• Blog Stats

    • 85,300 hits
  • Archives

  • Enter your email address to subscribe to this blog and receive notifications of new posts by email.

    Join 164 other followers

  • Copyright notice

    This blog entry and all other text on this blog is copyrighted, you are free to read it, discuss it with friends, co-workers and anyone else who will pay attention.

    If you want to cite this blog article or quote from it in a not for profit website or blog then please feel free to do so as long as you provide a link back to this blog article.

    If as a school teacher or university teacher you wish to use content from my blog for the education of students then you may do so as long as the teaching materials produced from my blogged writings are not distributed for profit to others. Also at University level I ask that you provide a link to my blog to the students.

    If you want to quote from this blog in an academic paper published in an academic journal then please contact me before you submit your paper to enable us to discuss the matter.

    If you wish to reuse my text in a way where you will be making a profit (however small) please contact me before you do so, and we can discuss the licensing of the content.

    If you want to contact me then please do so by e-mailing me at Chalmers University of Technology, I am quite easy to find there as I am the only person with the surname “foreman” working at Chalmers. An alternative method of contacting me is to leave a comment on a blog article. If you do not know which one to comment on then just pick one at random, please include your email in the comment so I can contact you.

  • Advertisements

What is solvent extraction ?

Dear Reader

You might ask the questions of what is the liquid-liquid extraction of metals and why is it important ?

I would say that the liquid-liquid extraction of metals is the more correct term for what is normally understood as “solvent extraction”. Solvent extraction is not the extraction of a solvent from a thing, but instead it is the extraction of a substance from one solvent into another. Commonly one of the liquids is water but there is no reason why the ideas of solvent extraction should not be applied to non aqueous systems such as silver being extracted from molten lead into molten zinc or the extraction of an organic species from methanol into hexane.

A typical metal ion in an aqueous phase is bonded to water molecules and/or chloride anions, this forms a water soluble metal complex. The metal complex normally has a very low solubility in an organic layer, but like many rules some exceptions do exist.

For example osmium(VIII) forms an organic soluble oxide (OsO4) which can be thought of as having formed from a Os8+ cation and four water molecules. The waters would have bonded onto the osmium before undergoing hydrolysis to hydroxyl ligands (OH-) which then react further to form oxide ligands. I suspect however that [Os(H2O)n]8+ is an impossible complex for several reasons so I do not think it will be possible to make it.

But I know that compounds such as [OsO3(OH)2] are well known, and this can be thought of as a hydrated form of OsO4, or alternatively OsO4 can be viewed as the acid anhydride of [OsO3(OH)2]. Yet another alternative is to stop thinking for a moment about osmium and then move onto some other things.

Well back to the solvent extraction of metals, if we take a hooks and eyes view of chemical bonding the electron poor metal ions can bond to the electron rich parts of solvent molecules such as the oxygens in water. These bonds to water can be broken and replaced with bonds to molecules with big long fatty groups, for example charged cobalt atoms in water are normally found with six waters around them. For example using 3D x-ray crystalovision (known to chemists as X-ray crystallography) we can glimpse a charged cobalt bearing six waters in the way it is thought to be in aqueous solution.

Cobalt with six waters

Cobalt with six waters

This cobalt ion was seen in a paper by some gifted chemists who made coordination complexes with tetrahydrofuran-1,2,3,4-tetracarboxylic acid.[i] I have to admire their compound, in some ways I wish I had done that chemistry myself.

Now it is well known that cobalt(II) salts in dilute water solutions are pale pink in colour, this is a sign that the cobalt is in a octahedral complex with six waters attached. However under some conditions deep blue complexes of cobalt are formed, this is normally a sign that a tetrahedral complex of cobalt as formed. It is well known that the solution of cobalt in bis-2-ethylhexyl hydrogen phosphate in a hydrocarbon solvent is deep blue. This made me want to look to see if X-ray crystallography has ever been used to characterize the idea mononuclear (complex with a single metal atom) of cobalt with two dialkyl phosphate ligands.

When I did the search I was in for a shock, I found a coordination polymer of cobalt and bis tert-butyl phosphate in which tetrahedral cobalts were linked by the dialkyl phosphates into a 1D chain. I also found bis-(((4,7,7-Trimethyl-3-oxobicyclo[2.2.1]hept-2-yl)phosphinato)-((4,7,7-trimethyl-3-oxobicyclo[2.2.1]hept-2-yl) hydrogen phosphinato))-cobalt(II) which is made from one cobalt(II) ion, two molecules of bis-(4,7,7-trimethyl-3-oxobicyclo[2.2.1]heptan-2-yl) phosphinic acid and two molecules of the conjugate base of this rather long named phosphinic acid.

I looked in the Cambridge database for metal complexes in which two dialkyl phosphate ligands chelate to a single metal atom to form a complex. None were found, I soon concluded that the oxygen oxygen distance is too large as the phosphorus atom is tetrahedral, this is different to a carboxylate where the central carbon is trigonal planar.

Here are the XYZ coordinates of a typical model phosphorus compound

O(1) -1.368 1.517 -0.000
P(2) -0.171 0.647 -0.000
O(3) 1.160 1.562 -0.000
C(4) -0.185 -0.423 1.516
C(5) -0.185 -0.424 -1.516
H(6) 1.876 0.950 -0.000
H(7) 0.715 -1.078 1.517
H(8) -1.102 -1.054 1.516
H(9) -0.177 0.220 2.424
H(10) -1.091 -0.199 -2.122
H(11) 0.726 -0.221 -2.122
H(12) -0.199 -1.495 -1.213


The two oxygen atoms are 2.5 Å apart,  now we can compare the distance with that in acetic acid. Again here is the xyz table, and in acetic acid the oxygens are 2.2 Å apart which makes it much more easy to form a chelate ring.

C(1) -7.031 -0.130 -0.000
C(2) -5.522 -0.130 -0.000
O(3) -4.918 -1.176 -0.000
O(4) -4.853 1.029 -0.000
H(5) -7.403 -1.179 -0.000
H(6) -7.402 0.394 0.909
H(7) -7.402 0.395 -0.909
H(8) -3.910 0.796 -0.000


Now some of my readers may ask “what is a chelate ring”, the idea of chelation is where a molecule bonds to a metal with more than one atom. I would like you to do a thought experiment, we have gone to the beach and a fish has bitten my foot, now how do I remove it ?

I have to open its mouth and then I can pull it off and throw it back in the sea.

Now imagine that the crab has grabbed by foot, now to remove the crab I must release both of its pincers at the same time. If I remove only one at a time while I am dealing with the second one the first one will grab me again. The crab is able to chelate to me.

Things get even worse with a scorpion which has two front grabbers and a tail to grab with, and frankly if the octopus gets me with all eight tentacles then it is going to be very hard to escape from its grasp.

If a ligand is able to chelate to a metal then it normally makes it very hard to remove the metal from the ligand’s grasp.

Thinking about the extraction of cobalt we might suspect that it should be given by the following equation

DCo = k[(RO)2PO2H]2[(RO)2PO2]2

I suspect that the (RO)2PO2H will form hydrogen bonded dimers in the organic phase which could alter the equation to

DCo = k[{(RO)2PO2H}2]1[(RO)2PO2]2

This might be true when the concentration of (RO)2PO2H is very low, the problem will be that when the concentration of (RO)2PO2H is high then it increases the dielectric constant of the organic phase, this will disfavor the extraction of the lipophilic (fat loving) cobalt complex [Co((RO)2PO2)2 ((RO)2PO2H)2]

This will make the equation which relates the distribution ratio closer to

DCo = k[{(RO)2PO2H}2]0[(RO)2PO2]2 = k[(RO)2PO2]2

As the concentration of (RO)2PO2 depends on the proton concentration we can make the equation slightly more complex and more useful.


Ka = [(RO)2PO2][H+]/[(RO)2PO2H]


Ka / [H+] = [(RO)2PO2]/[(RO)2PO2H]


Ka [(RO)2PO2H] / [H+] = [(RO)2PO2]

So next

DCo = k{[(RO)2PO2H]}2 ka2 / [H+]2


DCo = k{[(RO)2PO2H]}2 ka2 [H+]-2

This equation does explain how to change the distribution ratio by altering the concentration of acid, by making a change of pH we can adjust our D value. Now as my physics teacher told me we should use dimensional analysis to decide if the equation is good or bad.

Now DCo = [Co]organic / [Co]aqueous

So the D value has no units

So the k value has the units of mol-2 dm6 as [(RO)2PO2] has the unit of mol dm-3

OK what next.

Ka has the units of mol dm-3

So we have

mol-2 dm6 . mol2 dm-6 . mol2 dm-6 . mol-2 dm6 = mol-4 dm12 . mol4 dm-12 = mol0 dm0

Which finally confirms that the equation

DCo = k{[(RO)2PO2H]}2 ka2 [H+]-2

Is a viable equation which is at least mathematically correct.

At this point I think we can call it a day and have another solvent extraction lesson another day.


[i] Liang-Fang Huang, Chang-Chun Ji, Zhen-Zhong Lu, Xiao-Qiang Yao, Jin-Song Hu, He-Gen Zheng, Dalton Trans, 2011, 40, 3183.


Go on, Have your say !

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

WordPress.com Logo

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

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s

%d bloggers like this: