On March 24, 1989, the Exxon Valdez ran aground on Bligh Reef in the waters of Prince William Sound, Alaska, spilling approximately 11 million gallons of Prudhoe Day crude oil. The subsequent contamination fouled surrounding ecosystem, impacting many marine organisms. A major effort was then launched to capture and rehabilitate Alaska's native population of sea otters affected by the contamination. Several issues have arisen out of this effort:

 At the time of capture, was the animal contaminated.

 If the animal was contaminated, then to what degree, and how might it affect the health of the animal.

Thin layer chromatography, TLC, is a relatively simple and inexpensive analytical technique that is very useful both in the laboratory and in the field. TLC's greatest asset in the field is its accuracy and quick turnaround time. With control animals as a comparison, differences were determined between squalene (natural polar oils) and the less polar Prudhoe Bay crude petroleum products. Crude oil exhibits very characteristic bonds and can be easily standardized to determine oil concentrations in the otter's coat. This presentation will demonstrate how TLC can be used to determine the more toxic biodegration products that can affect the sea otter coat from contamination of Prudhoe Bay crude oil.

Introduction

Analyzing possible contaminants from an environmental disaster can often be a frustrating and cumbersome undertaking. Field methodologies, turnaround time on results, and costs, typically present hurdles that must be overcome to provide timely and effective information to scientists and technicians in the field.

When material from the oil tanker Exxon Valdez was released into the waters of Prince William Sound, Alaska, the indigenous population of sea otters was severely threatened. Field biologists responsible for capturing and rehabilitating these animals had to make a determination of whether or not the animal was oiled. Once it was determined the animal was in fact "oiled", it was retained for rehabilitation. At that time it became clear that specific methods for analyzing petroleum contaminants was lacking.

Thin Layer Chromatography (TLC) is a recognized quantifiable analytical method that can be readily employed in the field. All of the materials needed are quite portable and can easily fit into a medium size tool box. Costs for running these tests are minimal, and turnaround time is very rapid, 10-20 minutes. This analytical technique can prove very useful in future events of this magnitude.

Methods and Materials

The best application of TLC is its effectiveness for the rapid quantification of chemical contaminants. When used in this manner, a detailed quality assurance/ quality control (QA/QC) program should be implemented to substantiate quantifiable results. (Friedman & Bruya 1989).

Crude petroleum consists of a vastly complex mixture of paraffinic, cycloparaffinic, and aromatic hydrocarbons containing low concentrations of sulfur and trace amounts of nitrogen and oxygen compounds. (Sax and Lewis 1987) As petroleum hydrocarbons take many forms, differences can be distinguished between crude products, diesel oil, hydraulic fluids, and others, as well as their oxidative or biodegradation products. For this reason it was important to prepare standards that matched the suspected contamination as closely as possible. Crude oil from which the standards were prepared was obtained directly from the Exxon Valdez. Petroleum contaminants detected in sea otter fur reflect components of the crude oil released into Prince William Bound.

Sea otter hair samples were collected under two categories, pre and post wash. Pre-wash hair samples were obtained by field biologists in capture boats and by the husbandry staff at the rehabilitation center in Seward, Alaska. Post wash hair samples were obtained by the husbandry staff after the animals were washed with Dawn dishwashing soap and water (16:1 ratio), and rinsed with fresh water. Hair samples obtained were collected from the back of the neck, xyphoid region, and head of the femur. All samples were placed in various non-contaminated containers and refrigerated or frozen for future analysis.

For this analysis a glass plate coated with 60 A silica gel was selected. This silica gel is the most common and versatile solid support for TLC. Tanks used for developing the TLC plates were 6 x 12 inch glass blocks containing approximately 1/4 inch of a selective eluding solvent. Due to the polarities of the various components of crude petroleum, hexane was chosen as the primary eluding solvent. Secondary eluding solvents include methylene chloride or methanol, or a mixture of the two. Calibration standards (Table I) were prepared for comparison against unknown contamination in hair samples.

Table 1. Calibration Standards

Different combinations of the prepared standards were used to best match the suspected concentration of the contamination.

To eliminate any bias on the part of the analyst, samples were selected on a random basis without any prior knowledge as to the condition of the animal, where it was captured, suspect degree of oiling, and final disposition.

A small amount of fur was placed into a tared 8 dram vial and weighed. The weights of the samples varied from 30mg. to 250mg. The amount of hair used for analysis was frequently restricted by the amount of sample available. Ideally a 1:1 sample to extraction solvent ratio would be employed to eliminate the need for a dilution compensation. With the minute amount of sample material available, this was not feasible. After being weighed the samples were extracted with 3mls of hexane. After several trials with different extraction volumes, 3mls was deemed the best to still observe detectable contaminant levels, and to compensate for dilution. Hexane was chosen for the extraction solvent as well as the eluding solvent. In addition to petroleum hydrocarbons that may be present, the animal will exhibit characteristics of natural oils, squalene, that may be quite polar in nature. Using hexane instead of a more polar solvent such as methylene chloride may elude the less polar hydrocarbons while the polar natural oils at the origin of the plate. Using methylene chloride after the plate has been eluded with hexane, will draw the more polar oils through the silica gel. Analyzing the samples in this manner, should help eliminate interferences when determining levels of hydrocarbons present.

Once the sample has been thoroughly agitated during the extracted phase, it is ready to be spotted on the TLC plates. TLC plates were prepared and analyzed according to Friedman and Bruya (1989). Reporting values were obtained by comparisons of levels obtained from hair samples to known standards. If the extraction and spotting volume of the unknown hair samples were different than those of the standards, the reporting value was corrected by using the dilution ratio to obtain the true reporting value.

For each sample extraction prepared an analytical replicate was run, so the precision of the sample spotting could be measured. To ensure that contaminants found in the hair samples matched that of the standards prepared from the crude oil, Rf values

After Rf values and corrected reporting values were obtained, a Moro detailed record of the animals coat conditioning and degree of oiling could be directed.

Results

At this point, hair samples have been tested from 36 different otters and 2 control animals. Levels of oil contamination for the 36 otters ranged from undetectable to 135,000 ppm (mg./L). Based on reported levels, 3 distinct groups of otters can be seen. Those that were contaminated at 6,000 ppm or less including those in which no contaminants were detected, those contaminated at 6,000-20,000 ppm, and those contaminated greater than 50,000 ppm. The latter group contained 4 otters, the middle group 9, and the first group 23 otters.

Table 2 demonstrates how samples were set up and analyzed. In all, 10 plates were analyzed with 4-5 hair samples and replicates per plate as well as comparison standards and method blanks. As stated earlier it is important to know whether or not the suspected analyte corresponds to that of the standards. Peak Rf values of the analytes matched almost identically to those of the standards indicating the contaminants were a match for the crude oil.

Table 2 Contaminant Levels and Rf Values of Four Sea Otters

Initially, standards of 50,000ppm and 100,000ppm were run. When it became evident that suspect hair samples indicated levels well below this, the standards were deleted from the analysis.

It should be noted that all results are preliminary and mope in-depth work is continuing to improve this method and establish a larger data base with which more concrete information can be derived.

In all there are over 280 hair samples from 80 different animals of which only 36 have been analyzed to date. An inconsistency that may arise from capture and sampling is the location on the body from which the sample was taken. If only certain parts of the animals body retain oil, then testing for, petroleum hydrocarbons may be a hit or miss proposition. As part of the on going work being performed, differences in contaminant levels are being looked at from hair corresponding to the 3 sampling regions of the body. This should establish a standard and consistency in future sampling and analytical endeavors. It is also important to obtain a sample that is representative of the animals coat. This should include the guard hair and undercoat.

At the time of the oil spill there was no known baseline work on record for coat conditioning on native Alaskan otters, thus there were no control animals for comparison with oiled and stressed animals. In 1988, the Marine Animal Resource Center received in good condition, a corpse of a large male otter that had been shot in Cordova, Alaska. To our knowledge this is the only otter on record that was obtained "pre-spill". This animal was used as the control animal to assist in oil level comparisons of suspect hair samples. These control samples also served as a back-up in verifying cross-contamination from extraction and spatting methods. These samples indicated non-detectable levels of contamination.

This method of oil analysis is not a single method for determining oil contamination, and coat condition, but can serve as an aid to biologists in the field when making decisions concerning the disposition of the animal. TLC will not indicate whether an animals pelage is light, medium, or heavily oiled but will give a concentration that will serve as a starting point in determination of the animals overall coat condition.

Thin Layer Chromatography was not a test in use in the days following the oil spill. It was a test employed in a laboratory setting on hair samples obtained for future analysis. It has proved itself many times in the environmental hazardous waste industry, and has shown itself a useful tool for analyzing crude petroleum in hair samples. Given the portability and low cost of running this analysis, it should be considered as a plausible method in the future.

References

1.  Christian, G.D.: Analytical Chemistry, 3rd ed. New York: John Wiley and Sons, Inc., 1980.

2.  Friedman, A.J; Bruya, J.E., On-site analysis of environmental contaminants using Thin Layer Chromatography. Seattle: Friedman and Bruya Inc.,1989.

3.  Sax,I.N.; Lewis Sr., R.J., Hawley's Condensed Chemical Dictionary, 11th ed. New York: Van Nostrand Reinhold Co. Inc. 1987.