Formation of Membrane-Like Material

Further features and advantages of the invention will become more readily apparent from the following description of experimental work and of the application of the method to the decontamination of dioxin contaminated soils as well as to the recovery of oil or bitumen from tar sands.

A first experiment was conducted to investigate the nature of the membrane-like material. A solution of methylene chloride and bitumen was placed in a beaker and water was admixed to form an interfacial membrane-like material. Bubbles of air were introduced from underneath so as to rise up through the membrane-like material and establish bubbles of air on tethers of membrane-like material extending a few millimeters in length. Additional bubbles were added below the membrane-like material, which were then caused to move to the bottom of a tether with a bubble above it. The lower bubble was then observed to distend the tether, and move up inside it to join the bubble at the top. The experiment showed that the tether was a flattened tube consisting of two sides face to face, and that the inner material, never exposed to water, but only to solvent, bitumen and air, remained separate and distinct. Thus, the membrane-like material has a water side and an oily side, each defining a surface. As long as the membrane-like material remains in the water, the material remains stable, but if it is removed through the air interface, it dissociates by loosing its solvent component. If it is disturbed as with fragments of membrane-like material settling loose on the surface of the membrane, these fragments eventually recombine with the membrane-like material.

A second experiment was performed to determine if the bitumen was responsible for the formation of the membrane-like material, together with the solvent and water, or if some specific component of the bitumen was responsible and to determine whether or not such a component was extractable. A quantity of bitumen was extracted with methylene chloride from a tar sand and left standing in a beaker for several weeks without the addition of water. The solvent gradually escaped and a skin formed on the surface of the bitumen. This skin layer was removed and saved. A sample of the bitumen material left after removal of the skin was added to methylene chloride and a water layer added on top. It was unexpectedly discovered that no membrane-like material could be formed. Only an interfacial film was formed, and the membrane-like material did not appear. Upon returning the skin layer to the bitumen and solvent solution with added water, the membrane-like material once again formed as before. Thus, a specific ingredient of the bitumen was found to be responsible for the formation of the membrane-like material.

A third experiment was made to determine if solvents, other than methylene chloride, and polar liquids other than water would operate in the practice of the invention. It was found that the following solvents or random combinations thereof would produce membrane-like material:

Methylene Chloride

Trichlorethylene

Chloroform

Perchlorethylene

Carbon tetrachloride

Trichlorofluoromethane

Dichlorodifluoromethane

FREON TF

p-Xylene

On the other hand, it was found that the following solvents which are often disclosed in the art as being suitable for the solvent extraction of oil sand, tar sand, or oily compounds would not produce membrane-like material, but that these could be used as diluents prior to the formation of the membrane-like material, thus reducing the amount of the membrane-like material forming solvent from the above list that must be used:

Gasoline

Kerosene

Fuel Oil

Naptha

Ether

m-Xylene

o-Xylene

Hexane

Toluene

The isomers of xylene were unexpectedly found to be distributed between both the above lists.

It was also found that the following polar liquids would produce membrane-like material:

Water

Ethyl alcohol when used with the appropriate solvent and membrane-like material forming component, but the following would not:

Ammonia.

A fourth experiment was performed on a sample of oil saturated sand from Bakersfield Calif., where the clay fraction was later determined to be 40% by weight. The sand and clay were contacted with methylene chloride and agitated so as to bring the oil into solution with the solvent. Water was added slowly with a wand so as to form membrane-like material throughout the mass, and the excess water was allowed to collect on top. The sand and clay were carefully extracted from the bottom and were found to be water wet and oil free. It was found on several repeats of the experiment that an optimum concentration of solvent solution existed at which the membrane-like material would form, and then move through the mass of sand and clay as an interfacial zone. Too much solvent and the material was weak and would not separate cleanly from the sand, and particularly from the clay. Too little solvent and the interfacial zone would break up and form isolated regions enclosing sand, clay and oil. At the correct concentration of oil or bitumen, it has been found that the separation is clean and complete. This optimum concentration occurs when the concentration of oil or bitumen is low so that the solvent is nearly clear for this particular material. It was also found in several repeats of the experiment that if the sand mass was agitated the mass of sand increased in bulk due to the formation of membrane-like material forming globules which were then coalesced by applying moderate shearing forces to induce settling. If the introduction of the water from underneath was done sufficiently slowly, then such bulking could be avoided. Thus, it was confirmed that the membrane-like material did form with its water side towards the substrate. However, several failures of the experiment occured, indicating the need for further experiments.

A fifth experiment was performed to check if surfactants, used in the fourth experiment to clean the laboratory glassware, were interfering with the separation and causing the failures referred to in the fourth experiment. These surfactants are specifically those used for oily compound separation from surfaces by solubilization. The surfactants consist of an oil soluble part and a water soluble part, which may be anionic, cationic, zwitterionic or non-ionic. These parts are conventionally referred to as the heads, usually water soluble, and the tails, usually oil soluble. The fourth experiment was repeated with the same tar sand and oil sand but with the addition of the commercial surfactants commonly found in glassware cleaning formulations, such as ammonium lauryl sulphonate and sodium lauryl sulphonate. In all cases, the experiment failed, and an emulsion was formed. The following explanation was developed: The few layers of oily compound and solvent mixed with them that lie adjacent to the substrate form a convenient oil layer that allows the oil soluble tails of the surfactant to penetrate, leaving the water soluble surfactant heads exposed. Since the oily compound, now including the surfactant, is attached by coulombic forces to the substrate, and the surfactant is likewise held, the membrane-like material, if it forms at all, does so with its water side out and a water layer forms outside the oily layer, at a location determined by the surfactant heads thus effectively inhibiting the separation of oily compound from the substrate by preventing the water from reaching the substrate. Thick oily layers have been observed on the sand substrate, with the consequent loss of large amounts of solvent, when surfactants are present in even small amounts. Thus, the formation of the membrane-like material is inhibited by at least one observable mechanism, namely when surfactants are present. Upon repeating the fourth experiment, but taking care to eliminate surfactants, the failures referred to disappeared and the results became consistent with clean separations being achieved with a wide range of oily compound source material.

A sixth experiment was made to further define the properties of the membrane-like material in liquids, without the presence of a substrate. The conditions of the first experiment were repeated but with the separation of the membrane-like material effected by bubbling air through the interfacial zone so as to extend the membrane-like material into the water layer, and then using a wire brush to capture the material and deposit it into a vessel containing fresh solvent. Repeating the extraction from the fresh solvent several times resulted in a relatively concentrated material. This material was then added to fresh solvent, with a water layer added on top. The liquids were then shaken together and a zone of membrane-like material appeared at the top of the solvent layer and at the bottom of the water layer. It was unexpectedly discovered that all of the brown color, left over from the extraction and concentration of the material had been incorporated into the membrane-like material. After some minutes, the brown materials, residues of the original bitumen, diffused back into the solvent until nearly the original color was restored. This was unexpected. The mixture was shaken again, and this time the color was unexpectedly not taken up by the membrane-like material, but rather remained in the solvent. The formation of the membrane-like material thus involves a non-reversible step. Repeating the experiment with oil soluble dyes placed in the solvent showed the same effect; the dyes were first incorporated into the membrane-like material, and later at least partially released. A repeat of the experiment with dioxin showed similar results, that is, there was an affinity between the membrane-like material and dioxin; however, unlike the dyes, the dioxin remained absorbed in the membrane-like material.

A seventh experiment was performed to determine if the specific component derived from bitumen used in the second experiment and found to be responsible for the formation of the membrane-like material could be shown to exist in bitumen and crude oil from sources other than Athabasca. The extraction and test methods developed in the previous experiment were used. It was found by repeating the experiment that the membrane-like material forming component was present in oil or bitumen from the following sources: Athabasca, Venezuela and Utah at full strength relative to Athabasca. It was also present at slightly reduced strength in samples from New Mexico, Texas, Peru (South America) and Columbia (South America). It was also found in lesser strength in samples from Bakersfiled California. Additional samples of refined oil also showed the effect but with one notable exception, a refined oil from Pennsylvania did not shown any membrane-like material forming component.

An eight experiment was performed to determine the best method to be used to recover oil or bitumen from tar sand, together with the solvents used to recover it. First, a membrane-like material forming component was obtained as in the second experiment. Then, a sample of oil wet tar sand from Utah was extracted with kerosene as a solvent and most of the bitument was removed. Next, the membrane-like material forming component in a solution of methylene chloride was added and mixed thoroughly through the sand mass. Water was added slowly from the bottom so as to cause the membrane-like material to form adjacent the sand. Finally, the sand was washed several times with water and upon examination was found to be free of oil and solvents. An instrumentation examination of the water wet sand showed 19.9 parts per million of methylene chloride remaining, an extremely low value.

A ninth experiment was performed to determine if materials other than those extracted from crude oil or bitumen could be used to form membrane-like material. A quantity of cholesterol was added to methylene chloride and water was added on top. Upon shaking the mixture, the membrane-like material was found to have formed. It was thinner and weaker than that obtained from crude oil, but the extended structure was the same.

A tenth experiment was performed with a view to removing 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) from a sandy soil substrate contaminated with same. Unlike the earlier experiments, the amount of contaminant present was extremely small, of the order of 100 parts per billion. The maximum concentration of TCDD on any of the samples run in the experiment was 10 parts per million. These concentration levels contrast sharply with bitumen or oil samples previously run which have equivalent concentrations of 130,000 parts per million or 13%. The soil was first dried using heat and then 8.0 grams of the dry soil were placed in a separatory funnel; 15 ml of methylene chloride were added and the mixture was shaken. A solution of 250 ml of hexane and 50 ml of methylene chloride with membrane-like material forming component added, was made-up and tested for its activity to form membrane-like material of the desired strength. A few milliliters of the mixture was added to water in a flask and the mixture shaken. When the membrane-like material formed in a sufficient amount to produce at the interface between the water and methylene chloride a visible membrane strong enough to hold air bubbles 5 to 10 mm above the average membrane surface, the mixture was considered to be of sufficient strength. The ratio of hexane to methylene chloride was selected to produce a density of less than 1.0 when 30 ml of the solution was added to the mixture of soil and methylene chloride in the separatory funnel. The 30 ml of solution containing methylene chloride, hexane and membrane-like forming component were added to the separatory funnel, and the mixture shaken. About 50 ml of water were added and the mixture was shaken again. The separatory funnel was set aside for the contents to settle, and several layers formed. The solvent layer was on top, a zone rich in membrane-like material was at the bottom of the solvent layer, adjacent to the water layer underneath, and finally the soil was at the bottom. A sample of each of the layers was taken and analysed by the U.S. Environmental Protection Agency approved GC/MS/MS method for dioxin. Each sample was spiked with 10 ng of isotopically labelled C.sub.4.sup.37 -2,3,7,8-TCDD and 50 ng of C.sub.12.sup.13 -2,3,7,8-TCDD. The test results showed that the original TCDD and the isotopically labelled TCDD compounds could be reliably recovered from all of the samples, except the membrane-like material. No recovery was possible of either the original or the isotopically labelled TCDD from the membrane-like material. Using mass balance calculations, it was estimated that 53.7% of the original TCDD plus 100% of the two isotopically labelled TCDD compounds were absorbed in the membrane-like material. In another run, 42.9% of the original TCDD plus both isotopically labelled TCDD compounds were absorbed in the membrane-like material. The strong absorbtion of the dioxin by the membrane-like material was confirmed. The reduction in the TCDD concentration in the soil was from 79 parts per billion to 7.5 parts per billion.