Instructables
Picture of Make a Graphene Supercapacitor

This Instructable (loosely) follows the method of graphene production described by C&EN, who cite a paper published by Parvez, Wu, et. al. In the paper (Exfoliation of Graphite into Graphene in Aqueous Solutions of Inorganic Salts), it is explained that putting a graphite electrode and a platinum electrode into a solution of inorganic salts (ammonium sulfate works best, they say) with a 10V supply (graphite = anode, platinum = cathode) causes the graphite to be torn apart by electrochemical activity to become graphene! Graphene has attracted much hype and study recently. One of its near-term applications may be in high density supercapacitors.

 
Remove these adsRemove these ads by Signing Up

Step 1: Making the graphene

Picture of Making the graphene

To make the experiment conditions described in the paper, pure platinum was required for electrode material. Unfortunately, all I had on hand was 90%/10% Pt/Ir wire, a deplorable situation no doubt shared by many in the audience. (Those who do not have either on hand may try gold, or carbon, but I have not tested these.) Also, instead of graphite sheet of some sort that was used by the referenced experimenters, we used a piece of graphite mold blank. This choice is probably why our experiment time was around 19 hours but the published one was 3-5 minutes. Surface area to volume ratios...

To make graphene, you need:

10-15V supply, a car battery would work

0.1 molarity solution (NH3)S04, in water, and awareness of how to calculate molarity (ammonium sulfate weighs 132.14 g/mol)

graphite, we used a piece of a mold blank (it may be possible to use carbon fiber but I haven't tried)

Pt electrode or platinized electrode or possibly many of the noble metals (we used 90%/10% Pt/Ir wire)

See attached lab notes for a detailed description of the exfoliation procedure as completed by me.

DIY-Guy3 months ago

Quite interesting!

Graphene has been made by pulling sticky cellophane tape off a graphite block. Could this give a nice insulator/conductive/insulator/conductive surface if cut and stacked for a capacitor? Or is attaching electrodes too difficult with clear tape as a base?

quicksilv3rflash (author)  DIY-Guy1 month ago

I've never tried to attach electrodes to such a structure -- try and let us know!

They (electrodes) may need to be electroplated onto the graphene to get a grip.

megapix1 month ago

Nice work! Hope you will keep us updated on your progress. Oh, I like your lab notes too: "insect shows no sign of life, even after drying." ROFL.

thereyno3 months ago

what is the capacity of it sir? can we calculate?

quicksilv3rflash (author)  thereyno1 month ago

Probably not too well, since we don't know the equivalent parallel resistance, unfortunately.

fobos73 months ago

hola. mucho muy interesante. Gracias por dejar verlo

quicksilv3rflash (author)  fobos71 month ago

De nada

AJMansfield3 months ago

As a possibly cheaper approach, you might try producing your defect-free few-layer graphene by shear exfoliation in liquids (AKA throw the graphite in a blender.)

quicksilv3rflash (author)  AJMansfield3 months ago

Looks like I'd need to have a 'high-shear mixer' with a 6000 rpm spindle speed... I think the 15V supply and ammonium sulfate is easier for now :)

You can get a blender for $15 easily capable of 6000 rpm. Typical max kitchen blender speeds run around 11000 rpm, nearly double the needed speed. The one slight issue might be blade durability, but as long as you can get cheap replacements it shouldn't be a problem.

quicksilv3rflash (author)  AJMansfield3 months ago

*adds to pending experiment list

FYI the electrodes in high-end automotive spark plugs are often platinum or iridium.

Urnammu3 months ago

The carbon electrodes should also be taken into consideration. Both charging and discharging occurs through these, so I think the resistance is much larger than 10M.

tholopotami3 months ago

I congratulate you for entering the jungle! There are so many wild animals out there!

The major difficulty I suppose is to evaluate what kind of graphene you produce after exfoliation. Unless you have access to scanning tunneling microscopy (STM) or other high tech characterization methods you do not know the morphology of the film which may be crucial. On the other hand you do not know the quality of the graphite that you start with. This can be improved. I suggest to look for Highly Ordered Pyrolytic Graphite (HOPG) which is of a high quality (and expensive). We use it as a reference surface for STM. This is made out of complete sheets e.g. 1cmx1cm on the whole surface. I would not worry about the Pt/Ir wire, platinum has a higher concentration. Never mind that it has a small surface.

Unfortunately I could not read your lab notes as they were attached. Please include some information in your instructable, like capacitance or discharging times. Do you have access to a capacitometer of any kind? You really need it.

I would also suggest to prepare thin capacitors (less than 1mm thick and a few cm^2 surface ) - for example with your graphene type product mixed with an electrolyte or epoxy glue instead of the polyethylene bags - between glass surfaces with metal electrodes.

Try to make capacitors first, supercapacitors will follow! Keep us informed about your next steps.

quicksilv3rflash (author)  tholopotami3 months ago

I have reposted the lab notes with the "correct" .txt extension. I have the parts necessary to make an RCL meter, and will do so as I continue with these experiments. Is it really necessary to use Pyrolytic Graphite as feedstock for this process? I'd think that something like carbon fiber would be faster to exfoliate due to its higher surface area. If I'm making capacitors, I should be able to tolerate quite high concentrations of impurities in my exfoliation product. Also, the self-discharge of this capacitor was *so high* that I don't know that a regular capacitance meter would give accurate results.

Thanks for the data. I went through your numbers and I think that the
discharging is faster than logarithmic. It becomes logarithmic only for voltages smaller than 0.9 volt in the time interval 120-480sec. From this data the time constant is T=RC=1161sec and for R=10MOhm -> C=0.12mF. If you apply this to a LED with r~200Ohm it will discharge in 0.02sec. Here is the plot:

plot.jpg
quicksilv3rflash (author)  tholopotami3 months ago

"... becomes logarithmic only for voltages smaller than 0.9 volt in the time interval 120-480sec. ..." This may be because my capacitor is restricted to 0.7V or so to avoid excessive self-discharge. I tried 2.2V because

this page

http://www.tecategroup.com/ultracapacitors-supercapacitors/ultracapacitor-FAQ.php

on electrochemical supercapacitors listed a 2.7V maximum, and I've seen 2.5V as the typical minimum rating for commercial supercaps. My capacitor does not appear to match commercial specifications, unfortunately.

C=0.12mF refers to millifarads not microfarads. So my estimation is C=120microFarads.

My suggestion - as well as of other people in this post- is to concentrate
in making thin capacitors, charge accumulation does happen mostly
close
to the interface. The fact that you selected these two bags may be an
explanation of the large transient currents through the bulk when it
is discharged (from 2.2V to 0.9V).

One possible idea: Make a
sandwich metal plate electrode/ thin epoxy glue on the
plate/graphene/thin food plastic membrane (with epoxy glue on both

sides?) / graphene/ thin epoxy glue on the plate /metal plate electrode.
For reference and comparison you could select the surface equal in size
to that of your bag capacitor.

I would press it on a vice and let it settle.

The structure above is a simple capacitor. The next step would be to make another device substituting the central membrane with a paper washed in electrolyte.

In
any case monitor current vs time through a known ressitance both when
charging and when discharging to get an idea about your device. In the
charging phase I would select e resistance in the 100K-1M range.

In case you wonder about my interest , I work in microelectronics and I
have some experience with oxide capacitors with leakage as well as with
composite organic/oxide /semiconductor capacitors. I look forward
to see new results from you.

quicksilv3rflash (author)  tholopotami3 months ago

Thank you for plotting my data! I think the capacitance may be more than 120nF, though, because the EPR (Equivalent Parallel Resistance, which governs self-discharge rate) of electrochemical double-layer capacitors tends to be very high. I would be very surprised if the EPR of this capacitor were higher than 100kOhm, which means that only using the meter's input resistance will give an erroneously low value. I don't have good data for this, though, which puts this experiment firmly in the proof-of-concept realm.

TL;DR: I'll build more of them and take much more careful data on their electrical properties. I hope you all have success in your own tinkerings with this material.

From the technical point of view it would be worth making thin capacitors putting the graphene like material in an insulating polymer matrix.

A capacitometer would help you measure the capacitance of an uncharged capacitor and also measure the leakage current. There are cheap multimeters /capacitometers.

However the most interesting thing to look at is the nature of your exfoliation material. Is it graphene , graphite , carbon clusters or a mixture of the above? Dissolve them and check them with an optical microscope at least.

Kiteman3 months ago

Hardly a "super capacitor", but it's a start!

Have you had any thoughts of where to take your research next? I would start by changing the separator material (thin polythene bag?), then changing the overall area of the two halves.

quicksilv3rflash (author)  Kiteman3 months ago

Next, I don't mix nonconductive cellulose with my graphene product, as I suspect this setup doesn't allow enough electrical contact between the tiny bits of graphite.

Try thin layers, as the charge is stored on the surface of the conductor, not in the bulk material.

quicksilv3rflash (author)  Kiteman3 months ago

I thought loosely-packed bulk material would enable ion sorting as in an electric double-layer capacitor -- would having thin layers produce much more capacitance than just using metal foil? Are the layers "fuzzy," allowing much higher surface area than their macroscopic area would suggest?

Pass.

I think you're onto an interesting avenue of research - keep it up, and keep us updated.

jschnabel3 months ago

Good job! I would recommend you write your measurements in farads, the measurement of capacitance, it would give much more information on your capacitor. Also to increase your capacitance make the sides into two flat plates, this increases the surface area of the capacitor.

Jan_Henrik3 months ago

Nice project, but what is now the capacity of the capacitor?

quicksilv3rflash (author)  Jan_Henrik3 months ago

So low that I couldn't get it to light an LED. Very high self-discharge above 0.7V. Proof of concept only with this method, unfortunately. The graphene needs to be processed differently, I think, to be a more usable capacitor electrode material. I remember reading somewhere that graphene powder could be compressed and sintered...