I also posted this on the other forum, thought you might all find this of interest
In my research to find a inexpensive means to clean waste motor oil I found a series of papers written around 1925 on using sodium silicate to clean it. The gist is that SS of a 1.38 specific gravity is added to WMO at a 1 to 16 ratio.
It'll settle out the particulates to a underlying water layer, evidently with great efficiency and speed.
One paper is here:
http://pubs.acs.org/doi/abs/10.1021/ie50181a009And below is most of the text of another:
Purification of Mineral Oils.
Mineral oils which have been used for lubricating the crank cases of internal combustion engines, for insulating electrical transformers or for oil-immersion switches become contaminated with finely divided carbon which cannot be removed by filtration. The recovery of these oils will, in the future, assume a greater importance than it has in the past.
Van Brunt and Miller found that by agitating mineral oils con-
taining such colloidally dispersed carbon with a relatively concentrated
silicate solution they could cause it to pass completely into the silicate.
As the silicate solution is much heavier than the oil it is a simple matter
to throw the oil upon a body of water and allow the silicate droplets
as they quickly settle out to carry the carbon past the interface into
the aqueous layer, leaving the oil free of suspended matter. It then
remains only to remove the light fractions by heat to obtain an oil ready
for re-use. They worked with oils from the crank case of internal
combustion engines.
Na 2 0, 3.3Si0 2 at various concentrations yielded a sludge which sepa-
rated slowly from the oil at room temperatures, as indicated by Figure
149. It is evident that higher concentrations are more effective. Fol-
lowing this suggestion, Na 2 0, 3.3Si0 2 specific gravity 1.38, was tried in
various amounts. Relations between silicate and oil by volume indicate
that there is nothing gained by using more than 1 : 16 for sedimentation
at ordinary temperatures. Raising the temperature to 80° C. brought the
whole clarification of a 5 cm. layer of oil lying on water within 5-6
minutes and amounts of silicate solution down to 1 per cent were found
to be sufficient for some oils.
Contaminated crank-case oils are not simple suspensions of carbon
in hydrocarbon liquids, and some were found which did not yield to
this treatment, either hy failing to clarify or hy producing a sludge
which was not easily dispersed hy water. All these were brought into
line by adding a mixture of acid manganese resinatc and stearic acid
in the proportion of 1 : 200 and 1 : 1000 respectively, followed by a
1:40 by volume addition of Na 2 0, 3.3Si0 2 , specific gravity 1.38. A
single exception was an excessively dirty sample which required twice
the amount of silicate.
Fig. 149. — Reclamation of Crank Case Oil.
Left hand tube — Oil poured on water. Center tube — After the silicate and
suspended materials have dropped out of the oil and passed into the water layer.
Right — Same after settling.
Better dispersion of the sludges and hence easier operation of
mechanical devices was secured by using a still more concentrated sili-
cate, — Na 2 0, 1.6Si0 2 , specific gravity 1.67.
The method of agitation found to be most satisfactory is to break
up the silicate into small droplets, just enough to give complete contact,
but not enough to form an emulsion which may occur if too much agita-
tion is used in a relatively clean oil. This can be prevented by adding
carbon black. Air was also found to be essential and must be broken
into fine bubbles during the period of agitation. The authors comment
that this is undoubtedly connected with the fact that not only the more
polar bodies in the oil but also the silicate tend to enter the oil-air
interface. They promise a theoretical discussion of the action of the
silicate solution from the point of view of colloid chemistry.
This should be of great interest. The observation may, however, be
made that the process seems to be essentially a wetting of carbon, col-
loidally dispersed in the oil, by a silicate-soap solution heavy enough
to settle rapidly from the oil and soluble enough to be easily dispersed
in water.
The presence of resin or oleic acid would guarantee the formation
at least of traces of soap and acidic materials ; manganese or other
metallic salts would cause a precipitate in the concentrated silicate
which might, like the carbon black, favor the separation. This, of
course, is not a complete description of the observed phenomena, but
the process affords a striking example of the ability of silicate solutions
to wet surfaces which completely repel water, as well as their part in
suspending carbon after bringing it into water. Soap solutions and
other alkaline salts were tried and found greatly inferior to the silicates.
Other workers have used silicate solutions to wet colloidal carbon
and removed the sludge with the aid of centrifugal apparatus. While
oils of satisfactory quality may thus be produced it is neither so con-
venient nor economical to completely free the sludge from oil as in the Van Brunt method.