h a l f b a k e r y
These statements have not been evaluated by the Food and Drug Administration.

meta:

account: browse anonymously, or get an account and write.

 user: pass:
register,

# On SuperCapacitance

Exploring the physical limits
 (+3, -4) [vote for, against]

A "capacitor" can be a very simple device. It's first version was called a "Leyden Jar", and was invented in 1745 (see link). In modern form, just take an ordinary glass jar and wrap the outside with aluminum foil, and also put a layer of foil on the inside of the jar. Make sure the two layers don't come any closer to each other than the thickness of the jar. Also make sure the jar itself rests upon some sort of non-conducting surface, like a wood tabletop. Then attach one wire to each layer of foil, so that you can put some energy into this capacitor, and withdraw it later.

Capacitors only store Direct Current electricity. If you hook this jar to a 9-volt battery, the battery will output some electricity that will be stored in the jar, until the voltage of the stored electricity equals the voltage of the battery. This could take several seconds. WARNING! CAPACITORS CAN BE DANGEROUS!!! You might be able to safely put your tongue across the terminals of a 9-vold battery, and feel a tingle, but if you do that to THESE wires, you could be burned or worse. The small battery is strictly limited in how fast it can deliver Amperage, the essence of a flowing electrical current; a capacitor is NOT so limited. An electrical short-circuit, such as one involving a wet tongue, is an invitation for a capacitor to dump all its stored energy AT ONCE. It's not the voltage that causes electrocution; most people can easily handle a 30,000 volt shock after scuffing shoes across a carpet on a cold dry day, and then touching a metal water pipe. It's amperage that kills, and most capacitors can easily dump dozens or hundreds of amps in an instant (and some capacitor banks are specially built to do millions of amps; see the Z-machine link). That carpet-scuffing shock involves only a fraction of an ampere; messing with Leyden jars has occasionally proved to be fatal.

Historically, the total amount of energy that a capacitor can store has actually been rather small. If this seems surprising, it is a consequence of difference between "energy" and "power". Let us consider a dam holding back a lot of water, to make this difference clear. The total energy available is represented by the amount of water behind the dam. The power of the dam is represented by the rate at which water flows from the reservior through the dam. Even if no rainfall adds water to the reservior, the dam can produce power for a considerable time (much like a battery). If the dam is made to disappear, and the whole body of water rushes through its former location at once, that is a LOT more Power in action, but the total amount of Energy released would have been the same. Power is the RATE at which energy does things, and a shorted capacitor is always equivalent to that disappeared dam. Even without a lot of total Energy available, the Power it can exhibit can be formidable.

It happens that the total Energy that a capacitor can store is directly related to the total surface area of the foil sheets. The thickness of the sheets doesn't matter; only the total area separated by an insulator matters. You can now make a MUCH MORE DANGEROUS capacitor at home this way, using two rolls of aluminum foil and two rolls of plastic wrap, all the same width, perhaps 30cm (edge view; ignore the dots):
. ----------------Layer of foil-->(unroll)
----------------Layer of plastic wrap-->(unroll)
. ----------------Layer of foil-->(unroll)
----------------Layer of plastic wrap-->(unroll)
Note how, at left, the two layers of foil are not "at the edge"; they must not be allowed a chance to contact each other. The layers also need to be offset from each other, as in this "top" view (ignore the dots):
. ______edge of first foil layer
_|_____edge of both plastic layers
||-------dotted line is edge of other foil layer
||at far left is end of plastic layers
|| . . all four layers are overlapping here in middle
||at near left is end of foil layers
||-------dotted line is edge of first foil layer
|______ edge of both plastic layers
. |_____edge of other foil layer
Now return to the first sketch above, and start rolling up the four layers, from left to right. If you are unrolling the raw materials as you roll up the assemblage, you will find that the "outside" layers must be unrolled faster than the "inside" layers. This is natural. Stop when you become afraid of just how dangerous this capacitor is becoming. Make sure that the final ends of the foil layers remain completely insulated by the plastic wrap (cut the metal before cutting the plastic!). The final thick assembled roll, hopefully with plastic on the outside, will have metal foil sticking out of both ends of the body of the roll. At each end the extended metal foil can be squished down to a minimum size; you would attach wires to those points.

The maximum voltage that a capacitor can store is related to the quality of the insulator between its conductors. Plastic wrap is pretty thin stuff, and probably can't be expected to withstand much more than 12 volts. IF IT CAN (and I'm not saying it will), then you could hook it up to an ordinary 12-volt car battery, and charge this capacitor. Don't be surprised if you get a big spark when you make the connection! BE AWARE that hydrogen gas that naturally occurs in the vicinity of lead-acid car batteries can explode if a big spark happens. (Have I made it clear: THIS IS DANGEROUS!!! ?) If the plastic wrap can't withstand the 12 volts, then you will see an even bigger bang when the capacitor self-destructs. If the plastic CAN withstand the 12 volts, then THE CHARGED CAPACITOR WILL BE QUITE DEADLY IF MISHANDLED. All by itself, a car battery can put out a dangerous amount of amperes, and all car batteries have warnings about it. But that is nothing compared to the flood of amperes this capacitor can release, when its two wires are shorted together. YOU HAVE BEEN WARNED!!!

However awesome that particular capacitor may seem, IT IS STILL INADEQUATE for the purpose of storing a decent total amount of energy, such as you might want to use to power your air-conditioning system on a hot night. That is, if you had a method of gathering solar energy during the day, you would want to gather and store some extra for use at night. Batteries are seldom more than 75% efficient at storing and releasing energy; capacitors are almost 100% efficient. Therefore the search goes on, for ever-higher-capacity capacitors. Current versions are called "supercapacitors", or even "ultracapacitors".

They've almost reached the point where you can power your cell phone with one for a decent amount of time (remember, the huge power surge I've talking about before involved an electrical short-circuit; when a proper electrical "load" is used, a capacitor works like a battery --just not as long-lasting, until very recently). A cell-phone supercapacitor would be cool because it can be charged at maximum rate (equal to a short-circuit discharge rate), and be fully charged in seconds. It also will be good for many thousands or even millions of charge-discharge cycles; capacitors have no moving parts (or even moving chemicals like batteries do) to wear out. MOST of the entire chemical-battery industry will be doomed when supercapacitors become capacious enough (likely within 5 years); plan your investments accordingly!

And now it is almost time for the Idea this post is really about, for which all the preceding was background information. As mentioned, the amount of energy that a capacitor can store is directly related to the total surface area of its conductors. It also is related to the ability of its insulating layers to withstand voltage. The more voltage a fully-charged capacitor can handle, the longer it takes to "drop" to a useless level, when powering something.

So, we want to constuct our Ideal Capacitor out of layers such that its conductors are as thin as possible, and its insulators are as resistive as possible.

For the first, I recommend "graphene". This is the stuff, rolled up, of which carbon nanotubes are made. It is one layer of a chunk of graphite. It is one atom thick, a completely 2-dimensional array of hexagons, and an excellent electrical conductor. See link.

For the insulating layer, I want to suggest something that I'm not sure is known to exist, but OUGHT to be able to exist. The chemical compound "boron nitride" should be able to form a 2-dimensional hexagonal lattice identical to graphene. I have a link that mentions "a repeating array of hexagons in 3-dimensional space", but that's not quite what we need here. PERHAPS the 3-dimensionality of the array is only minor, something like this:
B< (the odd symbols represent chemical bond angles)
__>N (ignore the underscores)
B<
__>N
I've portrayed a vertical zigzag; a horizontal zigzag would have all the borons on one side and all the nitrogens on the other side of the imaginary 2-dimensional plane between them. That will be good enough!

Anyway, the boron-nitride chemical bond is a very strong bond, and such bonds always make excellent electrical insulators. Therefore we should seek to construct our Ideal Capacitor out of two layers of graphene, and two layers of boron nitride, rolled up very much as I described earlier for two layers each of aluminum foil and plastic wrap.

 — Vernon, Feb 23 2009

Leyden Jar http://en.wikipedia.org/wiki/Leyden_jar
As mentioned in the main text [Vernon, Feb 23 2009]

Z machine http://en.wikipedia.org/wiki/Z_machine
As mentioned in the main text [Vernon, Feb 23 2009]

Graphene http://en.wikipedia.org/wiki/Graphene
As mentioned in the main text [Vernon, Feb 23 2009]

A little about Boron Nitride http://arstechnica....nly-momentarily.ars
As mentioned in the main text [Vernon, Feb 23 2009]

Boron nitride http://en.wikipedia.../wiki/Boron_nitride
[xaviergisz, Feb 24 2009]

Boron Nitride dielectric properties http://www.springer...t/v642715376h84up1/
It's a Springer Journal article on the material properties of Boron Nitride (hexagonal). [Jinbish, Feb 24 2009]

'Supercapacitor', press release by Rensselaer Polytech http://news.rpi.edu...o?artcenterkey=2280
Seems like they've done some carbon nanotube stuff to develop a supercapacitor. [Jinbish, Feb 24 2009]

Mylar http://en.wikipedia...biaxially_oriented)
Mylar is a "trade name"; the chemical name is rather longer. [Vernon, Feb 24 2009]

Mylar's structure http://en.wikipedia...ylene_terephthalate

Dielectric http://en.wikipedia.org/wiki/Dielectric
Not quite the same thing as an insulator. [Vernon, Feb 24 2009]

HUGE Progress report! http://spectrum.iee...-mobile-electronics
I suspect this technology will be found everywhere in just a couple years. [Vernon, Sep 30 2015]

hyper capacitor http://www.hyper-capacitor.com/
Energy density better than lithium batteries. [travbm, Oct 29 2015]

Cut out the unabridged history of the capacitor and put a link to a wikipedia article or something.
 — phoenix, Feb 23 2009

So, to summarize - your idea is to make a capacitor "out of two layers of graphene, and two layers of boron nitride, rolled up".
 — jutta, Feb 23 2009

I liked the capacitor background reading. Vernon's prose has a certain charm and it was educational for me.
 — bungston, Feb 23 2009

 //certain charm//

USING GODDAMN CAPS!
 — Jinbish, Feb 23 2009

 So.

 First, your thoughts on the lethality of capacitors charged to reasonable voltages are slightly whacky. (For what it's worth, your capacitor charged to 9V will behave very much like a 9V battery if you stick your tongue on it; this is because your tongue will have a resistance of a few hundred ohms: neither the internal resistance of the battery, nor the lack of any internal resistance in the capacitor, is relevant here). But, in general the background information about capacitors is fine.

 Oh, and the resistance (more particularly, the breakdown voltage) of regular Saranwrap is more than enough to handle a few hundred volts, as long as there are no nicks or holes in it. Regular mylar-film capacitors have thinner dielectrics and can be made to take hundreds or thousands of volts.

However, the original idea here seems to be saying "make a better capacitor by using these two amazing materials, one of which I'm not sure exists, and neither of which I really know much about". Also, I'm pretty sure that bond- strengths are not the only relevant factor in deciding dielectric constant, breakdown voltage and what-have- you.
 — MaxwellBuchanan, Feb 23 2009

What you're describing is called intercalated graphite, and there's quite a body of work on the stuff, though most is of basic research, rather than applied. I think the practical problem with this would be in controlling where the graphite sheets are electrically connected, and where they are not.
 — colorclocks, Feb 23 2009

 I think a layering of graphene and insulator would act like a semiconductor bandgap. So in a similar way that quantum dots act like semiconductors, this would be a quantum *plane* semiconductor.

There would be lots of interesting applications such as photovoltaics and lasers.
 — xaviergisz, Feb 24 2009

Having never heard most of the terms used I was glad you provided the tutorial and links. I doubt I would have ever stumbled across the term Leyden jar. Thank you.
 — 2 fries shy of a happy meal, Feb 24 2009

 [Jinbish], if the HalfBakery allowed us to do italics and bold and underlining to stress a point, I would be happy to forego the caps.

 [MaxwellBuchanan], thanks for the data about plastic wrap. I don't mind having erred on the side of caution in the main text here. Next, please remember this IS the HalfBakery. We are allowed to talk about things we don't know all about. If we did, we would be posting at the Patent Office, instead of here. I AM aware, however, that likely the high-capacity device I described here would likely only be suitable for a couple of volts., mostly because of the thinness of the boron-nitride layer.

[colorclocks], a cross-section of this proposed capacitor may resemble intercalated graphite, but I did try to specify a long long spiral of just four sheets, rather than lots of individual sheets.
 — Vernon, Feb 24 2009

//We are allowed to talk about things we don't know all about.// Well, true enough and fair enough. It just seemed to me that this was innovative only in terms of using a different material, and the background on the material seemed a bit thin. I am now wondering whether a capacitor could be formed from layers of thinly-sliced dill pickle (conductor) and cheddar (insulator).
 — MaxwellBuchanan, Feb 24 2009

Oh dear, I think I just ate a pastrami and pickle capacitor. On granary. I hope it wasn't charged.
 — coprocephalous, Feb 24 2009

 I'm with [MB], I'm afraid, [Vernon]. I also have a funny feeling that two layers with of BN aren't going to cut it. You'll just end up with a material that is 2 parts graphene and 2 parts BN and will act like a poor quality semiconductor in across the grain and a half decent conductor in the coaxial plane.

 That's not to say that such materials aren't interesting. As [xaviergisz] suggests, the quantum effects may be interesting and [colorclocks] suggests that there is work on such graphene materials. However, in terms of trying to reduce 'd' in the equation C∝A/d , I don't think this idea is going to make the cut.

//On granary.// : Woff! I'd hate to see the back-EMF on that one!
 — Jinbish, Feb 24 2009

Apparently graphite conducts only in the direction of the plane. If you had perfect graphene layers, would you actually need an insulating molecular layer?
 — Loris, Feb 24 2009

 [xaviergisz], thanks for the link; the alpha/hexagonal form of BN is exactly what I was trying to describe.

 [Jinbish], your link states, and I quote: "In the normal direction BN is also an excellent low-loss insulator." Other text in the article indicates the "normal" direction is perpendicular to the plane of hexagons, and that insulating property is exactly what we want in a capacitor-layer.

I notice none of you thought to mention the possibility of building the proposed capacitor with extra layers of BN. That is, something like this:
. ----------graphene sheet-->
----------boron nitride sheet-->
----------boron nitride sheet-->
. ----------graphene sheet-->
----------boron nitride sheet-->
----------boron nitride sheet-->
Ignore the dots and roll the 6 layers up from the left, as described in the main text for 4. This doubles the distance between the graphene layers, allowing a higher voltage or less chance of tunnelling (or, perhaps more important for manufacturing, allowing any imperfection in one BN layer, that would allow current through it, to be placed next to a non-imperfect spot on the adjacent BN layer).
 — Vernon, Feb 24 2009

[Vernon]: Ok - what's your point?
 — Jinbish, Feb 24 2009

[Jinbish], in the main text I wrote, "we want to constuct our Ideal Capacitor out of layers such that ... its insulators are as resistive as possible." You seemed to be saying that one boron nitride layer can't be good enough. Therefore, use two (or even more). Simple.
 — Vernon, Feb 24 2009

Is boron nitride particularly good as a dielectric? Is it better than, say, mylar? I think these are basic numbers that ought to have been in the idea to justify it; otherwise it's "build train wheels out of camembert", no?
 — MaxwellBuchanan, Feb 24 2009

 I think BN is a decent dielectric, which was why I provided the link. The thing that I am not quite sure about is 'why use these materials'? I understand that you can use graphene in layers of a couple of atoms deep, and possibly BN too - and then plugging that into C varies A/d gives you larger capacitances. The bit I don't get is that surely at such scales of thickness you get large leakage currents?



Ok - looks like some dudes have made a capicitor using aligned carbon nanotubes and a liquid electrolyte... (linky). I think that the right term is "Electric double-layer capacitor" and the capacitor isn't so much two distinct plates as a material that has distinct layers in it that produce the capacitance effect.
 — Jinbish, Feb 24 2009

[MaxwellBuchanan], one way of looking at a dielectric is to think of how much voltage it takes to break its molecular bonds. It logically figures that the stronger the bonds, the better the dielectric. Mylar contains a mess of mostly carbon-hydrogen, carbon-carbon, and carbon-oxygen bonds. The bond strength of boron nitride is certainly greater than the carbon-hydrogen bond, and comparable to the other two (can't be almost as hard as diamond without it).
 — Vernon, Feb 24 2009

I'm not sure that dielectric properties depend on bond strengths; copper is not a dielectric, whereas air is, to an extent. Anyway, how does BN actually compare as a dielectric to polythene or Mylar? This is presumably known?
 — MaxwellBuchanan, Feb 24 2009

 [MaxwellBuchanan], there are multiple factors. In copper there are atomic bonds and also there are loose electrons. An applied voltage directly affects the loose electrons. In water there are often dissolved ions that would also be directly affected. In other substances, such as air (or deionized pure water), bonds have to break to GET loose electrons.

Perhaps we are confusing "dielectric" with "insulator"? I added a link. I'm pretty sure that for a capacitor, the ability to resist a flow of electrons, between the two conductor-sheets, is what we want more than anything else. It should be obvious that the more tightly-bound are the electrons of an insulating substance, the better for that purpose.
 — Vernon, Feb 24 2009

 There are indeed multiple factors, and I don't think we're confusing insulation with dielectric strength. As far as I can tell from the Wikipedia article on dielectric strength, there's no immediate correlation between bond strengths and breakdown voltage.

 If I had to guess, I would say that dielectric strength has a loose analogy with tensile strength. Tensile strength "should" depend on bond strengths, but in reality it seldom does; it has much more to do with crack-stopping properties of the material and on defects. Likewise, I suspect that dielectric strength has less to do with bond energies, and much more to do with the ability of the material to prevent avalanching of electrons at the point of incipient breakdown, and on the frequency of defects which will nucleate these avalanches.

Anyway, no point labouring it. If you think BN will be a good dielectric, then that's OK.
 — MaxwellBuchanan, Feb 24 2009

Hyper capacitors are being made today but not reached commercial production.
 — travbm, Oct 29 2015

 [annotate]

back: main index