Boron: A Better Energy Carrier than Hydrogen?

... while with boron the easy part is the handling. If drivers send slabs of (even less ignitable!) boria across the Earth, and get back boron, both forms' manageability helps this energy tap be efficient. But no cars seem to use it yet.

Contents

Return to mass and volume table 

The very large volume for ash from liquid hydrocarbon, over 100 cubic metres, is mostly due to gaseous carbon dioxide. The calculation is based on its density at zero Celsius, 0.001977 kg/L. If the problem of where hydrocarbon for many billions of cars will come from in the first place is solved, the release by those cars of many billions of tonnes of this gas presents another difficulty. It seems likely to have global consequences.

But its release has immediate local consequences that we seldom think about. We are very accustomed to hydrocarbon-derived limits on our thinking about cars, limits that hydrogen- or boron-burning cars do not impose. (Other limits still exist, of which more will be said a little later.) What if everyone in a crowded shopping mall remained in his or her car? Not as many people could fit, evidently, but maybe these are very small cars. If they burn boron they don't pollute the indoor environment at all. If they burn pure hydrogen the mall's air conditioning may need to dehumidify the air.

This situation doesn't arise with today's carbon-burning cars because everyone in the mall would get very short of breath, and flee or die of suffocation, long before anyone would be bothered by humidity. This is true regardless of how well tuned and clean burning the hydrocarbon cars are, because carbon dioxide is the cause of the trouble.

This chemical is also naturally produced by human life processes, and we absolutely need to get rid of it. If there is too much in the air, so that blood carbon dioxide has no concentration downgrade to escape along, it doesn't matter how much oxygen there is. If a hydrocarbon combustor in perfect working order, that just means all of the carbon it is exhausting is in the form of dioxide. Malfunctioning cars can emit carbon monoxide, a strong poison, but that is not the main reason why today's cars can't come in the house.

If they could, would that a good thing? Difficult to say, what with thinking limited by existing structures, but if we try it for 20 years or so and then consider going back, maybe some benefits will come to mind, or indeed maybe no benefit will have turned up and the cars, after a brief foray, will have returned to their almost exclusively outdoor habits.

An Insidious Hazard of Carbon-Free Combustibles

If one tries to keep a submarine warm with a charcoal fire, and the carbon dioxide that would otherwise be an immediate hazard is made harmless by being captured in alkali, eventually there will be a shortage of oxygen. This is the above mentioned limit that exists even when combustion is carbon free. Infringing on it is dangerous because animals don't naturally detect oxygen-deficient air the way they do carbon dioxide excess.

This hazard is a remote one because consuming all or most of the oxygen in an enclosed space produces a huge amount of heat, and it would be very unusual for the enclosure to be able to lose the heat without also exchanging the oxygen-depleted air for fresh. E.g. consider a house trailer sized room 2.5 metres by three metres by five metres. It contains 45 kilograms of air and 10 kg oxygen at sea level. In the depth of winter someone is heating it from inside by burning a carbon-free fuel, and the outside is so cold and the trailer is so thermally leaky that it takes 10 kW of heat to keep it warm.

The trailer's occupant gets 17.87 megajoules per kilogram of oxygen if he combines it with hydrogen, so a hydrogen burner will take half the oxygen in 2.5 hours.

The unlikely thing, unless the trailer really is hitched to a submarine, is that it could leak all that heat without being well ventilated. In fact this trailer would be unsafe in the summer too, because just breathing in it would raise the carbon dioxide concentration from 0.03 percent to 0.2 percent in 2.5 hours.

When oxygen is combined with boron the heat yield per kilogram of oxygen is 26.53 MJ so in that case, for a 50 percent depletion to exist, the trailer must still be on the same air after 3.7 hours.

An Oxygen Extractor that Quits before You Do?

It's possible that boron's unignitable nature will make a total difference, and make that slight extra time irrelevant. While hydrogen can be ignited very easily, even in air with low oxygen, a designed system is needed to ignite boron.

Part of the necessary commitment is extracting pure oxygen from air. Air gives up oxygen not directly to a boron flame, but to an oxygen extractor, and if that device can be so made that it will take plenty of oxygen when there is plenty, and none at all if the oxygen concentration is getting down anywhere near a level that would inconvenience air-breathing life, then dealing with the anoxia hazard will have been folded into a design task that had to be done anyway.

Return to mass and volume table 

Return to main text  

Gibbs free energy data were derived from enthalpy and entropy data in the American NIST web accessible database:

W.G. Mallard and P.J. Linstrom, Eds.,
NIST Chemistry WebBook,
NIST Standard Reference Database Number 69, November 1998,
National Institute of Standards and Technology,
Gaithersburg MD, 20899
(http://webbook.nist.gov).

Return to main text 

Boria is a glass-forming oxide. Tables of physical properties list both a vitreous (glassy) form and a crystalline form that is quite a bit denser and minutely lower in energy. Because of this energy difference, if you burn boron and the ash you get is droplets of the glassy form, your final energy yield is about 99 percent what you would get if it went all the way to crystalline B₂O₃.

Energy yield calculations in this document are based on burning to glass. The one percent is not entirely lost because when the ash gets back to the central power plant for recycling, it still is glassy, and this means the one percent is still in there. So the energy cost of recycling is less. Not a whole percent less though.

If boria crystals are heated to 450±2 Celsius they definitely melt. But if the liquid is cooled down through this temperature, it supercools, and if it is taken on down to room temperature, the supercooling is locked in. This is glass.

When it is reheated, nothing special happens at 450 Celsius. The temperature range between 440 Celsius and 460 Celsius could be considered the softening range, but this depends on how soft is soft. Metre-long centimetre-thin rods supported in the middle might slump into a horseshoe shape in a few days at 300 Celsius, while visitors from the planet of the tungsten people might find the liquid at 1,500 Celsius still too viscous for swimming. (These estimates are not based on any calculation, experiment, or actual consultation with the tungsten people.)

Return to main text 

Return to mass and volume table 

Based on hydrogen density 0.07078 kilograms per litre. This is the density of liquid hydrogen at its normal boiling point, 20.28 K, minus 252.87 Celsius.

Return to main text 

3,310 Celsius for methane, 3,370 Celsius for hydrogen.

These flame temperatures were predicted with EQS4WIN, a program from Mathtrek Systems. It can also predict the compositions of the flames. Here is what it says boron and four-thirds the oxygen it can combine with should turn into at constant 100-bar pressure:

2B(s) + 2O2(g), 298.15 K ---> 0.476B2O3(g) + 
                             0.293O2(g) + 
                             0.013B2O2(g) + 
                             0.375BO(g)  + 
                             0.647OBO(g)  + 
                             0.291O(g),  4,640 K

4,640 K is the adiabatic flame temperature, i.e., the temperature in a flame that has no way to lose heat and therefore is as hot as can be. If heat loss were possible then the molecular fragments shown in the last four lines would sort themselves out into the diboron trioxide and leftover oxygen shown in the first two.

This is what is predicted for methane.

CH4(g) + 8/3O2(g), 298.15 K ---> 1.687H2O(g) + 
                                0.688CO2(g) + 
                                0.729O2(g) + 
                                0.312CO(g) + 
                                0.391OH + 
                                0.003OOH + 
                                0.103O + 
                                0.094H2(g) + 
                                0.044H , 3,580 K

Return to main text 

Return to main text 

A solid lump of boron is not a practical fuel configuration, but if it were, it would be four times smaller in volume, a golf ball next to a tennis ball, compared to a hydrocarbon tank to deliver the same energy. Boron's combustion energy density is the highest there is.

This is almost true of its atom density also. Among solid forms of pure elements only diamond has more atoms per cubic centimetre, and unlike boron, diamond is stable only under high pressure.

Boron's mass density is not exceptional. The beta-rhombohedral crystalline form has density 2.34 kg/L and that is the basis of the energy-specific volume figures used here. WebElements gives a density of 2.46 kg/L but that is the density of alpha-rhombohedral boron. Neither form is quite as dense as aluminum, 2.70 kg/L, but since the average boron atom is only 0.40 as heavy as an aluminum that isn't surprising.

When boron is referred to as red, as at http://www.speclab.com/elements/boron.htm, alpha-rhombohedral is the kind being referred to. But boron fibre as it now-a-days comes off the chemical vapour deposition line is beta-rhombohedral and, as shown in the Introduction, dark brown.

Return to main text 

Return to main text 

A slug of dense liquid fuel can completely fill a tank and then flow from it through a narrow channel into a combustion chamber. This can seem like a very odd attribute of liquids if one thinks about it. Imagine if an alien world were covered with solid life forms filled with liquid travelling through narrow channels ...

A crystal the shape of a tank's interior cannot do likewise, but a long filament can. So for the solid substances the theoretical density, 2.34 kg/L in boron's case, was multiplied by two pi over (five root three) to approximate the effective deliverable density from a cylinder partly filled with coiled filament. Two pi over the product of five and root three is near 0.725520. So boron's motor-deliverable density is counted as 1.6977 kg/L.

Why two pi over (five root three)? This is the product of three volume fractions.

• Eight-ninths, if a coil of many concentric cylindrical layers has the innermost layer at one-third the radius of the outermost one. The other ninth is the cylindrical void surrounded by the innermost layer. It might not be a void during the coil's winding, it might be a spool.

• Pi over root 12. The filament is circular in cross section. If, after the first layer of windings has been wound, each new loop in every subsequent layer is laid into the groove between two adjacent ones in the layer below, then the circular cross-sections will be in a hexagonally close packed array, and fill pi over root 12 of the cross section plane. That is near 0.906900. The three-cornered spaces between the fibres take up the other 0.093100.

• Nine over ten. The fibre itself is counted as being only 90 percent as dense as a perfect crystal.

Evidently this is a reasonable fuel form but not the only reasonable one, so the exact fraction in a given implementation could be different, maybe better.

The same density reduction, i.e. 1.38-fold volume increase, was applied to all the solid fuels. Solid ashes were considered as being made into round rods of as much density as might readily be obtained on a vehicle. For crystalline oxides, this was two-thirds of the X-ray crystalline densities tabulated under "X-Ray Crystallographic Data of Minerals", p. B-186, CRC Handbook of Chemistry and Physics, 60th edition, CRC Press 1980. This table did not provide data for lithia or phosphorus (V) oxide, so its entry in "Physical Constants of Inorganic Compounds", p. B-92, was used.

Only one listed oxide, boria, is purely unable to crystallize when rapidly condensed from a flame. (Or maybe two, but silica was treated as condensing to nanocrystalline quartz.) Boria was counted as forming round glass rods of density 1.812 kg/L. This is 71 percent, not two-thirds, of its WebElements crystalline density in stablest form.

These reduced densities were further multiplied by the packing fraction of round rods that are as long as their container: pi over root 12, near 0.9069.

Return to main text 

Return to main text 

Pulling thread off a skein and forcing it into a pressurized chamber may seem like pushing a rope uphill. However in a heat engine there will be cool working fluid being forced into this chamber, and if the channels it enters by surround the thread's entry path, and it flows faster than the thread, it can drag the thread in.

Or at least, that's one method that logically ought to work. The task of pulling the thread up the pressure gradient turns out to be very small in terms of energy requirement, because the volume per unit time of thread that must enter the chamber is small, because boron has a high energy density.

E.g. suppose the chamber needs to produce 16.7 kilowatts of thermal power. This yields a boron volume flow rate as follows,

(16700 J/s) / (57.97 MJ/kg) / (2,106 kg/m³),
which is 0.000288 kg/s / (2,106 kg/m³),
which is 0.000000137 m³/s

Now if the pressure in the chamber is 100 bars, 10 megapascals (MPa), and outside is one bar, 100 kilopascals, then the rate of addition of pressure-volume energy in the thread is 0.000000137 m³/s times 9,900,000 pascals, and that is 1.35 watts. In a perfectly efficient system that would also be the required power. Since several kilowatts of the 16.7 kW (thermal) will become mechanical power that can be tapped for this, even with some losses it will still be a small fraction of output.

Return to main text 

Return to mass and volume table 

Condensing and storing 237 litres of water over a 1,000 km trip could be a real awkwardness for hydrogen-powered vehicles. But despite the scoffing of men in the street, experts have concluded that just letting it go, as vapour, is a pretty good solution.

In fact it is what already happens to water from hydrocarbon combustion. Hydrocarbon combustion produces less water, only 93 litres (as liquid) in this case, but even so, the Earth annually gains a cubic kilometre of water this way, enough to raise sea level by a tiny fraction of a millimetre.

Return to main text 

Based on the density from CRC tables of the glassy form of the oxide, 1.812 kilograms per litre. This is the only form of boria that matters in the context of using boron and boria as, respectively, energy supply and demand carriers.

At room temperature and pressure the stablest form of boria is a crystalline one with a density of 2.46 kg/L (CRC60, p. B-62) or 2.55  kg/L (WebElements). Despite its stability it is hard to make.

Another, still denser crystalline form that is not stable is reported in the Kirk-Othmer article.

Return to main text 

It's interesting to consider that if it were absolutely required, a gasoline-powered car could retain its carbon dioxide. There would have to be a large heat exchanger ("radiator") to get nearly all the combustion heat out of it, and a sturdy pressure tank.

One pound of octane gives 3.08208 pounds of carbon dioxide. In the 1,000-kilometre example, 65.7 kg fuel gives 202.5 kg carbon dioxide. On a hot day (30 Celsius) this can be kept liquid by a pressure of 7.164 megapascals, just over a thousand psi, and its volume then is 338 litres, 3.7 times the original fuel volume.

Return to main text 

Gasoline has various reported densities ranging from 0.695 kg/L to 0.75 kg/L. Its lower free energy yield, which is the yield if the water is produced as vapour, is between 44 and 46 megajoules per kilogram.

For the eight-carbon alkane known formally as 3-ethyl-3-methylpentane the data are 0.7274 kg/L and 45.66 MJ/kg. Since these are precisely known and fairly well centred in gasoline's range of variation this octane was taken as a fair representative of the whole commodity.

Return to mass and volume table 

As a basis for comparison, the energy to propel a car 1,000 km at 125 km/hr is set at an arbitrary round value of 3.00 billion joules in this table. If the joules are from gasoline, the car is getting 26 miles per US gallon.

Return to main text 

Actually 150 litres is less than a quarter the bulk of a three-gigajoule, thousand car kilometre liquid hydrogen tank. 373.8 litres is the volume of the liquid hydrogen itself, but that is only the start. The vessel around it roughly doubles the volume and triples the mass.

The volume of a hydrogen tank must include some wall thickness to keep the rate of inward heat leakage low. Too many watts coming in means not only rapid loss of fuel to evaporation but a hot catalytic oxidizer at the end of the relief line. At least one would hope it means that. This is an aspect of automotive hydrogen tank design rarely discussed but a catalytic oxidizer seems like an obvious and necessary part of it.

A 373.8-litre volume, if spherical, is 0.8937 metres in diameter. That is the inner diameter of a spherical hydrogen tank for a 1,000-kilometre-range car. Suppose the walls are 0.1 metres thick. Their volume then adds another 311 litres, because the volume of a sphere 1.0937 metres in diameter is 685 litres.

What are those walls made of? Layers of vacuum alternating with thermal radiation barrier layers, preferably reflective.

There must be structural strength to support atmospheric pressure from outside, plus strength to resist whatever pressure the hydrogen inside is allowed to exert before it begins to exit the relief valve. This is not necessarily a difficult structural problem, not unless it also is demanded that average density of the tank wall not greatly exceed liquid hydrogen's. If in fact the wall averages a balsa wood-like 0.2 kg/L then the tank, to contain 26.5 kg hydrogen, must on its own mass 62 kg.

This is why the apparently very small mass of the hydrogen ends up being part of a massive system. (Still more so for metal hydrides.) In fact with the above estimates it seems a 1,000-kilometre hydrogen tank masses 88 kg full, more than either a gasoline tank or a spool of boron.

Return to main text 

Return to main text 

The forms natural boria is found in are borates, combinations with other oxides, e.g. borax Na₂B₄O₇·10H₂O.

Return to main text 

This World Health Organization document's Section 1.2, Production, uses, environmental fate, and sources of exposure, gives a good summary of how much boron is handled today, including these surprising facts:

Boron enters the environment mainly through the weathering of rocks, boric acid volatilization from seawater, and volcanic activity. Boron is also released from anthropogenic sources to a lesser extent. Anthropogenic sources include agricultural, refuse, and fuel wood burning, power generation using coal and oil ...

Dissolved oxidized boron is vital to plants including trees, and they contain quite a few parts per million of it, so if you burn them, you are going to get volatilized borates. It seems possible that despite boron's not being a household word, the smell of borate is an essential part of the smell of wood smoke, and as such has been part of human experience for many thousands of generations.

The document also gives an indication of the order of magnitude of today's boron extraction industry:

Economic borate deposits are rare, occurring in arid regions of Turkey, the USA, Argentina, Chile, Russia, China, and Peru. Total world production of boron minerals ... was approximately 2 750 000 tonnes in 1994. About 800 000 tonnes of commercial borate products, expressed as B2O3, were manufactured from the boron minerals.

Return to main text 

Return to main text 

One would probably prefer to have stored boron under some kind of cover, but it isn't necessary! A coil of boron filament left out in the rain will not lose its goodness. According to A. B. Berg in the 1970 Encyclopædia Britannica, "Boron",

Boron is chemically almost inert at ordinary temperatures, although the fine powder is ignited by elementary fluorine or slowly converted to boric acid by the action of hot nitric acid

It seems likely this is the same Berg as the one mentioned in UCLA professor M. Frederick Hawthorne's lecture, FROM MUMMIES TO ROCKETS AND ON TO CANCER THERAPY:

... General Electric Company undertook a research and development program in Malta, N.Y., which actually made commercial quantities, several hundred pounds of diborane, pentaborane and decaborane. These more stable boron hydrides are the contributions of Alfred Stock.

All went well until about 1948, when the plant was cleaned and washed out with carbon tetrachloride, not realizing that carbon tetrachloride and decaborane form an explosive mixture equivalent to nitroglycerin. The plant was blown away. So, live and learn. Anton Berg at USC had warned them that this might happen. They did it nonetheless.

Several hundred pounds is enough for one rather small rocket. Possibly "tons" is what was meant.

WebElements at the University of Sheffield, U.K., gives similar evidence that elemental boron is a very different thing from decaborane:

Crystalline boron is inert chemically and is resistant to attack by boiling HF or HCl. When finely divided it is attacked slowly by hot concentrated nitric acid.

Note the one extra word, hot concentrated nitric acid. Both sources agree that the boron must be in very small pieces, so that it has a lot of surface area. This may not apply to the reaction with fluorine, because boron trifluoride is gaseous and does not seal off a boron surface the way the oxide does.

Neither elemental fluorine nor hot conc. nitric acid are ever found in Nature. Therefore artifacts of massive elemental boron left outside will last a very long time.

This inertness suggests that at normal environmental temperatures, unlike aluminum, boron does not have the oxide coat that prevents it from catching fire when hot. At room temperature naked aluminum reacts instantly with air or water. This can be confirmed by filing aluminum under water. The newly exposed metallic surface takes oxygen and leaves hydrogen, which appears as minute bubbles. But while aluminum is metallic, boron is not. Like graphite it may have just a layer of adsorbed air.

Supporting this notion is the fact that lumps of glassy boria start out as transparent as ordinary soda lime glass but quickly acquire a milky appearance through the growth of a thin superficial layer of boric acid. Boric acid dissolves in water, so if boron grew a coat of oxide at normal temperatures, water could attack it in two steps, and it would not be as inert as it is.

Return to main text 

Return to main text 

Helium as working fluid in a heat engine has very simple behaviour. Its heat capacity is always the same, it always is a gas and never changes state, and it doesn't react with anything. But however pleasant it might be in car motors it can't reasonably be expected to be available for them and is mentioned here only as a simple example.

Return to main text 

Flowing along a turbine blade is something boria can do and other oxides with high specific binding energies cannot. Boria condenses at a very high temperature, well above 2,000 Celsius, but remains able to flow down to much more moderate temperatures.

Here is a rough calculation showing that even though liquid boria has very high viscosity, centrifugal force in a turbine motor is abundantly strong enough to move it. Viscosity and density data are from "Boron Oxides, Boric Acid, and Borates", Kirk-Othmer Encylopedia of Chemical Technology, 4th edition.

Viscosity is a measure of the force in a fluid undergoing shear that is proportional to shearing area and also to the speed gradient. That gradient is in metres per second per metre, which is just inverse seconds. This means a viscosity times an area times a speed gradient, in inverse seconds, gives a force. The units of viscosity itself must therefore be units of force, over units of area, over inverse seconds. This simplifies to pressure units times time units, e.g. pascal-seconds. Thus:

viscosity * m²/s --> force in N,
therefore viscosity --> N / m² * s --> Pa·s

If we have a tension ring-supported turbine blade turning 500 times per second whose inner tip is at radius 0.02 m, angular speed being 3,142 Hz, the tip's centripetal acceleration is 197,000 m/s². If adequate flow exists here it will also exist farther out where the acceleration is greater. Suppose the boria coating is 100 microns thick, and its temperature is 1,000 Celsius. What shear rate must exist right next to the blade?

At 1,000 Celsius boria's density is 1.523 kg/L so in the centrifugal 'g' field it weighs 300.0 kN/L. 300 million newtons per cubic metre. Times the supposed 100-micron layer thickness gives the shear stress, 30 kPa. Viscosity is 7.4 Pa·s, seven thousand times the viscosity of water at room temperature.

shear stress / viscosity = speed gradient
30 kPa / (7.4 Pa·s) = 1,015/s

Ten microns out from the blade, if that speed gradient were constant (it isn't), the fluid would be flowing 0.010 m/s.

Actually at ten microns the weight that causes the shear is only that of the remaining 90 microns, and it turns out that this reduces the speed there to only 0.0385 m/s. Taking into account both the increasing flow speed farther from the surface, and the diminishing shear rate, the average speed is ...

Density * Acceleration / (3 * Viscosity) * (Layer thickness)²

The layer is 100 microns thick so the average speed in this instance is 0.135 m/s.

Since all this is happening 0.02 m from the turbine axis, the area of the outward-flowing film is 100 microns times the circumference, at that radius, of all the turbine blades. Suppose their circumferences sum to just twice the circumference of the 2 cm-radius circle. That makes 0.25 m. The total stream area then is 0.000025 m². This times the speed, 0.135 m/s, makes a flow of 0.0000034 cubic metres per second. 3.4 cubic centimetres per second.

This represents 88 kilowatts of boron combustion thermal power, enough to keep a large car moving quickly, and 100 microns is not a very thick coating. Therefore, Q.E.D., the oxide flow is sufficient.

Compare silicon dioxide, the other one of the big two glass-forming oxides. At 1,500 Celsius vitreous silica still is more than a billion times as viscous as water. Its metastable liquid state becomes absolutely stable -- the crystalline form melts -- at 1,723 Celsius rather than 450 Celsius. Getting liquid silicon dioxide to flow along a turbine blade would require that blade to be well above 1,723 Celsius.

What about non-glass-forming strongly bound oxides? Their tendency, once they have condensed from their formative flame as liquid, is to freeze to crystals at a temperature not much less. E.g. aluminum oxide, which condenses at around 3,000 Celsius, roughly a kilokelvin higher than boria, and freezes hard at 2,054 Celsius (WebElements), 1,600 K higher than boria's theoretical freezing point.

Once an aluminum flame was diluted with working fluid down to a temperature a turbine blade could take, it would be a suspension of corundum crystals, not droplets. These would not stick to the turbine blade surface and run out along it. Rather, each one would take a little bite and soon the blade would be a stump.

Return to main text 

Return to mass and volume table 

Methanol, also known as wood alcohol, is included at the suggestion of a reader. It burns cleanly, yielding, per thousand car kilometres, 192.7 kg carbon dioxide. Compare the 202.5 kg produced by equivalent 3-ethyl-3-methylpentane.

Methanol's mass density is 0.7914 kg/L as liquid at 20 Celsius. Its specific energy is 21.38 MJ/kg and therefore its energy density is 16.92 MJ/L. This means 1.96 times as much volume as equivalent 3-ethyl-3-methylpentane. Not as bad as hydrogen, which has 4.14 times the volume of the hydrocarbon.

But while a hydrogen vehicle's fuel reservoir system is more massive and much more complex than a gasoline tank, in part because of hydrogen's superlative bulk, the intrinsic mass of the absolute actual basic hydrogen is superlatively low.

Methanol does not have this virtue-in-principle. It is not only almost twice as bulky as gasoline, it is heavier by a factor exceeding two.

Return to mass and volume table 

Return to main text 

0.014 kg/s divided by density, 2.11 kg/L, means the volume of thread pulled in in one second is 0.0066 litres. If the thread is 0.1 mm in diameter, the length that has that volume is 834 metres. Thicker filament evidently is necessary, or multiple strands. A single strand at 0.5 mm diameter will bring in the right amount of boron if its speed is 33.4 m/s.

Return to main text 

The liquid boria has precipitated out of the working fluid, and is not changing in volume like the working fluid, so it isn't cooling itself directly by that mechanism. It may still be in contact with the fluid. The fluid is going down the centre of the expansion channel, pushing on turbine blades, and the boria is running along the channel walls.

What if the contact isn't enough, and the oxide just stays hot? Interestingly this doesn't mean major inefficiency. When the hot gaseous oxide in the flame mixed with, and condensed in the working fluid, it transferred something like 97 percent of the combustion energy. What's still in the liquid oxide is comparatively tiny.

Return to main text 

As noted in section Boron of the Bright Flame, boria newly formed in a boron-oxygen flame can start out at or above 4,600 K. Elsewhere we have considered a muscle-car case where at certain times 800 kilowatts of heat are needed in the combustion chamber. What kind of flame can broadcast this much power?

Actually only two-thirds of the power can move this way. The other third comes from the heat of condensation of boria to liquid, which occurs at a temperature -- mid-2,000s Kelvin -- that might in some circles be considered warm, but is far too cold for radiation heat transfer. The flame and the other fluid must mix for this final transfer to happen.

But let's approximate freely and see how big a 4.6-kilokelvin flame it takes to emit 533 kW. If it is an ideal blackbody it emits 25 watts per square millimetre. Suppose it is an ellipsoid with radii 'r', 'r', and seven 'r'. Surface area estimated to be 19.32 pi 'r' squared. (One cannot just multiply the sphere's four pi 'r' squared surface area by seven.)

The minor radius 'r' comes out to be 18.7 millimetres. This suggests a boron-oxygen flame 262 mm in length (twice seven times 'r') and 38 mm wide could radiate the required thermal power. This will readily fit in a combustion chamber 400 mm long, and so seems not out of line for a car.

Return to main text 

Return to main text 

Might an oxygen purifier / inert gas excluder also be good for a motor burning something else? It depends on the separability of the ash, and that in turn is associated with density. Boria is very separable from oxygen because it is very much denser, and it is denser because it liquefies. If it remained gaseous it would be only a little denser, not enough to separate.

A hydrogen combustion motor could conceivably have hydrogen for its working fluid, with combustion provided by an inverse flame -- a jet of pure oxygen reacting with the hydrogen around it. Since the hydrogen would be in excess, the oxygen would be totally consumed. So indeed, an oxygen purifier might be put to good use here.

But only if the water vapour coming off the inverse flame would liquefy, and be separable. Liquefying water in an internal combustion motor is difficult. But if it were done, the departing liquid water would not carry much hydrogen out with it, so little hydrogen and no oxygen would be wasted.

Hydrocarbon motors now-a-days depend on nitrogen to be the bulk of their working fluid and to dilute the flame and limit its temperature. This leads to nitrogen oxide formation.

Even so, it's hard to see any gain in switching to pure oxygen. Carbon dioxide cannot be liquefied in a motor, so if the motor is burning a lean fuel-oxygen mixture the departing CO2 will sweep excess oxygen out with it, and waste it. If the mixture is fuel-rich the excess fuel will be wasted, and also partly turned to soot. Finally, a balanced (stoichiometric) mixture will burn very hot, and nothing will be present to dilute it. Hydrocarbon motors would really work better on the neon planet.

Actually there is a possible diluent for a pure oxygen flame burning nearly stoichiometric hydrocarbon fuel: high-pressure carbon dioxide.

The dilution produces a mixture of steam and carbon dioxide. All of the first is fuel ash, as is some of the second. After the mixture has expanded and done some work it can, by cooling, be made to release its water as liquid. Then when it is compressed again a fraction of it equal to the fraction of the carbon dioxide that is newly generated can be liquefied, drained off, and as much carbon dioxide as before can return to the flame.

This is discussed as the "semi-closed CO2/O2 Brayton cycle" option in Bolland, O. and Mathieu, P., 1998, "Comparison of Two CO2 Removal Options in Combined Cycle Power Plants", Energy Conversion and Management, 39 (16-18), pp.1653-1663, also available 10 December 2001 as a PDF file at http://www.tev.ntnu.no/Olav.Bolland/flower97%5Cflower97_b.pdf.

Return to main text 

Return to main text 

Some publications on combustion of particulate boron exist but apparently nothing about flames propagating along fibre and consuming it. But much of what will happen can be predicted.

The high pressure flame is very bright and boron is dark, so as the flame gets near a particular small spot on the boron surface, that spot is strongly heated by light from the flame. Being transparent, the oxide coat can't get heat this way but can get it by conduction from the underlying boron -- except there now is doubt this coat exists.

As the boria gets hotter oxygen is increasingly able to diffuse through it. The oxygen pressure is high on the outer, liquid surface and near zero at the inner solid surface, so the diffusional flow of oxygen is inward. The oxygen pressure at the boron surface is very small because oxygen arriving there is immediately consumed, producing new oxide and much heat.

So two mechanisms, flame light absorption and combustion with diffused oxygen, both produce heat right at the boundary between solid boron and liquid oxide. With the approach of the flame, both mechanisms speed up. Soon the oxide at the surface boils. This lifts off the rest of the oxide coat and exposes the hot boron directly to gaseous oxygen. This is where the flame proper begins.

Nothing has yet created any heat inside the solid body of fibre. It's dark in there and oxygen can't get in. So the only way heat can get in there is by conduction. This means that even when the naked boron fibre is shrouded in flame, and molten on the surface, it still is solid in the middle. It will continue to advance longitudinally within the flame just as, up until the oxide coat peeled off, it was advancing towards it.

With the coat gone, and turbulence caused by vapour expansion, gas rich in oxygen will move across the liquid boron surface, laterally. When the coat was there it sat motionless and oxygen could only diffuse straight in, but now it can blow over the surface like wind blowing over a landscape. As it does so it scours out boron and carries it away as gaseous boria.

Something like weather on a micron distance scale, and fractional microsecond time scale, will occur: a surface current that is consuming boron will heat itself and expand. But in expanding it will push itself away from the fibre, and be less dense where still in contact, so its rate of self-heating will be reduced. The heat recently gained will now be radiated away faster than it can be replaced. So the gas current will shrink. This shrinkage will be the more because each two molecules of gaseous ash have replaced three molecules of oxygen.

When it shrinks, it will do so around its new centre, which is away from the fibre because it pushed itself away. Other gas will be pulled in between it and the fibre.

Each particular surface microgust gets laden with oxide vapour, but overall there is an abundant excess of oxygen, because of the boron motor's ability to retain unconsumed oxygen and bring it around again. So the fibre will be completely consumed and suboxides, such as diboron dioxide OB-BO, will not survive.

All this predicts why a flame, once in being, will eat all the fibre that the fibre it is on leads to. In a gas with a high enough pressure of oxygen. But if cold fibre is fed into a chamber with cold oxygen, or even oxygen that is somewhat hot from compression, nothing is going to happen.

A likely way around this is initially to heat the oxygen that will be nearest the cold boron filament with an electrical discharge. This discharge will heat the oxygen well above boria's boiling point and stand in momentarily for the flame.

Return to main text 

Return to main text 

Boron has chemical symbol B and carbon has symbol C. This is one of only two instances where chemical elements with consecutive atomic numbers -- 5 and 6 in this case -- also have, for their symbols, consecutive letters of the alphabet. The other is the pair nitrogen and oxygen, elements 7 and 8, symbols N and O. Boron, carbon, nitrogen, oxygen, and fluorine appear in the periodic table as five single-letter symbols together, the only such run longer than two: B C N O F.

Moreover, if the parallelism of atomic number and symbol letter begun with B(5) and C(6) is extended three more steps, it reaches F(9) -- and so agrees with the actual ninth element, fluorine. This may be only the outcropping of a rich vein of deep significance ...

The B C N O F series can be seen at the WebElements site, with links from each symbol to information on that element.

Return to main text  

These energy yields are more precisely Gibbs free energy changes between the chemicals before and the chemicals after. So are all the chemical energy measurements in this document that don't specifically refer to heat.

For each of boron, 3-ethyl-3-methylpentane, etc., the contained energy that it and oxygen together lose when reacted at constant pressure -- its enthalpy of combustion, DH, given below -- differs by a few percent from the free energy, DG, given in the main text.

Reagents   Products   DH/ (kJ/mol)   DG/DH   B(s) + 3/4O2 1/2B2O3(gls) -626.68 0.9427 Al(s) + 3/4O2 1/2Al2O3(s) -837.85 0.9443 1/2H2(g) + 1/4O2 1/2H2O(g) -120.91 0.9453 C(s) + O2 CO2(g) -393.52 1.0023 1/26C8H18(l) +25/52O2 9/26H2O(g) +4/13CO2(g) -194.46 1.0318

Exergy measurements show how much energy per unit of fuel is available for conversion to work, without regard to where it originates.

A motor that burns complex molecules, e.g. 3-ethyl-3-methylpentane, to simpler and more numerous ones can take some heat from the environment through the dispersal of those waste molecules. This is reflected in the DG/DH ratio column above. With 3-ethyl-3-methylpentane the total energy available for conversion to work is 1.0318 times the combustion enthalpy.

No actual cooling of the environment occurs. 3.18 joules are pulled in for conversion, but then, out of 103.18 joules, only 20 to 30 become work. The main effect on the outside world is the immediate addition of 73.18 to 83.18 joules.

When a car goes along a level road the 20 J to 30 J of propulsion work also becomes heat again after a very slight delay. All the work the motor does, it does against friction. 100 percent of the (fuel plus oxygen) enthalpy becomes heat in the air near the road, and a relatively tiny amount of preexisting heat that was in that air, and was borrowed, is quickly returned.

The ratio column shows boron, aluminum, and hydrogen all behave oppositely to 3-ethyl-3-methylpentane. Rather than having all their combustion enthalpy convertible and borrowing three percent more on top, they lose almost six convertible percent.

This is because the products are (in water's case) larger molecules than the reagents, or (in the cases of boria and aluminum oxide) solids. Manageability has its price, and for these ashes it's nine percent.

Return to main text  

Return to main text 

The energy density shown for carbon is based on the mass density of graphite, 2.267 kg/L (WebElements). Diamond is 56 percent denser but that only brings its combustion energy density to 84 MJ/L, not quite as good as beryllium and far short of boron. Diamond is slightly unstable with respect to graphite but this adds less than half a percent to its combustion energy.

Return to main text 

One vendor of pyrolytic boron nitride claims that artifacts of this material "can be exposed to oxygen indefinitely at temperatures up to 750°C", which is a dull red heat. So of course can we; we'd burn. If the implication that boron nitride won't is true, it suggests the nitride also grows a thick, flawless coat of liquid boria. Without that, it would burn very well, because boron nitride is much less stable than the oxide.

Return to main text 

Also known as magnesia, and when rocks are made of it they are said to be made of the mineral periclase. It is a very stable solid, m.p. 2,852 Celsius.

Return to main text 

Involatile: having low vapour pressure; tending to remain in a condensed phase, i.e. solid or liquid. The opposite of volatile.

Volatile materials have high vapour pressure. If they are combustible, this makes them easier to ignite. Their vapours mix with air or oxygen. "Volatile" sometimes is used as if it meant "easily ignitable" but it strictly doesn't. Things that cannot burn at all can still be volatile, e.g. water and carbon dioxide, the two ashes of hydrocarbon combustion.

Pure iron melts at 1,535 Celsius and the vapour pressure of boria at that temperature is 29 pascals (0.29 millibars). People who live with winter know that at -32 Celsius ice sublimes very slowly. That is the temperature where the pressure of water vapour is 29 Pa, so liquid boria would evaporate with about the same slowness if it were lying on partly liquid and partly solid iron.

This rule for boria vapour pressure,

ln (P / 1 kPa) = 5.849 - 16960 / T

is given in the 4th edition Kirk-Othmer Encyclopedia of Chemical Technology, in Table 4 of the "Boron oxide and borates" article by Robert B. McBroom of the U.S. Borax Research Corporation. 'T' is the absolute temperature and the range of validity is from 1,331 K to 1,808 K, which is 1,058 Celsius to 1,535 Celsius. 29 Pa is what this equation gives at the top end of the range.

The temperature where boria actually boils, i.e. has vapour pressure equal to 1.01325 bar, 101.325 kPa, is in the range between 2,000 Celsius and 2,500 Celsius, most likely near the top of that range. A value of 1,860 Celsius is often given but is wrong.

The vapour pressure of ice is given in a table in the CRC Handbook of Chemistry and Physics, CRC Press, p. D-195 in the 60th edition.

Return to main text 

Return to main text 

In descending order of combustion energy density solid boron, beryllium, aluminum, carbon, and silicon all have more of it than 3-ethyl-3-methylpentane.

That is not an exhaustive list. As seen in the kilograms per gigajoule chart the show there is really over once the atomic number reaches six, but that isn't true for energy density. Some quite heavy elements that would never make sense as motor fuel, e.g. zirconium, nonetheless have denser combustion energy than liquid hydrocarbons.

Yet sometimes a lack of energy density in alternative fuels, compared to liquid hydrocarbons, is presented almost as a law of Nature:

Petroleum-based fuels (gasoline and diesel) offer excellent energy content by volume. By comparison, alternative fuels have less energy density. (http://www.ece.umr.edu/areas/power/Energy_Course/energy/vehicles.html)

As with all alternative energy storage media, the energy density (W-hr/kg) of liquid nitrogen is relatively low when compared to gasoline ...
(http://www.aa.washington.edu/AERP/CRYOCAR/HomePage/NoFramesIndex.htm)

So what has made boron, etc., unthinkable as alternative fuels?

Perhaps it has been an unstated assumption that pure oxygen could not be supplied to a vehicular combustion chamber. When this is done, complete retention of solid or liquid ashes is possible, and fuels that produce them can be considered. But if other gases besides oxygen are let in, they must be let out again still in gaseous form, and some fraction of a solid or liquid ash stream would escape with them.

As discussed in Internal Boron Combustion Engine Considerations, this can be avoided with pure oxygen because all of it can be incorporated in the ash.

Incidentally the second quote refers to a remarkable investigation where a vehicle, instead of carrying around something that can burn hot and be the hot end of a heat engine, carries around liquid nitrogen for the cold end. All of the heat that gets converted to propulsion energy starts out in the air around the vehicle. In very favorable conditions the system may, per kilogram of liquid nitrogen consumed, convert as much as 0.4 MJ of environmental heat to propulsion energy.

This, and liquid nitrogen's 0.8081 kg/L density, make 57 gallons of liquid nitrogen yield less propulsion than 2.8 gallons of gasoline, and less than one gallon of boron. This is an extreme example. Most alternative fuels don't fall this short.

Return to main text 

Return to main text 

Boria dissolves slowly as boric acid, and much of the world's boron is already in the sea in this form. The boron concentration is variously reported as 4.4 and 4.6 parts per million by mass. Most of the mass of boric acid is hydrogen and oxygen so 4.6 ppm B translates into 26 ppm boric acid. In the whole ocean this means roughly 30 trillion tonnes of it.

A few hundred kilotonnes of scattered boria lumps on the sea bottom would, if washed over by a bottom current, have a plume of elevated boric acid concentration downstream of them, but because the process of dissolving is slow, the elevation would be slight, i.e., the concentration would not be much above the natural level.

Return to main text 

Boron will be sent forth to be fuel again ... and again, and again, and again. Boron can't get bored.

What about the oxygen a deoxidation plant separates from boron? Will it just be let go? Yes. There is some call for pure oxygen, but not conceivably enough for the huge volumes boron deoxidation will make.

The most notable demanders of pure oxygen are the very combustors that the boron will burn in. Since the same boron is used over and over again, it might seem the same oxygen could also be reused, and the combustors could be spared the work of isolating new oxygen from the local air.

But what with oxygen's being a gas that if kept would have to be kept in something, and the hazardous nature of compressed or liquid oxygen, and the large amount involved -- 2.22 kg for each kilogram of boron -- new oxygen turns out to be a lot less trouble.

The best thing to do with the oxygen, really the only thing, is store it in the Earth's atmosphere.

Return to main text 

Return to main text 

Diboron trioxide, B₂O₃.

Other names include boron oxide and boric oxide and anhydrous boric acid (or boric anhydride).

Analogy with silicon dioxide, silica, suggests diboron trioxide should be called bora. This was proposed in 1931 by S.R. Scholes but did not catch on. Four letters may be just too few.

Return to main text 

Is +157 kJ/mol a lot of energy for an energy-absorbing reaction to absorb? To put it in context, here is the sulfur/boria process side-by-side with another reaction that is already industrially important, limestone calcination:

Reagents   Products DG°/(kJ/mol) 1/2B2O3(l) + 9/8S2(g) 1/2B2S3 + 3/4SO2 +156.89 CaCO3(s) CaO(s) + CO2(g) +130.21

When boria and gaseous sulfur react, volume diminishes. Most of the reagent volume is sulfur, and most of the product volume is sulfur dioxide, and the produced gas has only two-thirds the volume of the consumed one. (Sulfur is gaseous in this context because it boils at 444.674 Celsius, far below the required temperature for either of these reactions.)

Calcining limestone makes gas where there was none. If limestone were heated in a sealed vessel and the carbon dioxide could escape only by turning a dynamo, the resulting electricity could heat a resistor in the vessel. This could provide a fraction of the heat. But since the DG° measure measures free energy, not just heat, it already takes this potential efficiency pretty well into account.

Return to main text 

Return to main text 

"Boron", P. A. Lyday, USGS 2000 (http://minerals.usgs.gov/minerals/pubs/commodity/boron/120498.pdf).

Return to main text 

There are a number of university extension service documents on boron supplementation for crops, including one from the University of Michigan.

Return to main text 

3,600 megajoules, one megawatt-hour, is the amount of chemical energy in 29 US gallons of gasoline. That would mass 174 pounds, plus the mass of a container. The same in boron would mass only 145 pounds, and unlike gasoline, boron is safe to touch and won't flow, so a strong SUV driver could bring a megawatt-hour on board by lifting just that much.

But the largest 2000-model SUVs can take on fuel containing 1.4 to 1.5 megawatt-hours. Even in boron this would be an impossible lift for just about anyone. Presumably this will let future SUBV operators off the hook. They can divide a boron reload into several handy 60 lb aliquots.

More data on fuel mass-to-energy ratios are given in section Boron the Energetic, allowing, for instance, calculation of how massive a megawatt-hour in silicon would be: 3.6 GJ * 32.79 kg/GJ = 118 kg, 260 lb.

Return to main text 

Informed petrocolumnist "Policy Pete" frequently comments on projections of hydrocarbon availability. Read him here:

(http://208.49.25.221/policypete/policypete.htm)

Return to main text 

At 10 Celsius. Room temperature operation would be slightly less efficient.

The researcher is Hiroyuke Nishide, Waseda University Professor of Macromolecular Chemistry (http://www.appchem.waseda.ac.jp/fm-eng/nishid-e.htm).

Return to main text 

Industrial sturdiness is the presumed explanation for the three-tonne mass of the OXP-30 separator from Oxair. No particular endorsement of this product or slight against unmentioned competitors is intended.

The "30" is the number of kilowatts of compressor power the device requires. It purifies a little over 11.4 grams per second of oxygen, and at 30 percent efficiency, that can support 45 kilowatts of hydrocarbon power, 84 kW of boron power.

Return to main text 

The boria disk actually is meant to be melted and used as a transparent inert high-temperature bath -- a "liquid encapsulant" -- within whose protection crystals of certain semiconducting compounds can grow.

It probably would break if thrown onto a hard surface, but glass breaks because it is very notch-sensitive, and a very recently cooled glass form, whose surface has not yet had any microscopic notches worn in it, can be amazingly tough.

Toughness in a glass form can also result if it is of a composition that has a high coefficient of thermal expansion, and freezes from the outside in. The cooling interior shrinks and the exterior, having frozen earlier, can't deform plastically to accommodate it and instead is put into compression. Pure boria is such a composition. (This is so even though alloys of boria and silica are low-expansion, and can hold water while over a Bunsen burner flame, or go from kitchen oven to dining table.)

In any case the ingot wouldn't remain shiny and transparent. Atmospheric water vapour would make its surface opalescent with a very thin, slippery tarnish of boric acid. Where an athlete touched this solid acid her sweat would dissolve it, producing a solution that is harmless to human skin, but corrosive to boria glass. White, flaky handprints would form.

Return to main text 

Return to main text 

As a combustion fuel boron is favoured over hydrocarbons in use today on several performance measures. Most of them favour beryllium even more. So why not beryllium?

A very small fraction of the very small beryllium fraction in the Earth's crust occurs as the mineral bromellite, which is beryllia. Although very toxic if pulverized or dissolved, bromellite dissolves less in water than does aluminum oxide (mineral names corundum, ruby, sapphire), melts higher (2,507 Celsius according to WebElements), and is at least as hard. So it presents little natural hazard.

If there were cars that ran by oxidizing metallic beryllium, they could hardly produce the oxide in any other form than very fine powder. One might envision a variation of the Verneuil synthetic ruby-growing technique where beryllia replaces alumina and is born in the flame, not just melted there, but the boule would have to, and this is the impossible part, capture all the beryllia. Any that was still suspended in the gas downstream of the boule would tend to destroy both engine parts and people investigating why the engine had failed.

A recent price for 43 kg/GJ of beryllium oxide is US$9,500/GJ (Mineral Commodity Summary Beryllium, USGS 2001). For boria 59 kg/GJ translates to US$120/GJ (P.A. Lyday, C.H. Lindsay, and R.R. Coleman, Minerals Yearbook Boron, Table 4, p. 10, , USGS 2001). (These are PDF files.)

To work as an energy carrier, an atom of either of these rare elements can't be oxidized once and discarded, like an atom in petroleum. Instead, like a cell in a very durable rechargeable battery, it must be kept and recharged.

Since a gigajoule will propel a car approximately 300 km, the raw material cost of a boron/boria battery big enough for a car (including enough cells on board for several hundred kilometres' range, some at a power station being recharged, and some in transit to and from there) would add at most a few percent to the total. A similarly spread-out beryllium/bromellite battery would cost more than most cars.

If there were a magical way to cancel either beryllium's prohibitive toxicity or its very high cost, but not both, its use as an energy carrier would remain unlikely.

Return to main text 

Return to main text 

If "ash" is (as noted in the main text) reasonably defined, adding per-joule fuel masses and oxygen masses will give a very good approximation of per-joule ash masses. But the same sum answers other questions, such as what mass of (fuel plus oxygen) must be on board a rocket in order for its motors to have a joule to work with.

Like a rocket, a lead-acid battery is not an air-breathing device. All the atoms it will get energy from are already inside it. It may therefore be fair to include in an air-breathing system's mass not just onboard fuel but atmospheric oxygen, especially if that oxygen spends some of its time on board too, as it does in boria awaiting recycling.

If producing electricity is the goal, the battery's ability to do it directly while combustion energy must go through a heat engine and then a dynamo means the (fuel plus oxygen) mass might fairly be multiplied by three. The heat engine and dynamo together may not reach 33.3 percent efficiency, but neither will the battery reach 100 percent.

The result for (boron plus oxygen) is 177 kg per electrical gigajoule. When they are young and frisky, fully charged lead-acid batteries mass about 7.5 tonnes per electrical gigajoule.

Return to main text 

Return to main text 

Zinc has been recommended as an energy carrier on the grounds that it has a high energy density and can readily be oxidized in a fuel cell with room-temperature aqueous potassium hydroxide electrolyte.

A heat engine burning zinc or a fuel cell oxidizing it would take just 50 kg of oxygen from the surrounding air per gigajoule of released energy -- less oxygen than hydrocarbons take. That is zinc's best showing in the basic fuel performance charts. The two where it shows worst are fuel mass and ash mass.

On the ash volume chart the story is not so simple. Zinc oxide is solid, worth US$1 a kilogram (Table 16, J. Plachy, USGS 1999 (PDF)). Like boria it is therefore potentially a zero-emission ash: easy to hold onto, and worth holding onto. Its holding bin would be bigger than for any other solid ash listed, but if cars that run on liquid hydrocarbons had ash tanks, those tanks would in turn be much bigger than a zinc oxide reservoir.

Section Boron the Dense's data support the claim of high energy density: where a gigajoule in 3-ethyl-3-methylpentane occupies 30 litres, in a 200-kg piece of solid metallic zinc it would just about fill up 29 litres.

Heavy though it is, clean dry zinc oxide would not be too heavy to ship hundreds of kilometres to a power station, and bring back as zinc. The energy consumed in transport would be small compared to the energy the zinc would deliver. This can be estimated by multiplying by four the small losses calculated for ocean-spanning energy transfer using boria and boron in section Boron the Wieldy.

Such a power station might reduce zinc oxide by non-electrical means (cf. the boria reduction proposed in Boron Decombustion, cf. also, with a grain of salt for the claimed efficiency, http://solar.web.psi.ch/daten/projekt/zno/roca/roca.html), and would need no electrical distribution grid connection. But if the station is so connected, and the grid also reaches conveniently near to where a zinc fuel cell is, its zinc oxide can be electrolysed back to zinc without making a long journey, and without being removed from the electrolyte it was born in.

It might seem that an electrically recharged zinc oxidation cell is just a secondary battery, not a fuel cell, but in the plan Metallic Power Inc. says it is putting into effect, a dozen secondary batteries might be divided in 13 parts: 12 travelling fuel cells and one stay-at-home electrolyser.

This makes electricity the actual long-distance energy carrier, but as far as the cartridges that shuttle zinc and zinc oxide between fuel cell and electrolyser travel, so far does zinc carry energy.

Interestingly, as long as the number of cars powered by zinc fuel cells stays under a thousand, the zinc oxide in their ash reservoirs will actually be more valuable than energetically equivalent boria, despite boron's accurate reputation as a rarer element than zinc. The boria is twice as costly per unit mass, but less than a quarter as massive.

Most present-day traffic in zinc is in the elemental form, not the oxide. With boron it's the other way around. Supplying elemental boron for many thousands of boron cars will require new deoxidation plants. But the large existing market for boria will easily be able to accommodate those B car operators who want to sell the used B atoms back. Contrastingly, if there came to be thousands of zinc fuel cell motorists, and they wanted to sell the quarter-tonne loads of zinc oxide that each would frequently possess, the price such a load would fetch would diminish, perhaps below zero.

Return to main text 

Return to main text 

The alert reader may be saying, "Why not just have a 108-litre boria chamber? When a vehicle has been freshly loaded with 32 litres of boron, the 108 litres of boria don't yet exist. By the time they do, the boron won't. They could time-share."

This might be workable if boron pellets could be taken from the bottom of the chamber and slabs of boron oxide put in at the top. The glassy oxide is less dense than the pure element, so would tend mostly to stay on top, and could easily be cast in pieces too large to traverse whatever narrow exit channel the boron pellets are taking.

But there would have to be other benefits, beyond just reducing the system from 140 L internal volume in two chambers to 108 L in one, and there seem to be negative benefits in this scheme, such as the operator's having to know not to fill the chamber to the top with boron, and the difficulty of shepherding hard pellets from the bottom of a deep layer of them, a few at a time.

Contrastingly, there might be positive ones in having the boria in its own space. Retaining these glassy lumps (photo of one commercially prepared for another purpose here) is primarily an economic necessity, not a safety-related one. Spilling them on a road in an accident would not be catastrophic; they'd just have to be picked up.

That means boria bins in a vehicle would not need to be in the protected inner regions that gasoline or hydrogen tanks need, and compete with passengers and cargo for. They could be out at the edges, where space is much less scarce. Rather than needing collision protection, they might even help to provide it, although of course they'd be empty just after refuelling.

Interestingly, boron itself offers a similar locational flexibility. One could not put a hydrogen or hydrocarbon tank right next to the engine, lest it get heated and build up excessive pressure. But there's no harm if a bin full of boron bits gets warm.

Return to main text 

Return to main text 

Boron's room-temperature inertness to water, and the fact that boron oxide reacts with water and the product, boric acid, dissolves in it, are evidence that there is no room-temperature oxide film. If it existed it would not be protective the way insoluble alumina is protective to aluminum.

In "Phase Changes in Boron Ignition and Combustion" E. L. Dreizin, D.G. Keil, W. Felder, and E. P. Vicenzi review recent evidence that this film never turns up at any temperature. They report experiments where electrically heated boron filament would absorb oxygen but not burn even if held in air at a temperature of bright orange incandescence. A little hotter and it would seem to react suddenly with that oxygen, and burn with more from the air around it.

(COMBUSTION AND FLAME 119:272-290 (1999),
available 6 November 2001 at http://www-ec.njit.edu/~dreyzin/Papers-pdf/boronignition-C&F.pdf)

Return to main text 

Return to main text 

This is a sample footnote. You can return to where you were by clicking on the above link, or, with most Windows web browsers, by holding down the Alt key and pressing the left arrow key.