From Tech Trends September 1996
Novel Slurry Washing For Hydrophobic Organics
By Dr. Peter R. Jaffé, Princeton University
Researchers Dr. Jae-Woo Park from the University of Hawaii and Dr. Peter R. Jaffé from Princeton University are looking for a site to test their soil-slurry washing technique that has decontaminated soils in standard bench treatability studies. The technique is particularly intriguing because it removes low-solubility nonionic organic pollutants such as polychlorinated biphenyls (PCBs) which are not readily amenable to bioremediation or conventional pump-and-treat methods. One of the factors contributing to obstinate removal is that nonionic organic contaminants in the subsurface are effectively sorbed onto the naturally occurring organic materials in the soil. The technique is based on first transferring the sorbed pollutant from the soil to anionic surfactant-coated oxide particles and then separating these anionic surfactant-coated oxide particles with the sorbed pollutant from the soil slurry via a magnetic separation technique.
For the study, two different soils were examined. Soil 1 was a sandy soil with a relatively low organic carbon content. Soil 2 was a topsoil with a higher organic carbon content than Soil 1. The anionic surfactant Emcol CNP-60 was used. The oxide used in the research was Pferrico 3570 (cobalt treated gamma ferric oxide). The nonionic organic contaminant was phenanthrene.
The objective of the research was to: (1) describe the distribution of an anionic surfactant and a nonionic organic contaminant in a system consisting of oxide, soil and water; (2) study the effectiveness of the transfer of the nonionic organic contaminant from soil to the treated oxide in a soil slurry, followed by the separation of the treated oxide from the soil slurry; and (3) investigate the use of this soil-slurry washing technique for practical applications.
To prepare the soil for the bench treatability study, 10 grams of soil and 54 milliliters (mL) of deionized water were put into a disposable glass centrifuge tube, to which was added from 46.4 to 55.6 micrograms of [14C]phenanthrene. The tube was tumbled in a tumbler for 24 hours to facilitate mixing and to allow for a complete sorption of the phenanthrene onto the soil.
The first step of the removal technique consists of adding surfactant-treated magnetite, which was shown to sorb a significant fraction of the pollutant, to the tube, thus transforming the tube into a reactor. For this phase, 1 gram of oxide and 2 mL of Emcol CNP-60 solution (10.68 grams per liter) were added to the tube.
After an equilibration period, the treated magnetite and pollutant sorbed to it are removed using a magnetic separation technique. The researchers assumed linear relationship for the sorption of the anionic surfactant to soils and magnetite over the relevant concentration range; they obtained a simple expression for the surfactant distribution between magnetite, soil and water as a function of the solution pH. Distribution of the nonionic organic contaminant between soil as a function of the sorbed surfactant concentration, magnetite as a function of the sorbed surfactant concentration and water could be described in terms of partition coefficients normalized with respect to the sorbed surfactant mass. The results from actual experiments and theoretical estimates were in good agreement. The mass of anionic surfactant sorbed onto the oxide particles was about 30% of total surfactant uptake capacity of the oxide. The amount of phenanthrene transferred divided by total phenanthrene in the system was approximately 36.1% for Soil 2 and 85% for Soil 1. That is, phenanthrene in Soil 2 went from 5.1 milligrams (mg) per kilogram (kg) of soil to 3.25 mg.; and, phenanthrene in Soil 1 went from 5.1 mg. per kg. of soil to 0.76 mg. These figures are for a single extraction.
Additional extractions could improve
extraction rates.
Numerical simulations were conducted using a simple model that was verified against the batch experimental results. Various scenarios were investigated, illustrating the differences in pollutant removal rates for different oxide- and surfactant-application rates in multistage soil slurry reactors. It was shown that the most significant improvement in the pollutant removal rate could be achieved by using surfactants that have a low degree of loss from the oxide to the soil.
A mathematical model was developed to simulate the operation of a multi-step soil-slurry washing process. The batch-type multi-step soil-slurry washing process can also be implemented as a multi-step continuous flow process. In a continuous flow system, the soil, slurry, oxide and surfactant are applied to the first reactor continuously. As the slurry is pumped to the next reactor, the oxide is removed continuously via magnetic separation; and, a new oxide and surfactant dose is applied to that slurry stream.
The technique is ready to be scaled up to a pilot demonstration using contaminated soils from a site. The researchers are looking for a site that has relatively low organic carbon content, on the order of 1% or less, and has a high octanol water partitioning coefficient on the order of 10,000. The existing patent on the process is held by Princeton University; and, they are looking for financial collaboration to put together the pilot scale demonstration.
If you are interested, contact Peter R. Jaffé at Princeton University by telephone (609-258-4655), FAX (609-258-2799) or the Internet (jaffe@soil.princeton.edu). For more detailed information about the research, see "Phenanthrene Removal from Soil Slurries with Surfactant-Treated Oxides," Jae-Woo Park and Peter R. Jaffé, Journal of Environmental Engineering, June 1995, pp. 430-437.