About TCAW - Subscription Info
March 2001
Vol. 10, No. 03, pp 30–36.
Today's Chemist at Work
Focus: Trace Metal Spectroscopy


Mercury Rising

Deforestation and gold mining in the Amazon basin cause the release of toxic metal.

opening artOne of the less well known effects of soaring gold prices in the late 1970s was the creation of a gold rush in the Amazon basin. While price fluctuations from $100/oz to more than $800/oz literally enriched the lives of many people, the downstream environmental and health effects of gold mining in South America has potentially impoverished the lives of thousands of locals.

Part of the problem stems from how the gold is pulled from the ore. Across Brazil, thousands of garimpieros, itinerant gold miners, remove ore by washing a rock face with a high-pressure stream of water. The ore is then broken down in a hammer crusher, and the gold-bearing ore is sluiced with mercury in a process known as amalgamation. The amalgam is filtered manually and then retorted to release the mercury from the gold. The mercury vapor that results is distilled and reused, although a small fraction remains bound to the gold, to be released by the gold dealers during processing.

Although much of the mercury is recovered in the amalgamation process, it has long been believed that losses due to dumping or evaporation, about 100–150 tons/yr, were the main cause of environmental mercury contamination. Numerous studies supported this theory, showing astronomical levels of mercury in the soils and rivers surrounding the mining districts. There was also the medical evidence of mercury contamination in the garimpieros (see box, “The Mercurial Tea Party”).

Curses, Soiled Again
Marucia Amorim, a cytogeneticist at the Federal University of Pará (Belem, Brazil), was interested in the high mercury levels that she was finding in Brazilian garimpieros and grew concerned when she found similar levels in natives living kilometers from the mines. In 1992, Amorim contacted Donna Mergler, a neurotoxicologist at the University of Québec at Montréal (UQAM) who studied the effects of occupational exposure to environmental contaminants and, together with Marc Lucotte, director of UQAM’s Institute of Environmental Sciences, established a joint project to study the Brazilian mercury situation.

The researchers concentrated their efforts on the Tapajós River, one of the last major tributaries before the Amazon empties into the Atlantic Ocean (Figure 1). The area had long been home to garimpos (gold mines) and offered a range of sampling sites for the group to explore.

“We wanted to map the gradient of the problem away from the gold mining zone, “ says Lucotte, “so we chose an area about 400 km downstream of where the gold miners were active.” At the same time, Mergler wanted to study the local populations—some exposed to mercury, others not—so she chose villages equally far downstream.

“Both of our studies showed that people 400 km downstream were just as exposed as people 50 km downstream. That was quite amazing,” says Lucotte. “And, on the environmental side, we found that there was exactly the same amount of mercury in the river water whether 50 km or 400 km downstream. The same was true of sediments and in the terrestrial ecosystem.”

These results flew in the face of those who had based their careers on the hypothesis that gold mining was the sole source of mercury contamination. Lucotte puts this down to people founding their research on a linear correlation between gold and mercury, and performing their studies within too small a range. That being said, both he and Mergler stress that, within 50 km of gold mines, they believe that mercury released from shoddy amalgamation processes does play a role in the contamination. Overall, however, they believe that gold mining accounts for no more than 3% of the total mercury present.

Instead, Lucotte believes that the source of the contamination is natural, a product of volcanic materials released into the atmosphere over millennia and eventually deposited on the ground. With enough time, the mercury accumulation becomes significant. Lucotte explains that, unlike the soils of Québec, which last experienced glaciation about 10,000 years ago, the soils of the Amazon are ancient, perhaps more than 1 million years old.

“If the soils are very old, they have been accumulating mercury for a very long time,” Lucotte says. “And this is exactly what has happened in the Amazon.”

But, if this is true, one question remains. What caused the mercury to be released from the soil?

The Lorax
Taking sediment samples at several spots along the Tapajós River and testing for mercury in 5-mm increments, the researchers noted that the most recent sediments carried 1.5–3 times as much mercury as those deposited 40 years ago. A similar study of riverbank soils indicated even higher mercury levels. This history correlated well with the fact that the region had undergone massive colonization beginning in the 1960s, led in part by initiatives from Brazil’s National Institute of Colonization and Agrarian Reform. Under these initiatives, tens of thousands of families moved from the poorer regions of northeastern Brazil to the Amazon basin.

“During our [8-year] stay in the Amazon, we could actually see the propagation of the pioneer front of maybe 150 km,” relates Lucotte.

When the colonists move into a new region, they need a place to live and a way to support their families. Some go into gold mining, but this is a poor way to make a living. Instead, most colonists—and many miners—turn to farming, which requires the clearance of large tracts of forest, typically by “slash and burn” methods. The remote sensing center of the Brazilian Institute of the Environment and Naturally Renewable Resources (Brasilia) estimates that, from 1996 to 1999, more than 2.5 million hectares of the Amazon region were deforested. This is the equivalent of losing two-thirds of Switzerland. This rampant deforestation allows the soils, previously by the trees and undergrowth, to erode into the rivers and streams with each passing rain.

“When you have forest cover, this mercury is extremely stable in the soils,” explains Lucotte. “There is hardly any release to the aquatic ecosystem. The mercury is bound to clay, organic matter, humic acids, and so on.”

But, upon release into the aquatic ecosystem, everything about mercury changes. Although inorganic mercury (Hg2+ and Hgo) is highly toxic, the largest source of mercury poisoning in the Amazon basin comes from organic mercury—specifically, mercury that has been biomethylated.

As it enters the aquatic environment, especially the anoxic organically rich sediments, mercury is converted by sulfate-reducing bacteria into its more toxic organic species, typically methylmercury (MeHg). It is believed that the bacteria do this to limit the metal’s toxicity and increase the excretion rate. Organic mercury interacts very tightly with proteins and thus, when it enters the food chain, it is poorly eliminated. Initially taken up by algae, MeHg bioaccumulates as it moves along the food chain, passing into plants, then to herbivorous fish, to piscivorous fish, and finally to omnivorous humans.

Using hair samples taken from the riverine natives, this cycle can be illustrated by mercury levels that increase and decrease with the changing seasons. “Mercury levels in the hair followed the fish-eating habits of the population,” relates Mergler. “That is, in one village, during the high-water season—when waters are 6 m higher than in low-water season—there were mainly piscivorous fish with high levels of mercury. And [the natives] eat mainly herbivorous fish in the low-water season.”

“You could see the sinusoidal changes in mercury hair levels,” she continues. “The more they had mercury in their hair, the slower they were in motor tasks and the more visual dysfunction they had.”

But does this mean that everyone is in agreement?

Second Opinions
Interestingly, while the Canadian group’s first hint of an alternative source of mercury contamination was their inability to see a gradient of contamination away from the mining sites, this was not the experience of all such research groups. Studying children living along the Tapajós and Amazon rivers, Philippe Grandjean and colleagues noted a correlation between the levels of mercury contamination and the distance from gold mining operations.

“I am not an expert on sources of mercury,” admits Grandjean. “However, there is very little deforestation above the Mundurucu reserve where we found the highest mercury exposure levels. But there is a lot of gold mining.”

He also admits that their results are difficult to interpret with regard to mercury sources because exposure depends on diet, for which they did not control, so individual exposure levels may not correspond to the overall river contamination level. “Still, if the overall community averages show this gradient, then that is a rather strong suggestion,” he adds. “We made an effort to include almost all children aged 7–12 years in the villages, so our material has substantial validity.”

Mergler echoes Grandjean’s comment on the lack of dietary information. “In Grandjean’s study, where he examined the different villages at different times of the year, with no information on what they were eating, it is not possible to draw the conclusion that there is a gradient from gold-mining.”

The Fish-Eating Variable
Mergler described the cyclical fish-eating habits in two of the villages that her group studied, 100 and 250 km from the gold-mining operations, where the peaks and troughs of fish consumption change in opposite directions between high and low water seasons. “Depending on the time of year that a study would be carried out, we would have different mean levels for the populations,” she says. “In the high-water season, we would conclude that there is a gradient from the gold-mining, while in the low water season, we would have concluded the opposite.”

While it is apparent—and fully admitted by Lucotte and Mergler—that gold mining does play a role in mercury contamination, Anne-Hélène Fostier of the Universidade Estadual de Campinas (Brazil) offers an intermediate solution. “Gold mining also strongly damages the soil cover, which could represent a significant secondary source of Hg contamination of the aquatic system,” writes Fostier. She based her comments on studies of the State of Amapá in northeastern Brazil, where she noted some effect of regional mining on the total mercury load, but no more than the 3% suggested by Lucotte’s group.

Regardless of the source, mercury contamination is a fact. And the most immediate question becomes how to help those who are or will soon be contaminated. Lucotte and Mergler suggest a multidisciplinary approach that focuses on educating the local population.

In the short term, they are convincing the locals to use herbivorous fish as the preferred food source. In the longer term, the group is trying to assist the local people, with the help of the various levels of government, to find other methods of land use. But, this is no easy task in a region where credit is hard to come by, and it will require the development of new economic niches for agricultural and industrial products.

The problem is a big one that was long in developing. It will not be solved quickly. But, with a better awareness of their place in the ecosystem, the Amazonians are on their way.

Measuring Mercury
A schematic of Lucotte's cold vapor atomic fluorescence spectrometer.
The two predominant techniques for the detection and quantitation of total and organic mercury are cold vapor atomic absorption spectroscopy (CVAAS) and atomic fluorescence spectroscopy (AFS), with detection limits of 10–9 g/g (Hg/sample) and 10–12 g/g, respectively. Unfortunately, compared to AFS, CVAAS’s need for larger sample sizes and its lower sensitivity are relegating this technique to second-class status. In fact, the U.S. Environmental Protection Agency’s Measurement of Mercury in Water (EPA Method 1631, Revision B) requires AFS.

In AFS, Hg atoms are excited by the emission of light at a 254-nm wavelength. As the atoms return to the ground state, they emit energy as fluorescence (also at 254 nm), which the detector captures. Unlike CVAAS, which measures the total absorption of light and therefore cannot account for other atom types that may absorb the emitted light, AFS only measures the fluoresced light and therefore suffers less interference, resulting in higher sensitivity.

Because of the working conditions experienced by Marc Lucotte’s team, the Québec researchers developed their own AFS unit (see figure at right). To measure total mercury levels, milligram quantities of the soil or biological samples were acid-digested for several hours at 121 °C. The samples were then diluted, and aliquots were mixed under positive argon pressure with a solution of SnCl2 to generate Hg vapor. The argon carries the vapor to the fluorescence cell where the Hg is detected. To detect total Hg in water, the sample was first photochemically digested under UV light in a KMnO4 solution to liberate metal associated with organic complexes. These methods gave the researchers Hg detection limits of 5 pg/mg for the soil/biological samples, and 0.3 ng/L for the water samples.

Traditional methods for measuring methylmercury (MeHg) in soil and biological samples relied on an organic extraction and distillation process, but this method was time-consuming and required large quantities of sample. To address this, Lucotte’s group developed a saponification method where milligram samples were digested in a KOH/CH3OH solution for several hours at 68 °C. Next, the MeHg was converted to methylethylmercury, separated by gas chromatography, and quantified by AFS. Similarly, water samples were distilled in a solution of KBr, H2SO4, and ammonium pyrrolidine dithiocarbamate for several hours at 110 °C under nitrogen, and ethylated as before. These methods gave detection limits of 0.3 pg/mg for the soil/biological samples and 1 pg/L for the water samples.

Further Reading

  1. Mercury in the Biogeochemical Cycle; Lucotte, M., Schetagne, R., Therien, N., Langlois, C., Eds.; Springer-Verlag: New York, 1999; pp 41–52.
  2. Mineral and Metal Neurotoxicology; Yasui, M., Strong, M. J., Ota, K., Verity, M. A., Eds.; CRC Press: Boca Raton, FL, 1997; pp 177–188.
  3. Lodenius, M.; Malm, O. Mercury in the Amazon. Rev. Environ. Contam. Toxicol. 1998, 157, 25–52.
  4. Roulet, M.; Lucotte, M.; Canuel, R.; Farella, N.; Courcelles, M.; Guimarães, J-R. D.; Mergler, D.; Amorim, M. Increase in Mercury Contamination Recorded in Lacustrine Sediments Following Deforestation in the Central Amazon. Chem. Geol. 2000, 165, 243–266.
  5. Mercury Study Report to Congress; Report EPA-452/R-97-007; U.S. Environmental Protection Agency: Washington, DC, 1997.

Randall C. Willis is an assistant editor with Today’s Chemist at Work. Send your comments or questions regarding this article to tcaw@acs.org or the Editorial Office 1155 16th st N.W., Washington, DC 20036.

Return to Top || Table of Contents

 CASChemPortChemCenterPubs Page