Print your own Supercaps

[Gil] recently wrote in to tell us about some awesome research going on at UCLA. Apparently by layering some oxidized graphite onto a DVD and tossing it into a lightscribe burner, it’s possible to print your own super capacitors; some pretty high capacity ones at that.

For those that are unaware, supercapcaitors are typically made using two electrolyte soaked, activated carbon plates separated by an ion permeable film. Since activated carbon has an incredible surface area huge energy densities can be reached, in some cases 1kJ/lb.

Laser-formed graphite sponge replaces the activated carbon in the researchers’ printed capacitors. A video after the break discusses  the whole process in moderate detail, meanwhile greater detail can be found in their two papers on the subject.

First one to print a transistor gets a bag of mosfets!

121 thoughts on “Print your own Supercaps

    1. I wonder if you produce graphene by this method or just reduced graphite oxide? Can you show that you really got graphene? Have you used for example Raman spectroscopy to verify your results?

  1. So given that capacitance is simply:
    Area * Permittivity

    What is it that makes graphine so “magical”? What’s the diff in doing this from just doing very thin electroplating?


    1. If you’ll read the papers, and do some reading on graphene, you’ll see that the graphene structures (pictured in the papers) create an absurd amount of surface area while being very conductive. As opposed to a metal film which is conductive with low surface area, and activated carbon which has decent surface area but poor conductivity.

    1. I, too, just shat myself after reading this and the papers.

      The paper claims a capacitance of 3.67 millifarads per cubic centimeter. This means that for a common DVD with a ‘printable’ area of 100 square centimeters (120mm outer diameter, 40mm inner diameter), you’d get a capacitance of 367 millifarads.

      0.367 F on a single DVD, using what appears to be a cheap-as-dirt manufacturing process.

      Unbelievable. I can’t even begin to imagine what’s possible once this technology is optimized.

      1. You forget the part where it uses graphene oxide, not graphite, which is what the editors also got wrong.

        Graphene is a bit more difficult to make at home, although you can make tiny flecks of it by scotch taping a bar of graphite.

      2. From the article opening paragraph:

        “We used a standard LightScribe DVD optical drive to do the direct laser reduction of graphite oxide films to graphene.”

        Readers Digest version: They made the graphene themselves.

      3. the intrinsic capacitance of single-layer
        graphene was reported to be~21 uF/cm2

        a compact disk has an approximate area of 109 cm/2

        from the math it tells me that one layer ~1/32″ thick the size of a cd would have a capacitance of 2289uF or .002289 Fareds so a little more math and we can find out approximately how thick this package would be for a density close to a fared
        would be 437 layers so about 13″ tall assuming each layer was 1/32 thick or 6.5″ if the layers were each 1/64″ thick

      4. To further clarify I was basing the thickness of the package on what I observed on the video that being what I would consider one layer, two plastic sheets with electrolytic sandwiched between I could see this possibly being as thick (thin?) as 1/128″ (about the thickness of 4 sheets of paper or so) this would bring the above sandwich down to about 3.41 inches

        oh and my math:

        21uF*109 = 2289 uF
        2289 uF = .002289 Farads
        1F/.002289F = 436.872
        436.872*(1/32) = 13.65″
        436.872*(1/64) = 6.83″
        436.872*(1/128)= 3.41″

      5. @Willaim what people seem to be not seeing is that this is not limited to the thickness of a CD disk. The disk provides a substrate for the film consisting of the graphene to lay on while it is “etched” by the laser in the Lightscribe. One of the backings suggested for the graphene was aluminum foil. Assuming a thickness of .006″ and your calculation of 437 layers, the resulting thickness would be 2.62200 for a single Fared.

      6. Chris, I didn’t interchange anything. I made a mistake. So sorry. I meant to state square, not cubic, centimeters. I ended up with a usable area of 10,053 square millimeters, equal to about 100 square centimeters.

        [(pi * (120/2)^2)] – [(pi * (40/2)^2)]

        Willaim, kindly refer to page 10 of the second linked paper. It states, “the areal capacitance of the LSG-EC was calculated to be 3.67 mF/cm^2….”

        If that figure is used, then a DVD with 100 square centimeters of area used as an LSG-EC would mean a total possible capacitance of 367 mF/cm^2.

      7. crap I just finished the PDF it states within that the layers were <100um or about 1/256" Bringing the total thickness to approximately
        436.872*(1/256)= 1.71"

        now bear in mind these calculations were performed with the best case scenario of 22uf /cm2
        while their tested results were

        3.67uF/cm2 in an aqueous solution of Sulfuric Acid (1.0M H2SO4) its a far cry from 22uf
        and brings the resulting 1Farad package back to a thickness of: 9.77"

        400uf = .000400 Farads
        1F/.000400F = 2500 Layers!
        2500 * 1/256" = 9.77"

        previously in the document though the electrolyte appears to be 1.0M H3PO4- Phosphoric Acid mixed with water.

        Finally at the end they used
        Poly(vinyl alcohol)(PVA)-H3PO4 polymer gelled electrolyte
        much safer than the strong acid solutions and apparently well suited for increasing the durability of the cells with comparable energy densities to the acid solutions

        The last thing to keep in mind about this is even with one Farad its only good for around 1 volt so to get any appreciable amounts of power cells must be put in parallel to increase voltage and series to increase capacity so the resulting 10"X5" stack could be 1V@1F or 2V@.5F or 4V@.25F(250000uF) or 5v@.2F(200000uF) or 100v@.01F(10000uF) etc..

      8. two things, the PVA electrolyte is still an acid solution (Phosphoric Acid mixed with Poly vinyl alcohol) but gelled so at least slightly safer whereas it was a strong solution of water and acid.
        kind of like a gel cell lead acid battery compared to a normal flooded Lead Acid battery

        that and mF is micro farads the same symbol as uF

        not anything against you Steve just made the note to keep someone from getting confused.

      9. This is intended as a reply to William:

        “and mF is micro farads the same symbol as uF”


        ‘m’ is the SI prefix for Milli, 10^-3, where ‘μ’ is the prefix for Micro, 10^-6.

        In the papers they used μ to denote micrometer thickness of the layers / stacks. Later in the same paper they use m to denote millfarad capacitances. If they wished to denote microfarads they would have used μ again.

        So either they have millifarad capacitances and you are reading the paper wrong, or they negligently used the wrong SI prefix repeatedly after demonstrating that they have the capacity and understanding to use μ where appropriate.

        I suspect you are mistaken.


      10. Perhaps you are right and in that case makes it far more exciting… milli is typically used for things like voltage whereas capacitance is almost exclusively microfarads not quite sure why they would go to a relatively unused suffix in capacitance
        and using mF instead of uF is usuallly avoided to keep it from being confused
        lloking at one of the capacitors the panasonic they listed as 22mf whereas on the panasonic site its listed as .0022(F) so yes that would seem to be what they did but now it just makes me suspicious of the results…

      11. ok so if we assume they mean 3670 uf and not 3.67uf it makes for some exciting news!
        .00367F * 109 =.400F
        1F/.400F=2.5 Layers
        round 2.5 to 3 layers for good measure and we come to
        3*(1/256″) =~ 1/8″ thick x 5″ for a 1 Farad unit.
        now that gets me excited!

    2. @Dax it states in the PDF:
      ‘Because of the simplicity of the device ar-
      chitecture and the availability of the graphite
      oxide precursor,which is already manufactured
      on the ton scale, these LSG-ECs hold promise
      for commercial applications.’

      I guess once the graphite oxide is hit with the laser it turns to laser-scribed graphene

      1. Graphene manufacture:
        Graphite oxide has attracted much interest recently as a possible route for the large-scale production and manipulation of graphene, a material with extraordinary electronic properties. Graphite oxide itself is an insulator,[15] almost a semiconductor, with differential conductivity between 1 and 5×10−3 S/cm at a bias voltage of 10 V.[15] However, being hydrophilic, graphene oxide disperses readily in water, breaking up into macroscopic flakes, mostly one layer thick. Chemical reduction of these flakes would yield a suspension of graphene flakes. It was argued that the first experimental observation of graphene was reported by Hanns-Peter Boehm in 1962.[16] In this early work the existence of monolayer reduced graphene oxide flakes was demonstrated. The contribution of Boehm was recently acknowledged by Andre Geim, the Nobel Prize winner for graphene research.[17]

        Partial reduction can be achieved by treating the suspended graphene oxide with hydrazine hydrate at 100 °C for 24 hours,[9] or by exposing graphene oxide to hydrogen plasma for a few seconds,[15] or by exposure to a strong pulse of light, such as that of a Xenon flash.[18] However, the conductivity of the graphene obtained by this route is below 10 S/cm,[18] and the charge mobility is between 2 to 200 cm2/(V·s) for holes and 0.5 to 30 cm2/(V·s) for electrons.[15] These values are much greater than the oxide’s, but still a few orders of magnitude lower than those of pristine graphene.[15] Inspection with the atomic force microscope shows that the oxygen bonds distort the carbon layer, creating a pronounced intrinsic roughness in the oxide layers which persists after reduction. These defects also show up in Raman spectra of graphene oxide.[15]

        Additionaly, exposing a film of graphite oxide to the laser of a LightScribe DVD has also revealed to produce quality graphene at a low cost. [19]

    1. Well, it’s a week and a half too early for April fools. ;D

      I wonder how feasible it is for someone to pull this off without a university laboratory though. But then, I saw a video with a dude making Aerogel in his garage, so I guess anything’s possible!

  2. So, oxidized graphite = what? Graphite is a stable, solid layered form of carbon, right? How is it oxidized? Is there some oxygen molecules but it isn’t CO2 or CO?

    (Or could I just put a sheet of plastic to my burner, blow some CO2 into it and BURNNN… to get same results?)

    1. Jump to: navigation, search Contents [hide]
      1 History and preparation
      2 Structure
      3 Applications
      3.1 Graphene manufacture
      3.2 Related materials
      4 See also
      5 References

      Graphite oxide, formerly called graphitic oxide or graphitic acid, is a compound of carbon, oxygen, and hydrogen in variable ratios, obtained by treating graphite with strong oxidizers. The maximally oxidized bulk product is a yellow solid with C:O ratio between 2.1 and 2.9, that retains the layer structure of graphite but with a much larger and irregular spacing.[1]

      The bulk material disperses in basic solutions to yield monomolecular sheets, known as graphene oxide by analogy to graphene, the single-layer form of graphite.[2] Graphene oxide sheets have been used to prepare a strong paper-like material, and have recently attracted substantial interest as a possible intermediate for the manufacture of graphene. However, as of 2010 this goal remains elusive since graphene obtained by this route still has many chemical and structural defects.

    1. I wonder how much power you could fit on a supercap cell the size of a SCUBA tank. The discs should fit inside nicely, with plenty of room for connections and circuitry.

    2. You can’t store more than 10V in super capacitors = no coil gun. sorry. But you can use supercaps to store power for your coilgun, so it will be still the same size (maybe slightly bigger), but it will be able to recharge much faster in matter of both:

      charging supercaps from external power supply
      charging high voltage caps from supercaps

      which means faster shooting with almost no time to recharge when supercaps are depleted completely.

      1. of course you can store more than 10 volts. not on one capacitor, you just wire multiple capacitors in series like we’ve always done. i have a 15v, 433 Farad array in front of me and it works just fine.

      1. Guess why it takes “seconds” – not “microseconds” to charge single supercap… Answer: internal resistance. You can have battery of 1V suppercaps but each one will still have just 1V to shoot through it’s internal resistance.

        Just try and see :-)

  3. This is pretty awesome. I really have to read deeper into this. Now I can start printing off super caps (among other things), and stop buying batteries for my smaller projects.

  4. This is awesome. If only there was a cheap way to get graphite oxide. Looking online shows it is very expensive on its own. Process seems easy enough after getting GO.

  5. Wait a second. It seems not many people are interpreting the video correctly. The cd is covered with plastic as merely a substrate to keep the graphite where it needs to be. You don’t need another disc for every burn. At the end of the video it shows a flexible supercap that is two thick layers of the burned/scribed material surrounded in plastic with an electrolytic solution trapped.

    So, in theory, these could be layered into any configuration or size you could want. I want to make some in conventional battery form factors, should they work.

    Now only to find this graphite oxide and an appropriate electrolyte.

    1. Yes, the DVD’s involvement is very minimal, but its what sells the story. Which is what makes me skeptical of these claims.

      I’ll look forward to someone else’s attempt at an LSG supercap to see how they go…

      1. The light scribe drive is actually instrumental in the high power densities described. The theoretical power density of graphene is about 500 f/g, we were previously only able to achieve only 1/5th of that. With the circular patterns that result as part of the light scribing process, approximately 250 f/g has been observed, 1/2 of the theoretical value.

  6. For the do-it-at-home peeps: (I read the docs)

    1. synthesise graphene (sorry, you’re on your own) and mix with water in 3.7 mg per ml
    2. spray onto PET plastic or aluminum foil and let dry at room temp and affix to top of CD/DVD
    3. blast with lightscribe drive 6 times.(just make it black)
    4. paint some spots with conductive silver paint to hook up aligator leads to two pieces of this stuff
    5. sandwich a liquid electrolyte and porous spacer in between, or mix their solid-state gel stuff and squash it in there (read the paper, it’s simple to make if you can buy the stuff you need)
    6. use it

  7. If this is *REAL* it is a complete game changer.
    Manufacturing costs of the super capacitors
    then drops to zero. Think the chemistry for
    supercaps is much more enviornmentally friendly
    than batteries. No more dependencies on exotic
    rare earth materials. Who cares if the energy density is lower if cost dramatically drops and there is virtually unlimited charge discharge cycles. Carbon oxide $220 for 500 mg.
    Lets get a home process published!

    1. We just need someobody with a chemistry set and access to Sigma Aldrich to try it out. I have reason to believe it’s quite real. I read the papers and it makes perfect sense to me, I’m just wondering how reproducible it will be. It seems to me that the only thing the paper is light on is details of “Why specifically Lightscribe?” Is it just the 788 nm wavelength? Or something else?

      The chemistry is very simple and easy to understand. What I want to know more about, and what the paper does not discuss, is why does a Lightscribe laser convert graphite oxide into such an extremely fine layer (7.6 micrometers) of non-exfoliated (which is key to keeping the high surface area) graphene?

      1. My take is that a Lightscribe DVD burner is a very consistent, precise and inexpensive machine for annealing or binding the graphite to the plastic separator. And patterns can be designed and implemented using convenient PC software. Nothing special, it’s ubiquitous and it works.

      2. There is another process using the flash from a camera which heats it rapidly (the article said “explodes”) to create the graphene. Don’t have the link but you can search. Most likely the Lightscribe does the same thing just not in one shot.

  8. Oh, and the paper spends a lot of time discussing how this LSG or laser scribed graphene makes an extremely durable flexible conductor, and its many uses besides as a capacitor electrode.

    And you should probably also note that the exceptional performance may be tied to a special proprietary ion-porous separator called Celguard 3501.

    1. Wiki it, they’ve been making carbon oxides in similar fashions since the 1800s. CO and CO2 are not the only oxides of carbon.

      On an interesting side note: it is possible to incinerate a diamond, the carbon can oxidize leaving not more than a tiny bit of ash.

  9. What’s even cooler is I suspect that they could use a mask and flash the film of graphene oxide with a Xenon flash or a high powered laser and achieve in microseconds what take some time (30min+ from memory?) with the light scribe.

    In short, easy to scale for mass manufacturing.

    I’m sure the light scribe is great for experimenting with different patterns however, which I suspect could be the secret sauce.

      1. Yeah. The secret is that ‘foreign’ governments often pay for top students to come to the united states while simultaneously the united states government pays the students too(in various ways). So in essence its either have 160 thousand dollars lying around AND be a 1/1000 student OR be a pretty good student from a number of countries that take a higher stance on education.

        Sources: the vast number of foreign students that are part of a masters program overseen by my significant other

  10. Can anyone verify the 1KJ/lb number ? That sounds really low for a super cap. And in the video he says that it can hold as much as a conventional battery. But a Li-Ion battery for example can hold as much as 150 Wh/kg = 540KJ/lb. I am missing something ?

  11. Correction to my last comment..

    Can anyone verify the 1KJ/lb number ? That sounds really low for a super cap. And in the video he says that it can hold as much as a conventional battery. But a Li-Ion battery for example can hold as much as 150 Wh/kg = 245KJ/lb. I am missing something ?

  12. >> Manufacturing costs of the super capacitors then drops to zero.
    At least one current version of supercap is based on activated charcoal, which is very cheap. Even the carbon aerogel varieties don’t exactly have terribly expensive raw materials.
    The production of graphite oxide described on wikipedia sounds seriously scary. They’re using things that are “oxidizers” at the “immediately explodes on contact with any reducing agent” level. Chlorate and fuming nitric acid. Oh my! permanganic acid. Oh dear! And the GO appears a bit on the explosive side as well (“exfoliates with an impressive increase in volume.”)
    It sounds a bit like the interesting part here is being able to obtain higher-quality graphene films from the GO than other methods.

  13. Thats ingenious, I was looking into hacking LS burners for the position sensors a while back.
    As it happens I have three broken drives here which probably can be modified with a Bluray laser diode and a slight optics tweak so that the existing sensor and associated circuitry can be used.
    Seems that the LS part is a lot less sensitive to diode degradation than the DVD/CD burner.

    1. I have a couple dead LS drives, and I love dissecting old optical drives for parts. No idea about the LS-specific components though, know any sources of info for their possible (re-)applications?

      Obviously I could just google, but you said you’d been looking into it. ;)

  14. “kJ/lb” – now that is a mashed up unit if ever I saw one! Surely you should use ft-lbf or BTU for energy :-)
    Even though the French invented it, even us Brits use metric.

  15. ok so if we assume they mean 3670 uf and not 3.67uf it makes for some exciting news! .00367F*109 =.400F
    1F/.400F=2.5 Layers
    round 2.5 to 3 layers for good measure and we come to
    3*(1/256″) =~ 1/8″ thick x 5″ for a 1 Farad unit.
    now that gets me excited!

  16. So just to be clear I did a bit of research and pricing etc…
    From what i can tell it looks like 1 gram of Graphite Oxide will get you to about 556 Farads. Which at 1.5 Volts (being generous here) means that you can manage to produce about .24 watt hours per gram of this stuff.
    The current rate in bulk is easily 150 per gram however add in other materials and lets say your at 160 per .24 watt hours. Although it will be very small and very dense the cost is still way too much to ever replace a battery.
    Were talking to get to a standard AA battery of 1.2Volts and 2600mAh capacity 3.12 Watt Hours
    You would need 13 grams which would cost you 2,000 USD…

    So although its great to be able to produce a 556 Farad 1.5 Volt Super capacitor that can charge and discharge quickly and last a large number of cycles 10,000+ its still not as cheap as making your own super capacitors

    They have managed to make super caps around 450 Farads 1.2Volts and the total cost to make each one is roughly around 10 USD.
    I prefer to save my money.

    1. High Purity Graphite from Sigma-Aldrig 332461 box of 2,5kg for 33£. Make some magick with chip chemical “Improved Synthesis of Graphene Oxide
      Daniela C. Marcano,† Dmitry V. Kosynkin,† Jacob M. Berlin, Alexander Sinitskii, Zhengzong Sun,
      Alexander Slesarev, Lawrence B. Alemany, Wei Lu, and James M. Tour”
      And you have DIY dolar/capcitor printer.

    2. yea graphite oxide itself looks spendy. However, chemicals for making it are relatively cheap, as graphite, phosphoric acid, sulfuric acid, and potassium permanganate are pretty cheap substances. Dunno what the steps involved are, but bulk prices for the components are in the $10-$20/L range, which, even with lots of loss, has you producing kilos of GO in the $200-$300 range.

      1. The recipe for GO from Marcano et. al. (2010):

        Graphite flakes (Sigma-Aldrich, cat #332461, ~150 um flakes) were oxidized using three different procedures: improved method, Hummers’ method, and Hummers’ method with additional KMnO4.

        For the improved method, a 9:1 mixture of concentrated H2SO4/H3PO4 (360:40 mL) was added to a mixture of graphite flakes (3.0 g, 1 wt equiv) and KMnO4 (18.0 g, 6 wt equiv), producing a slight exotherm to 35-40 °C. The reaction was then heated to 50 °C and stirred for 12 h. The reaction was cooled to rt and poured onto ice (~400 mL) with 30% H2O2 (3 mL). For workup, the mixture was sifted through a metal U.S. Standard testing sieve (W.S. Tyler, 300um) and then filtered through polyester fiber (Carpenter Co.) The filtrate was centrifuged (4000 rpm for 4 h), and the supernatant was decanted away. The remaining solid material was then washed in succession with 200 mL of water, 200 mL of 30% HCl, and 200 mL of ethanol (2x); for each wash, the mixture was sifted through the U.S. Standard testing sieve and then filtered through polyester fiber with the filtrate being centrifuged (4000 rpm for 4 h) and the supernatant decanted away. The material remaining after this extended, multiple-wash process was coagulated with 200 mL of ether, and the resulting suspension was filtered over a PTFE membrane with a 0.45 um pore size. The solid obtained on the filter was vacuum-dried overnight at room temperature, obtaining 5.8 g of product.

  17. @bakamoichigei Reply sent.

    This procedure would probably work
    with metal/sulphur mixtures to make
    memristors/EL/etc as long as air
    is kept out of the sandwich using
    an inert barrier such as glass.
    Might also work using sodium
    silicate, this is worth a try.

  18. They are using the Hummers method of creating Graphic Oxide. This involves mixing graphite with a water free mixture of concentrated sulfuric acid, sodium nitrate and potassium permanganate. This is stirred continuously for 30 min while the container is cooled by an ice cube bath. Then it is VARY carefully diluted and filtered. (It reacts like crazy when you start adding the water, so you have to go slow enough to keep the temp under 98C). The whole process is detailed in a 1957 paper by William S. Hummers.

    @Stefan “Is this real? It sounds pretty unbelievable.”
    The AAAS (Science Mag) thinks it is, and they think it will be a big game changer.

  19. Would I be correct in my observation that..

    a) The film is semi transparent.
    b) The described electrolyte could be replaced with one which was photosensitive (google for some likely candidates).
    c) This could therefore open up the possibility of a cheap flexible solar panel.
    d) Profit.

  20. For reference, Here is the “Modified Hummers” method they used. Does anyone have the full text of the Marcano paper that describes their improved method?
    “GO was synthesized from natural graphite powder (325 mesh, GAK-2, Ukraine) by the method of Hummers and Offeman. 18 It was found that, prior to the GO preparation according to ref 18, an additional graphite oxidation procedure was needed. Otherwise, incompletely oxidized graphite-core/GO-shell particles were always observed in the final product. The graphite powder (20 g) was put into an 80 °C solution of concentrated H2SO4 (30 mL), K2S2O8 (10 g), and P2O5 (10 g). The resultant dark blue mixture was thermally isolated and allowed to cool to room temperature over a period of 6 h. The mixture was then carefully diluted with distilled water, filtered, and washed on the filter until the rinse water pH became neutral. The product was dried in air at ambient temperature overnight. This preoxidized graphite was then subjected to oxidation by Hummers’ method. The oxidized graphite powder (20 g) was put into cold (0 °C) concentrated H2SO4 (460 mL). KMnO4 (60 g) was added gradually with stirring and cooling, so that the temperature of the mixture was not allowed to reach 20 °C. The mixture was then stirred at 35 °C for 2 h, and distilled water (920 mL) was added. In 15 min, the reaction was terminated by the addition of a large amount of distilled water (2.8 L) and 30% H2O2 solution (50 mL), after which the color of the mixture changed to bright yellow. The mixture was filtered and washed with 1:10 HCl solution (5 L) in order to remove metal ions. The GO product was suspended in distilled water to give a viscous, brown, 2% dispersion, which was subjected to dialysis to completely remove metal ions and acids. The resulting 0.5% w/v GO dispersion, which is stable for a period of years, was used to prepare exfoliated GO.”

  21. Thanks re. the tip about using a
    plastic sheet to hold the GO.
    I would expect that refitting the
    LS read disk under the hub and moving the sensor would
    then avoid having to use a disk
    at all, as long as the laser “sees”
    a valid surface that it can focus
    on everything should be fine.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s