Sciencemadness Discussion Board - Preparation of ferrofluids

Preparation of ferrofluids

math - 5-10-2012 at 03:26

Hello,

this topic may be small matter for most of you, but I've been recently interested in ferrofluids and would like to know more on their homemade preparation.

From what I've read, ferrofluids are a suspension of tiny magnetite particles, created from the reaction between FeCl2 and FeCl3 in the presence of NaOH.

Also, to avoid clumping, a surfactant is added, like tetramethylammonium hydroxide or citric acid.

Starting from iron metal, aq. HCl, NaOH, aq. NaClO and citric acid, do I have enough to synthetize some?

Could I dissolve iron metal in HCl to yield FeCl2, then bubble some Cl2 into this solution to get FeCl3 and finally mix both chlorides together with NaOH and get a ferrofluid?

Fe(s) + 2 HCl(aq) → FeCl2(aq) + H2(g)
2 FeCl2(aq) + Cl2(g) → 2 FeCl3(aq)
FeCl2(aq) + 2 FeCl3(aq) + 8 NaOH(s) → Fe3O4(s) + 8 NaCl(aq) + 4 H2O(l)

Some websites advise using aq. NH3 as base, though since it's weaker I guess using NaOH would need less?

Also do aqueous ferrofluids behave like commercially available ones, which I believe are kerosene suspensions?

Thank you

Vlassis - 5-10-2012 at 04:38

I dont know anout the ferrofluid, but you dont need to bubble Chlorine to get FeCl3.In the mixture of FeCl2 and excess hydrochloric acid, you can add some Hydrogen Peroxide, to oxidize further the FeCl2 to FeCl3.Or, you can wait for a few days, since the FeCl2(yellow) will get oxidized by air to yield FeCl3(orange)

[Edited on 5-10-2012 by Vlassis]

encipher - 5-10-2012 at 06:15

math,

You can buy PCB etching solution (FeCl3) from some place like radio shack. Then react it with iron metal to get FeCl2.

2 FeCl3 + Fe → 3 FeCl2

Once you have both FeCl2 and FeCl3, You can form magnetite through co-precipitation.

Fe2+ + 2OH- → Fe(OH)2
Fe3+ + 3OH- → Fe(OH)3
Fe(OH)2 + 2Fe(OH)3 → Fe3O4 4H2O

As for the source of OH-, (ammonia vs NaOH) there are two things to keep in mind. The rate at which you drip the base into the solution and the strength of the base. NaOH being stronger may result in the formation of larger particles, which is undesired and can lead to agglomeration, due to strong local concentrations at the location of the drop.

The issue of drip rate: you should have a very slow drip rate, again for the same reason as above. Increased stirring is also better.

As for TMAOH, it is quite difficult to acquire as an amateur chemist, as opposed to making a non-aqueous ferrofluid. If you already have TMAOH then you are set. If not, it is much easier to get oleic acid as a surfactant and take your pick of n-alkane or kerosene as a medium.

Hope this helps. For what it's worth, I've made ferrofluids quite successfully both aqueous and non-aqueous.

math - 5-10-2012 at 12:54

Easiest option for FeCl3 would be as Vlassis suggested, to add some H2O2 to aq. FeCl2 in HCl.

I can't get TMAOH or oleic acid, so I'd just use citric acid, if it works as a surfactant.

Since you've successfully made both aq. and non-aq. ferrofluids, which one do you think behave better, aesthetically speaking, when exposed to a magnetic field?

Could I expect from a homemade aqueous ferrofluid similar performance to a non-aq. commercial one?

Thank you

encipher - 5-10-2012 at 18:10

I guess depending on your starting materials your route may be fine for you. I made mine via the route I described above, back in '06. Speaking of which, I think I'd posted a topic about it from back then. Here's are photos of the ferrofluid I made.

I have since made better ferrofluids - that was my first attempt at it. Aesthetically speaking the medium doesn't matter. You can get pretty ferrofluids both ways.

What matters is the final particle size of the magnetite. If they are too large you will venture into MR fluid territory, which "works just fine" but isn't pretty. So focus on making sure it is very well stirred, heated, and don't do anything too fast. If

math - 6-10-2012 at 15:07

Thanks for your reply and nice pictures

I can now understand importance of smaller particle size and stirring (as well as NaOH dilution), but you also talked about heating, while I wasn't aware of any external heating needed for this reaction. Could you give me more details?

Thank you

encipher - 6-10-2012 at 16:29

Well heating is primarily used when using NH4OH as a base to drive off the ammonia. In your case it may not be necessary. Check out the experimental section in this paper: http://www.eng.buffalo.edu/~swihart/Reprints/Goodarzi_MatRes...

encipher - 6-10-2012 at 16:35

Going back to your original concern, you should also note that the surfactant used can affect the size and "geometry" of the spikes, aesthetically speaking. Depending on the mode of action of the surfactant (electrostatic, etc..) and the size of the molecule (oleic acid, stearic acid, citric acid, etc..). I don't think it will be aesthetically displeasing, one versus the other. It will just produce different styles. Like in my photo the spikes are very thin and needle-like, and another synthesis will produce wider, more round spikes.

MrHomeScientist - 9-10-2012 at 06:05

I tried my hand at making a kerosene-based ferrofluid using ammonia as the base and oleic acid as the surfactant. Here's a couple videos on my process. In the end I didn't get any spiking, so i may not have stirred vigorously enough. Using a magnetic stir bar also probably didn't help

Making iron(II) and iron(III) chloride - http://www.youtube.com/watch?v=iWpfHkWr5DY&feature=plcp
Making the ferrofluid itself - http://www.youtube.com/watch?v=LlQw9dfexBQ&feature=plcp

watson.fawkes - 9-10-2012 at 08:38

Quote: Originally posted by MrHomeScientist

Using a magnetic stir bar also probably didn't help

That's interesting. Did you get any sense that your particle size was larger than it should be?

encipher - 9-10-2012 at 10:37

Quote: Originally posted by watson.fawkes

Quote: Originally posted by MrHomeScientist

Using a magnetic stir bar also probably didn't help

That's interesting. Did you get any sense that your particle size was larger than it should be?

I think he may be referring to the fact that magnetite is magnetic, and he used a magnetic stir bar.

MrHomeScientist, I think a big issue is the mixing of the reagents. In my first attempt (and all subsequent ones) I dripped in the base using a buret *very slowly* with gentle heating. I was able to produce spikes successfully on my first try as you can see above.

In the video it appears that your particles are agglomerating but they are still "wetted". There are a couple reasons I can think of immediately that may cause this, but the main one that comes to mind is too large of a particle size.

The fact that some quantity of carrier is pooling with the magnetite indicates some degree of surfactant has successfully coated the particles. However, if the particles are too large then there is less surface area, less particles and therefore less overall volume. Large particle size also means no "spiking" - what you have there is probably some form of MR fluid, or even beyond that ( > D90).

Of course just judging from the footage there appears to be much magnetite looking particulate in the main solution so there are other issues going on as well.

Try it again, with moderate heating, slow dripping, extend the time and see what happens. I think with that in mind it should be pretty fool proof.

Hope this helps.

watson.fawkes - 9-10-2012 at 12:17

Quote: Originally posted by encipher

I think he may be referring to the fact that magnetite is magnetic, and he used a magnetic stir bar.

Yes, I'm sure he was. My suspicion is that the external magnetic field (external to the fluid, not the container) might drive particle agglomeration more than normal. The idea is that an outside field would reduce "orientation noise", that is, it would tend to align magnetic domains along the field, leading to more collisions that increased particle size over those that didn't. It's only an idea, though, and I was asking if there was any evidence for it.

arsphenamine - 9-10-2012 at 22:39

Ferrofluid particle aggregation could be abated using ultrasound,
much as is already done in zeolite manufacture, specialty ceramics,
nano-lubricant molybdenum sulphide, and micro-milled thermite.

In practical terms, you'd put the reaction vessel in a swept ~40kHz
ultrasound water bath and stir mechanically.

encipher - 10-10-2012 at 04:34

Quote: Originally posted by watson.fawkes

Quote: Originally posted by encipher

I think he may be referring to the fact that magnetite is magnetic, and he used a magnetic stir bar.

Interesting thought. Do you think it could be significant compared to a driving "force" like surface area?

MrHomeScientist - 10-10-2012 at 05:11

Thanks for all the comments on my comment! Encipher, those are some good tips. I hadn't considered adding the base in slowly; mostly I thought it was either a stirring issue or the stir bar magnet messing with things. I'd definitely like to try this experiment again.

I suppose the "evidence" for bad particle size is the fact that it didn't spike, but still responded to the magnet. The carrier fluid and surfactant seem to have worked well, because the magnetite was all drawn into the kerosene layer and no clumps form in the fluid.
The magnetic stir bar attracted a lot of the magnetite, however, so being stuck there it may not have been able to be "surfacted" sufficiently by the oleic acid. It's also possible that it attracted a lot of the smaller particles, leaving only the larger ones in solution. Come to think of it, the hot plate / stirrer I used also has a magnet in the plate itself doesn't it? That could have also contributed to disturbing the particles' growth. It's an interesting thought that the magnetic field itself might affect particle size.

watson.fawkes - 10-10-2012 at 07:55

Quote: Originally posted by encipher

Do you think it could be significant compared to a driving "force" like surface area?

It might be. I don't know for sure. What seems clear enough to me is that a static magnetic field can't help. If other means of stirring are available, it would seem prudent to use them.

Quote: Originally posted by MrHomeScientist

The magnetic stir bar attracted a lot of the magnetite, however, so being stuck there it may not have been able to be "surfacted" sufficiently by the oleic acid. [...] Come to think of it, the hot plate / stirrer I used also has a magnet in the plate itself doesn't it?

The surfaction rate is exactly what I was implicitly comparing to. If a surfactant molecule has adsorbed onto the surface of a particle, it's far less likely to have an accretion collision with another particle. Certainly increasing the density of magnetite particles near the stir bar, implicitly decreasing the surfactant density, won't help. You've suggested a second effect over the one I was thinking of. The one that first came to mind was the magnetic field lines formed in loops passing through the stir bar magnet and the stir plate magnet.

Quote: Originally posted by arsphenamine

Ferrofluid particle aggregation could be abated using ultrasound [...]

This seems like the most practical suggestion. Decent ultrasonic baths have dropped greatly in price over the last few years.

arsphenamine - 11-10-2012 at 11:07

Quote: Originally posted by watson.fawkes

Decent ultrasonic baths have dropped greatly in price over the last few years.

Harbor Freight stuff is okay, but fragile.

If possible, acquire a used lab or industrial unit one by Crest or Branson,
and be aware of the distinction between "desktop" and "benchtop" sizes.
Gunsmiths are another user demographic, thus may be a reseller, too.

math - 7-7-2013 at 13:52

Quote: Originally posted by arsphenamine

Ferrofluid particle aggregation could be abated using ultrasound,
much as is already done in zeolite manufacture, specialty ceramics,
nano-lubricant molybdenum sulphide, and micro-milled thermite.

In practical terms, you'd put the reaction vessel in a swept ~40kHz
ultrasound water bath and stir mechanically.

This may be a very silly question, but does an ultrasound bath have to be nested out of the magnetite-reaction beaker or could I for example just play really loud a 40kHz frequency thru speakers onto the mix?

On the other hand, since the ferrofluid appearance and properties seem to be linked to its particles size, smaller being better, I found something contradictive, as rapid mixing of Fe+3, Fe+2 and NH4OH as key for nanoparticles creation (?)

Rapid mixing: A route to synthesize magnetite nanoparticles with high moment http://apl.aip.org/resource/1/applab/v99/i22/p222501_s1?isAu...

while others seem to rely on Fe+3/Fe+2 ratio as pH control which in turn relates to particle size

Synthesis of Magnetite Nanoparticles by Precipitation with Forced Mixing http://link.springer.com/article/10.1023%2FA%3A1010091625981

Synthesis of magnetite (Fe3O4) nanoparticles without surfactants at room temperature http://www.sciencedirect.com/science/article/pii/S0167577X07...

Since some papers (though I don't have full access to them) seem to explain it as overly simple, I can't understand if this is due to lab experiments being extremely difficult by routine or if some of us are failing because a too-small particle size may again impair the ferrofluid appearance as would a too large particle size.

Thank you

math - 8-7-2013 at 15:24

edit: another question about ferrofluids

If salts-containing (aka "hard") water is used to prepare an aq. ferrofluid, would it impair the surfactant and hence the ferrofluid properties? How would the dissolved salt behave and diffuse within the ferrofluid when a magnet field is applied? Would it be in solution in every point in the same concentration (like following the whole ferrofluid movement, being trapped between the magnetite particles) or settle on the bottom while the magnetite particles move and rise according to the magnetic field?

Thank you

Acidum - 9-7-2013 at 12:35

Quote: Originally posted by math

...or could I for example just play really loud a 40kHz frequency thru speakers onto the mix?...

Oh my, this is indeed a very silly question!

First of all, you can not play frequencies above 20 kHz (few kHz give or take, depending on quality) on any "normal" speakers. They are not made for such frequencies simply because nobody can hear them.

Second, in order to transfer sufficient amount of energy from any outside "non-nested" source to the mixture, amount of dissipated energy will be tremendous, and it would cause significant damage to everything around, including your soft tissues.

...that is why people made ultrasonic baths and probes...

PS. You can find such baths in mineralogy or jewelry shops.

math - 10-7-2013 at 11:10

I knew it was a silly question

So if I can get a hold on an ultrasonic cleaner, to get it to work as it's intended, can I put the reaction beaker into the ultrasonic bath or do I have to perform the reaction directly in the ultrasonic bath?

Thank you