence is. Any thoughts? Eddill (talk) 17:29, 31 December 2013 (UTC)

Check the wording in the Operation section
Quote: "lithium is oxidized at the anode forming lithium ions and electrons". Why does it say "oxidized at the anode"? Surely "forming lithium ions and electrons" means "ionisation" not "oxidisation". From what I understood, lithium atoms are "ionised" at the anode, then travel to cathode to be combined with oxygen (thus being oxidised at the cathode). — Preceding unsigned comment added by 69.67.167.78 (talk) 19:17, 31 July 2014 (UTC)


 * Just to clarify this electrode situation: cathode is always negative, anode is always positive. Lithium or other metals in a battery are called cations, they wander to the cathode during electrolysis, or only interact as the negative cathode during battery operation, and oxygen/oxide and chlorine/chloride or even sulfate or hydroxide are called anions, and they only interact with the positive anode, during electrolysis/charging or battery function/discharge.


 * In normal battery discharging and electric energy providing operation lithium metal, a strong reducing agent, is oxidized at the cathode to positive lithium ions or cations via electrons it gives off, or pressurizes to large negative voltage, that get sent to the other electrode, while oxygen, a strong oxidizer, is reduced at its electrode, the anode, into negative oxygen ions or anions by the electrons it receives from the other electrode, or via the positive suction it provides to these electrons at its positive anode.


 * (All voltages being positive electron suction or negative electron pressure vs. a reference voltage, just like altitude up a mountain side is in reference to sea level or a town down the valley, but there is no such thing as absolute height nor absolute voltage. Sea level and the hydrogen electrode are usual references, even if not the most practical in most measurements, and the Calomel electrode or saturated silver chloride/potassium chloride electrodes are usually the practical reference electrodes in pH meters or practical electrochemistry, at reference voltage different from the standard 0.0000.. V of the platinized platinum surface gas bubbled hydrogen electrode.)


 * In battery charging or electrolysis operations the reverse process happens: electric energy is fed into the system, and converted into stored chemical energy. The reactions of the cations and anions still happen at their same electrode, cathode or anode, but they are reverse. In charging, lithium, a strong reducing agent, is this time reduced by even stronger reducing agent electricity electrons received from a windmill dynamo or solar cell into zero valent neutral elemental lithium metal from the positive lithium dissolved cations in solution at the cathode. (These dissolved positive cations do nothing at the positive anode, as they are already at max charge, however if it were possible to create a Li++ two valent ion with ease, then there could be interaction of oxidation, or receiving an electron from Li++ to Li+, and you could call Li++ a positive cation with respect to the more negative Li+ anion. But such things, which may be possible in a bend over backwards theoretical way, do not happen in practice due to the enormous energy requirements that would rip all water and the glass container molecules apart way before they do occur.) So to get back, in the reverse, chemical energy storing, electrolysis or battery charging operation, oxygen, a strong oxidizer, is in turn itself oxidized by an even stronger oxidizer positive charge, or electron suction force received at the anode, and turns into a neutral, zero valent, elemental oxygen gas molecule. Sillybilly (talk) 16:33, 7 January 2015 (UTC)


 * To recapitulate, anode is always positive, cathode is always negative. Metals like lithium or iron are reducing agents that are cations, always at the cathode whether discharging or charging, whether reduced or oxidized, while nonmetals like oxygen or chlorine are oxidizing agents that are anions. It's not that complicated.


 * Well, you can have silver or gold as metals that are very noble and oxidize very difficultly, and themselves can be used as oxidizers for lithium or iron, on a relative basis, but in their compounds like silver oxide or gold oxide, they still carry a positive charge, and are still called cations. I have not heard of such a thing as lithium goldide where gold would act as an anion carrying a negative charge that interacts with a positive anode. Semimetals can vary, such as phophorous, arsenic, antimony and bismuth, which have metallic tendencies, and nonmetalic ones, and phosphorous can be an anion in a phosphide like Li3P with a -3 anion valence charge, or cation in P2O3, with a + 3 cation valence charge, or P2O5, with a +5 cation valence charge, which is up for debate in case of phosphorous oxide on covalent bond shared electrong arguments, but less debatable for bismuth oxide Bi2O3, where the bonds are more polarized and more ionic in nature. Sillybilly (talk) 16:33, 7 January 2015 (UTC)


 * And while we're at it, this is what I came back here for in the first place, after reading this article last night (the way I'm posting it should turn it into a new talk section):

Mixed aqueous/aprotic
The aqueous/aprotic or mixed Li-air battery design attempts to unite advantages of the aprotic and aqueous battery designs. The common feature of hybrid designs is a two-part (one part aqueous and one part aprotic) electrolyte connected by a lithium-conducting membrane. The lithium metal anode abuts the aprotic side of the electrolyte while the porous cathode is in contact with the aqueous side. A lithium-conducting ceramic is typically employed as the membrane joining the two electrolytes.

I started writing this as only the first sentence following this paragraph, and might have included it on the main page like that, but then it turned into this long rambling, and I don't really know where to put it but here on the talk page, it does not really belong on Slashdot or a blog, maybe a blog or even Slashdot, but it was reading Wikipedia that incited me to write this, not a Slashdot article, when it would be proper to put it on Slashdot, or a random webpage or a book, when it would be proper in a blog. So writing this here in a sense is returning the favor to Wikipedia. There should be room for it on this talk page, and here it is provided via the GFDL automatically, while in other places it's ambiguous. Editors are free to remove it, but then it would still be stored in the history of edits. I wrote the following with good intentions, unlike the naughty millions of edits Wikipedia has to put up with from anonymous sources. It's not really encyclopedic content, but in effect all encyclopedia articles reflect some collective opinion of the authors and references about the status quo of the world, as there is no such thing as absolutely certain scientific knowledge. So the following could be used as a guideline in hunting down references in shaping the article to be more educational. Or even creating properly executed experimental references, as theoretical ramblings are mere opinions, but data, which is also an opinion somewhat, but data collected in an attempt of inquiring nature for answers independent of personal opinions or wishes, is king in science.

As a lithium ion conducting membrane is most likely also an excellent proton (i.e. hydrogen ion) conductor, and the aprotic (or proton free, or hydrogen ion free) solvent must contain counter anions (such as hexafluorophosphate) that solvate protons just as well as lithium ions, so in general only highly alkaline, i.e. (low in proton concentration) aqueous chemistries should be feasible in such mixed electrolyte systems. This issue does not arise to such an extent in the purely aqueous design, as there the lithium conducting membrane is in direct contact with lithium metal (or lithium intercalated graphite), in absence of anions that would also solvate protons, and thereby generate hydrogen gas. Hydrogen gas inside the solid membrane or even inside the graphite intercalation compound would be or is under extreme pressure therefore the reaction is driven backwards due to the Le Chatelier principles, however such pressures could be destructive to the materials involved, unless they are very strong. A possibility to mitigate the high pressure might be the formation of the hydride, where the hydrogen skips the intermediate zero oxidation gaseous molecular state (i.e. it only exists temporarily as nonmolecular, surface bound 0-valent atoms), and becomes an anion itself, contributing to extra energy storage density in a solid state.

Such proton reactivity and hydrogen gassing issues make the purely aprotic-only lithium chemistries as the only economically successful ones to date, even if conductivities in aprotic solvents are undesirably low. The key in aqueous chemistries, whether mixed aprotic or purely aqueous, is the membrane plus behind the membrane cathode material selectivity to lithium vs. protons. For instance mercury metal readily forms a strong metallic amalgam with lithium with full rejection of protons and high overpotential to hydrogen gas generation. See http://www.osti.gov/scitech/servlets/purl/132649 pdf file from 1995 (such technology is pretty much the Castner-Kellner process at its core from the 1890's, and even if there are improvement patents to it, the article is from 1995 where the 20 years patent time is about up anyway. Moreover none of this technology is radically novel compared to the 1890's world. And this discussion I'm writing goes down under the GFDL inside Wikipedia Talk pages) As mercury, especially its organomercury derivatives are a strong environmental issue, other zero or near zero oxidation state lithium solvents (i.e. materials that mobilize lithium from a solid bulk) that fully reject hydrogen should be found. This ratio of lithium-philicity vs. proton-phobicity could be a key to a successful and high conductivity aqueous lithium battery. So far graphite intercalation compounds neither provide high mobility, nor good proton rejection (carbon rejects chlorine well in carbochlorination reactions, but does not do the same with hydrogen). Using an aprotic solvent again has issues of low selectivity for lithium vs. proton in most present chemistries, but this could be a very good area to research (such as crown ethers, or soft anions that create a near metallic bond in a very low ORP (i.e. highly reducing) solvent environment that still somehow forbids hydrogen gas generation, or simply combination with protons. Also the so far used ionic membranes themselves are not very conductive, therefore they must be ultra thin, (as in, if the total resistance of the cell is 2 ohms, while if it were fully the membrane material it would be 200 megaohms, a 20 nanometer thick membrane might pose insignificant addition of resistance to the 2 ohms, but a 20 micrometer one might be dominant, plus if only 2nm tunneling effects might dominate all other considerations, and in tunneling single particle protons much outdo lithium ions which are the next smallest ions in ordinary conditions, and this tunneling ease or proton hopping ease is also the reason why proton based solvents such as water so outdo other solvents in conductivity or simlply ionic reaction rates, and why water is so important to the processes of life) so fragile and prone to defects and leaking, and cannot withstand great pressures that would simply allow eliminating the need of a mobilizing solvent behind them on the cathode side, and just simply use an extreme spring loaded pressurization that makes the soft lithium metal flow and maintain contact with the membrane that way (however a mechanically strong (say metal or diamond or nanotube) sieve on the aqueous and reactive non-bottleneck side of the membrane might be able to hold back such pressures just fine.) A fully ionized (i.e. higher than so negative, i.e. neutral ORP, or no longer strongly reducing) lithium in contact with water requires the solid membrane materials to have highly anionic charges, such as silicates do, to this date, but which are also very brittle and fragile due to the nonyielding and nonslipping and therefore nontough nature of ionic salt bonds. High rate batteries such as NiCd or Pb-acid have such direct metal to electrolyte contact zones and the transition from electronic to ionic conduction happens through an extremely thin and highly tunneling possible zone, and because of lack of such direct metal to highly conductive aqueous contact zone in lithium batteries, lithium batteries cannot be used in high drain rate requiring applications, such as car engine starts or power tools where lead acid and nicads still dominate, despite the low energy storage density by weight. Mercury metal is able to be in direct contact with liquid water, with highly mobile lithium content which though is kept at a fairly low concentration near the direct contact surface, plus, even with reduced zero-valent hydrogen atoms on the mercury surface molecular combination and gas bubble initiation and formation is difficult in a lithium-mercury amalgam (unlike in the case of the magnesium amalgam, where the available excess driving force is sufficient to surpass the required Arrhenius activation energy) and only causes 1% or less Coulombic efficiency (amperage) loss via hydrogen gassing during electrolysis in the above article. Part of the reason of the instability of magnesium amalgam vs. alkali metal or even calcium each of which are possible to electrolyze into a stable amalgam from aqueous solution, unlike magnesium, which is impossible due to an unstable amalgam and fast gassing, is the heat of formation, or combination of the alkali metals with mercury (where the mercury (very near the noble end of the activity series of metals or oxidizing end of eletrode oxidation-reduction potential charts) acts like a soft, metallic bonding electron rich weak anion to form a lithium-mercuride like compound) compared to not much happening with magnesium to form a similar magnesium-mercuride, to where the ratio of reactivity with water vs. reactivity with mercury is not good enough. (Another possible explanation of the difference might be that magnesium might provide some hydride type coordination and ease of gassing and molecule formation not available with the other amalgams, and it is that why a magnesium amalgam is unstable, not because of lack of good heat of combination between mercury and magnesium - I don't think anyone knows this chemistry answer in the present day. Of course mercury used to be used even in zinc batteries such as alkaline or Leclanche ammonium chloride, and cadmium possibly even lead acid, for reasons of decreasing hydrogen gassing, so it's not simply the heat of combination of say zn-hg or cd-hg that matters, but the ratio of such heat compared to drive to react with protons in water. Both Zn and Cd have high hydrogen overpotential themselves, and so does lead, while iron and manganese and aluminum not so, which is why batteries are based on these exotic zn, cd, pb cathodes compared to ubiquitous iron (which is behind zn or more noble than it in the activity series but has weak hydrogen overpotential), or even alumium, manganese or magnesium, which of course are more active and less noble than zn, manganese being the top metal still possible to electrolyze directly from an aqueous solution in absence of an amalgam surface, while in presence of an amalgam surface all metals are possible, including sodium in the Castner-Kellner process, or calcium in the Humphry Davy process of Ca metal preparation from 1805.) So a similar to mercury material could be sought, such as a gallium-telluride eutectic, where the gallium provides the low temperature liquidity and mobility, while the tellurium the somewhat anionic like mercury (or even silver, gold, platinum (with a caveat at Pt, Pd, Ru, Ni and the like which dissolve hydrogen or are non-hydrogen-phobic in their regular crystal structure, dissolved in an alloy they might behave differently, or if a lithium hydride is sought, they might be a storage reservoir for atomic zero valent hydrogen and useful), etc), but not too anionic along the lines of silicates or hexafluorophosphate or even the molten iodide stripper from the above article, where there is some heat of formation, but not too great, and the bonding is more of a metallic, or electron-rich and soft as opposed to hard and polar and ionic. If the heat of formation, or bonding, as the telluride of lithium is too strong, possibly antimonide would lessen it, or even bismuthide (polonium or astatine are not readily available nor desirable for toxicity reasons), or possibly stannide or plumbide. There could possibly be various combinations also instead of just pure elements, similar to the situation such as silicon and GaAs and CdTe all are great semiconductors similar somewhat to elemental silicon and germanium but not exactly similar in every respect, so if a magical material is sought that duplicates the behavior of mercury, or even improves upon it, and elements are unable to match it, combinations such as molybdenum diselenide or ruthenium diselenide might give a similar result when interacting with lithium metal in a metallic solution as say telluride or mercury would, where the selenide is more anionic compared to tellurium, but not enough so for the lithium metal to fully go after it and attack molybdenum or ruthenium out of its combination with it, in a replacement fashion based on the activity series of metals, especially if molybdenum has rich enough d-orbital bonds and increased metallic-covalent bonding to compensate for its weaker reducing electric potential, when in combination of selenium or tellurium. Also, just like a LiI-CsI eutectic is used in the above article, a potassium telluride or bismuthide or plumbide mix (or sodium, calcium, barium, strontium, rubidium, cerium, etc) might be more ionic in nature where lithium would prefer going after the soft anion metal, but the other metals trump it or near equal it in the activity series of metals, or in reducing potential, while still providing a modification to the properties of the metalloid soft anion material. The above article uses iodide which is both low melting (of 225C compared to higher for fluorides, chlorides and bromides), and insoluble in metallic mercury (where a telluride might act more like a simple metallic alloy and be highly soluble in liquid mercury, or at least Li2Te have a high saturation concentration compared to LiI, and even higher solubility for Li3Sb, or Li4Sn or Li4Ge, but less for Li3As.) There are articles about electromigration of metals in alloys even in the solid state, as in under huge amperages there are tendencies of some metallic elements in an alloy to migrate(with normal diffusion counteracting) toward the positive electrode, while others to the negative one, on the order of ppm scale after weeks or months of such treatments, but understanding and charting such behavior might be worthwile even possibly for understanding superconductivity behaviours (as one of the best superconductors used in hospital equipment is Nb3Sn, and another recently found superconductor is MgB2, both having a feel of some intermetallic compound at the right combinations.)

Basically what's sought is a behavior similar to mercury amalgamation, that strikes a yin-yang balance in stability of metallic combination with lithium, i.e. heat of formation, and rejection of any reaction with protons or simply just blocking hydrogen gas formation via a high hydrogen overpotential surface (which can be helped by acetylenic or other electron rich gassing suppression compounds used in acid pickling of rusty steel before electroplating, hogging active surface sites), all the while maintaining high conductivity throughout a battery, both electronic through the metal, and solvated ions in the aqueous or other highly conductive and ionic organic or inorganic molten salt solution. The lithium dissolving material does not have to be very liquid, in effect instead of a sharp liquid solid transition in viscosity as it happens when gallium melts, a softening and a gradual transition might be more optimal, where the lithium mobility is still very high, but the electrode maintains mechanical stability with minimal effort from an ultra thin silicate ionic membrane, or just a perforate sieve-cage. A perforated sieve cage might hold straight metallic lithium under extreme pressure with a liquid mercury membrane at zero pressure on the outside of it. Because of surface tension the mercury should not want to enter through the tiny holes into the lithium metal bulk, other than through some vacuum vapor phase saturation, and in the absence of a huge pressure inside the perforated cylinder reaction or battery action would stop, but if a huge pressure is applied, the soft lithium metal would squeeze through the tiny holes and dissolve into the mercury and react. This way you could have an internal on-off switch to the battery, and extreme lifetimes of a battery could be achieved, together with high reaction rates, as needed.

There are two types of high electrical conductivity near room temperature - ionic and electronic. In a lithium battery there is a necessity of transition from the metallic electronic lithium conduction to the ionic dissolved ion lithium conduction. And it is lithium itself that has to do the positive ionic conducting, because, unlike lead, which is barely above hydrogen in the activity series of metals, and it's like a self amalgam and does not gas much hydrogen even in 38%wt sulfuric acid in a lead metal/lead dioxide sulfuric acid battery, where it might be possible to insert a proton conducting Nafion membrane just fine, and have the protons do the conducting just fine (which in fact is what happens in a lead acid battery, proton conduction, as both lead sulfate and lead dioxide are mostly insoluble and remain at their respective electrodes), in a lithium battery lithium metal in contact with protons means localized reaction and gassing and loss of electric energy and locally turned into waste heat. There are of course exceptions to this line of thought, that it is necessary for lithium to do the ionic conducting part, for instance, potassium could be doing the conducting in a lithium battery, and lithium is unable to liberate it from solution, and because of the smaller sphere of hydration, potassium does have a higher conductivity than lithium does, in hydrogen bonded water solutions at least, while in aprotic solvents the situation might be opposite. However solubility and saturation come into play, and one might want to leave as much room in solution as possible for the lithium, as long as he has decent conductivity to begin with, as possible. (Unless a method similar to lead acid is sought, where it's not bulk dissolution but insoluble storage near the electrodes is sought.) The other method of conducting through the membrane, without using protons, is via negative charges. But it is proton tunneling that is responsible for high ionic conductivities near room temperature in in the best ionic solvents (with the exception of some room temperature molten salt like waxy-quaternary-amine superacid counterion solvents where there is no proton-tunneling-hopping, but the ion concentration is fully 100% compared to 0.0000001=10^-7 mol/liter proton concentration in distilled water at neutral pH 7, where the full ionization compesates for the lack of proton tunneling.) So sacrificing proton tunneling comes at a grave price (including water being of low price), but it does not have to be a full sacrifice. For instance, even though the hydroxide ion in say a potassium hydroxide water solution has a high conductivity, it does it via hydrogen hops, not oxygen motion, however the 2nd hydrogen in a hydroxide is so strongly bound that, as far as I know, during molten LiOH electrolysis the Li metal does not react with LiOH to form Li2O and H2, but instead the hydrogen generation happens through water formation at the anone from the 2OH-> H2O+O reaction, and the water migrating back into contact with the metallic lithium, and if a fully protected anolyte section can be created through a ceramic membrane where the water accumulates to concentrations that leat it easily turn into steam and be vented off, straight LiOH, or eutectics of it would actually a great molten salt electrolysis medium for Li metal (even for such simple things as Li3N ammonia generation from atmospheric air.) So even the most conductive anion in water is hydroxide, far surpassing other anions, and the small fluoride has similar bulky hydrogen bonded sphere of hydration issues compared to iodide or bromide as lithium does compared to potassium or rubidium, and in aqueous electrolytes it is very difficult to get away from proton conduction and get an anion conduction through a membrane such as oxygen or fluoride migrating through a silicate, which are insurmountably difficult compared to proton traveling through, or lithium.

So once again membranes are difficult for good battery applications (or even for fuel cell applications, where the low ionic conductivity membrane itself in micron thickness is the ionic conduction forcing electrolyte that blocks electronic conduction short circuiting, but because of the low conductivity it must be huge surface area or low thickness, both of which increase the chance of pinholes or membrane failure unleashing an avalanche of local reaction and reactants freely getting to each other and locally generating waste heat.) And by membranes we mean solid objects, because traditional batteries themselves are pretty much "membrane" applications of the liquid type, where leaks are not possible because liquids self-fill any holes and cavities within their bodies. Ionic conducting liquid membranes of low electronic conductivity type, where the reactants, such as zinc and manganese dioxide or lead and lead dioxide do not react at room temperature, without the aid of a dissolving agent, which this way is both a catalyst for the reaction, and the ionic conducting medium lacking electronic conductivity. In lithium solid membrane batteries there are issues related to size and leak failures due to scale, issues related to conductivity and reaction rate in high drain and high power application, requiring large volumes for a given wattage or peak amp rate compared to lead or cadmium cathode batteries in direct metallic/electrolyte contact, with a transition layer of less than a nanometer at the metal surface to ionic electrolyte, and easy metal almost quantum hop tunneling into the bulk solution and very low ohmic resistance from that. In this sense a bare metal in naked contact with the bulk ionic conducting solution is much preferable to a solid ionic membrane, other than a general loose structure salt-bridge like such as a ceramic flower pot (but not really ionic conduction only forcing solid membranes such as Nafion fluorosulfate) that create pH-like gradients, or concentration gradients such as water to be distilled out from a molten LiOH anolyte.

There might be such lithium dissolving substances that keep lithium metal near the zero valence state, or near enough it in a semi-ionic combination such as lithium-antimonide (or could be called stibnide), where the reduction potential is still maintained near -3V for lithium but the high overpotential surface blocks hydrogen gassing, or the potential is reduced to near 2V or whatever this voltage is in alkaline(low proton concentration solution where the Nernst equation changes the decomposition equilibrium voltage based on molarity; here's a better reference Pourbaix_diagram) solutions where water does not break up spontaneously, and the low molecular weight of lithium still provides high energy density, or a combination of these factors to provide 2.5 V at the cathode, and then of course you are free to add on top of that at the cathode, oxygen from the air being preferable, but the Co(III)-Co(IV) oxidation transition at higher oxidizing power than Au gold also being available, as presently practiced, true that both atmospheric oxygen and cobaltite crystal have reaction rate issues, however the atmospheric oxygen part could be helped via ferric chloride or cupric chloride catalyst circuit board etch mechanisms, or similar ways, both of which though require high proton concentration (similar to how permanganate, chromate or hypochlorite or hypobromite are more oxidizing in acidic environments and stable in basic environments) to clear the generated hydroxide ions from the O+H2O->2 OH reaction, so there may be a need for a strong pH gradient of low pH high acidity near the air electrode, and low acidity high alkalinity and high pH near the lithium electrode in the fully charged state, and, here we go again, possibly a proton conducting membrane that keeps these pH domains isolated, but in this case it would need to be fully proton only conducting. A very high pH domain near the lithium electrode still keeps the LiOH soluble, and lithium has this advantage vs. beryllium, the top chemical in energy storage capacity when the recyclable nonexhausted ash weight is concerned, and lithium is 4th down on the list from BeO, LiF, BeF2, Li2O in top storage density as shown in this comment, and hydrocarbons or hydronitrogens only outdo these in storage density because their atmospheric reaction product oxide is freely exhaustible as a gas, and in that, hydronitrogens are also carbon-neutral, and somewhat easier to recycle from the atmosphere from the ubiquitous nitrogen that's difficult to fixate compared to relatively scarce atmospheric carbon. Of course there could be a magic organic molten salt like aprotic beryllium/air solvent that leaves aqueous chemistries based on hydrogen hopping-tunneling in the dust, and do it at room temperature, and provide top chemical energy density storage with minimum ancillary equipment from bulk reserves of beryllium metal and oxygen from the air, similar to a flow battery where the working btattery part is small and the storage reservoirs huge. Such things would be top rechargeable energy storage devices for say space applications, but where there is an atmosphere, like on Earth or even inside a space station, atmospheric recyclable gaseous ash or gaseous oxide yielding hydronitrogens and hydrocarbons are still king, and the next best things are of course nuclear processes, for which we don't presently save small scale nore regeneratable/chargeable devices, and the lack of such technology in the age of haters and terrorism is a blessing.Sillybilly (talk) 16:57, 7 January 2015 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Lithium–air battery. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20150924061043/http://www.ibm.com/smarterplanet/us/en/smart_grid/article/battery500.html?lnk=ibmhpcs2%2Fsmarter_planet%2Fenergy%2Farticle%2Fbattery_500 to http://www.ibm.com/smarterplanet/us/en/smart_grid/article/battery500.html?lnk=ibmhpcs2%2Fsmarter_planet%2Fenergy%2Farticle%2Fbattery_500

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 17:17, 3 January 2018 (UTC)

Suggest: section "Potential Market"
Howard from NYC (talk) 22:04, 2 October 2022 (UTC)

anyone else consider it useful to include a section "Potential Market"? Provided in order to provide context why this variant on battery technology is worthy of greater attention. Keep in mind journalists and investors always start their research by digging into public and/or free sources of knowledge.