SciELO - Scientific Electronic Library Online

Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Journal of the Chilean Chemical Society

versión On-line ISSN 0717-9707

J. Chil. Chem. Soc. v.52 n.3 Concepción sep. 2007 


J. Chil. Chem. Soc, 52, N° 3 (2007) págs.: 1261-1265






1 Department of Inorganic and Analytical Chemistry, Faculty of Chemical and Pharmaceutical Sciences, University of Chile, P.O. Box 233, Santiago, Chile.
2 Department of Chemistry, Faculty of Sciences, University of Chile, P.O. Box 653, Santiago, Chile.


The concentration of trace elements and other analytes of environmental interest was determined in waters and sediments of the Coya and Cachapoal rivers (VI Region, Central Chile), which corresponds to a location highly polluted by copper mining activity.

Analysis showed the highest pollution at the sampling station closest to the copper smelter (Caletones), then pollutant concentrations decrease at sites far away from this source.

The statistical correlation between the different variables determined in waters and sediments was carried out through a principal components analysis (PCA). It was observed that the parameters fell into three groups: (a) analytes present in waters and sediments which come from the mining wastes (b) analytes of lithoge-nic character present in waters plus alkaline and alkaline earth metals plus organic matter present in sediments and waters, and (c) lithogenic elements present in sediments. On the other hand, by using cluster analysis (CA), it was possible to group the different sampling stations according to their extent of pollution.


The contamination of rivers by trace elements coming from mining activity is today considered a global problem of international concern because the ecosystems located along the river can be damaged.1

The impact of this kind of contamination depends on a number of factors including nature of the mining extraction process, nature of the ores, efforts in the control of contamination, and also on environmental factors. It is well-known that physical-chemical factors such as pH, redox potential and nature and concentration of organic matter, control the mobility and availability of trace elements.2-4

Sediment analysis has been widely used as an environmental indicator to assess the extent of contamination in an aquatic system. Its ability to identify pollution sources is widely recognized.5-10 Therefore, in order to understand the environmental behavior of elements including pollutants and taking into account that sediments are in continuous interaction with the liquid phase, it is more convenient to combine both water and sediment analysis. 8-10

Chile has experienced a dramatic increase in its mining activities in the past few years, bringing about investments and technological changes. Some of these changes, however, have had damaging effects on the environment. Mining processes produce a considerable amount of wastes part of which is dispersed in the air, water and soils. The surroundings of the Caletones smelter, located in the VI region of the country, in central Chile, is also impacted by this kind of mining activity. It has been demonstrated that heavy metals are emitted from this source.910 In that area, the Cachapoal River has received for years the effluent from mining activities, affecting agricultural and probably aquatic life close to the outlet of the Cachapoal River into the Pacific Ocean. Taking these antecedents into account, the biggest mining company in Chile is taking steps to ensure that copper becomes an environmentally sustainable element over time and will contribute to reduce any possible negative effects of its production and use throughout its life cycle.911

It has been internationally recognized that in Chile there is lack of scientific information related to the flow and mobility of contaminants in aquatic ecosystems12. This is particularly true of aquatic systems of the VI Region in Central Chile, which are affected by mining activities. In previous studies,910 it was found that arsenic was mainly present as As(V) in both water and sediment, and organic arsenic species were not detected. The presence of high concentrations of arsenic and copper found in all the samples under analysis suggests a real impact of the mining activity over the region. The presence of arsenic in water and in sediments probably resulted from an acid drainage, because pollution clearly diminished with the distance from the source. In the case of soils, they can be impacted by irrigation with river water. However, the high concentration values determined at sites far from the source suggest that soil pollution is more related to air particles deposited in soils. At the most impacted sites, most arsenic is immobilized as a form of As (V).9

In this study, the preliminary physical chemical characterization of water and sediment samples collected at different points of the Coya and Cachapoal rivers has been carried out. The main objectives of this study were to determine the element concentration in waters and sediments and to establish the extent of contamination in this aquatic system at different distances from the main pollution source (Caletones smelter). Further, statistical data treatment was carried out by principal component analysis (PCA) and cluster analysis (CA) in order to reduce the amount of physical chemical variables for correlation and grouping of the sampling points according their pollution extent.



All chemical used were of analytical grade. De-ionized water (NANOpure ultrapure water system, Barnstead) was used throughout. Working solutions of different concentrations of arsenic species were prepared from 1000 mg L-1 standard solutions. As(III) standard solution was prepared by dissolving extra pure reagent As203 (Aldrich, Milwaukee, WI, USA) in 0.1 mol L-1 NaOH solution (Merck). As(V) standard solution was from WAKO pure Chemical industries, Japan. Monomethylarsonic acid (MMAA) and Dimethylarsinic acid (DMAA) standard solutions were prepared by dissolving the pure reagents (Tri Chemical Laboratory Inc., Japan). A 5 % sodium tetrahydroborate (Merck) solution was prepared in 0.25 % NaOH just before use.

Multielement standards QCS-26 and ICP-200.7-5 (High Purity Standard, USA) were used for ICP measurements of total concentration elements (As, Cu, Zn, Pb, P, Fe, Al, Si, Mg, Ca, Na, K). An AccuStandard, Inc. multistandard was used for sulphate and nitrate determination.

A Na2C03/NaHC03 (3.5 mmol/L /1.0 mmol/L) buffer system was used as eluting agent in ion chromatography.

Nitric acid 65%, Suprapur® (Merck) y HF 40% (Merck) was used for sediment microwave digestion. Solutions of 1 mol/1 K2Cr207 (Merck), 5 mol/l FeS04x7H20 (Mallinckrodt Analytical Reagent) and ferroin (Sigma-Aldrich Co.) were used for organic matter determination in sediments.

Water samples for analysis of trace elements were preserved al pH<2 by adding drops of HN03 (Merck) and H2S04 (Merck) for determination of Total organic carbon (TOC).


An Inductively Coupled Plasma-Optical Emission Spectrometer (ICP-OES) Perkin Elmer Optima 3300 XL was used to determine the total element concentrations in water and sediment samples. The manifold used for arsenic speciation has been described elsewhere.910 It basically consists of a hydride generation (HG) system connected to a Cold-trap unit (U-tube half-packed with Chromosorb W AW-DWCS 60/80 mesh OV-3 (15 %) (Supelco, Bellfonte, PA)) and an ICP-OES.

A Dionex 120 ion chromatograph with an AS 14 analytical column and an AG 14 pre-column was used for determination of nitrate and sulphate in water samples. In situ measurements of pH, Eh (redox), and EC (electrical conductivity) were carried out at each sampling point by using a portable water quality multiprobe (Hydrolab) that had been calibrated on the sampling day.


The Coya and Cachapoal rivers were considered in this study. Figure 1 shows the sampling area. It comprises a wide area of Central Chile (VI Region) with intense agricultural and mining activity. The sampling program considered sample collection mainly from locations near the Caletones smelter on the NE to sectors next to Rapel Lake, located west of Rancagua City. Three sampling periods were considered: April 2003, April 2004 (autumn) and August 2004 (winter). Sampling station N° 9 (Tinguiririca river) was selected for comparison and assessment of background level in this area because it is not affected by heavy metal pollution.

Surface water samples were taken from both rivers. All water samples were collected in clean polyethylene bottles, filtered through 0.45 um, kept at 4°C and frozen on arrival at the laboratory until analysis. One additional water sample, filtered and preserved with nitric acid (pH <2), was taken at each sampling point for total element determination. Samples of sediment were also taken at each point.

Samples of loose surface sediment were collected in polyethylene bottles, using a plastic trowel. All sediment samples were air dried, ground and sieved (0.6 mm).


Total element determination was carried out by ICP-OES, following the US-EPA 6010B method.13 The sediment samples (0.3 g) were previously digested in a microwave oven Millestone mis 1200 mega following the US-EPA 3051 method.14

Arsenic speciation was carried out by hydride generation coupled to cold trap and ICP-OES.910 In the case of sediments, 5-g samples were extracted with 25 mL of 1 mol L-1 HC1 by shaking mechanically at 30 °C and 200 rpm for lh. Speciation was carried out in the extract.9

Determination of organic matter in sediments was carried out by Walkley-Black method.15 Nitrite and sulphate were determined by ion chromatography 16 and Total organic carbon was measured by Combustion-Infrared Method.16

Statistical analysis

In this study, the statistical analysis of the results was based on Principal Component Analysis (PCA)17-18 and Cluster Analysis (CA)17-19.

PCA and CA were carried out by using the statistical package Statgraphic Plus v. 5.1. The experimental data were standardized before the chemometrical analysis by subtracting their mean values and dividing them by their standard deviations.

The most relevant physico-chemical variables in both filtered waters and superficial sediments were considered. In addition, data standardization of each variable was done prior to statistical analysis because the concentration orders among the elements and the parameters differed greatly and statistical results would be highly biased by elements with high concentration.8

CA was performed from the original data in order to reduce the variant biases. The Furthest Neighbor method was applied for grouping of the different variables, and dissimilarity was defined by City-Block distance.



Determination of physical and chemical parameters in water was carried out at the nine sampling stations. As can be seen in Table 1, the lowest pH value and the highest values of both electric conductivity (EC) and redox potential (Eh) were measured at sampling station N° 2 (Coya River between Sewell and Caletones). This sampling point also showed the highest concentrations of trace elements (such as Cu, Fe, and As), anions and TOC, which indicates that the main discards of mining waste into the Coya river occurred between sampling stations 1 and 2.

On the other hand, it can be observed that downstream of sampling point 2 of the Coya River, concentrations of all parameters decreased due to dilution and sorption on the surface of sediments. After the Coya river joins the Cachapoal River, the concentrations of these parameters decreased to a higher extent, reaching the background level, defined by point 9, at sampling station 8.

Due to the available information, special care must be taken if some dredge or sediment removal occurs, because the sulfides and some metals retained in this river fraction can be released, contaminating these aquatic systems even more.20

On the contrary, the concentrations of P, Al, and Si in water remain almost constant at all sampling points (Coya - Cachapoal), which is consistent with the lithogenic nature of these elements. Contrary to Coya and Cachapoal rivers, Tinguiririca river (point 9) showed lower concentration of Fe, Al, Si, S042- and lower Eh and EC, which can be attributed to the different mineral composition of its sediments.

Regarding the seasonal effect, a slight decrease was observed in trace element concentrations in water in August 2003 (spring) compared with April, which is consistent with the increment in the flow rate of the rivers under study in this period. However, only in this period methylated arsenic species were determined, which is in agreement with the slight increment in TOC concentration making microbiological activity possible.21


The determination of chemical parameters in sediment was carried out at the nine sampling stations (Table 2). Similar to the observations in water analysis, higher concentrations of trace elements, including arsenic species, and organic matter were determined at sampling station 2. Downstream of this point the concentrations decreased to reach almost the background level at sampling station 8. In contrast, sampling stations 1-8 showed similar concentrations of P, Al, Si, Ca, Mg, Na and K among them. However a comparison with sampling point 9 (Tinguiririca River) indicates that concentrations of Fe, Al and Si were lower in the latter case, possibly due to the different composition of the parenteral material of both basins.

Seasonal variations were lower in sediments than in water, which can be explained because dilution affects sediments to a lower extent.

Multivariate analysis

The procedure of multivariate analysis was based on the observations obtained in relation to several variables for a number of samples (also called eigenvectors). Multivariate methods to analyze environmental data constitute an increasingly important part of statistics, because they allow an easier interpretation of the results.22-24

Principal component analysis (PCA)

PCA was applied in order to obtain a statistical correlation among the different components of the environmental matrices analyzed (river waters and sediments). The application of PCA allowed reducing the number of the variables in a few components with minimum loss of information, which was obtained by the linear combination of all the variables in consideration. The number of the main components was obtained on the basis of the sum of the eigenvectors (eigenvalue) being greater than 80% of the total variability.

The application of PCA to the set of variables determined in filtered waters as well as in surface sediments indicated that these 39 selected variables were represented by 3 new variables or principal components. These 3 new components account for near 87% of the total variance in the data set. Therefore, three vectors were used for statistical analysis. The eigenvectors and eigenvalues obtained in this analysis are shown in Table 3. In order to achieve the association of the variables to each principal component, the corresponding weights (eigenvectors) in the scores matrix were analyzed, and the highest weight observed allowed to associate the corresponding variable to that component. The eigenvectors classified the 39 physico-chemical variables (elements, compounds and field parameters) into the following three components:

• Component 1: Total As and its species in river waters and sediments, heavy metals generated by mining activities in river waters and sediments, field parameters (pH, Eh, EC), and geological components present in waters.
• Component 2: Lithogenic elements in river waters, alkaline and alkaline earth metals present in sediments, and organic compounds in water and sediments.
• Component 3: Lithogenic elements present in sediments.

 In other words, PCA indicated association between metallic and nonmetallic elements added to the rivers by the mining activity. In addition, the high concentration of polluting elements confirmed the established previously thought the measurement of field parameters (pH, Redox potential and conductivity) regarding to the contamination of the aquatic system. On the other hand, component 2 revealed that lythogenic elements present in high concentrations in river waters demonstrated similar behavior to that of lythogenic elements present in smaller concentration in sediments and with the organic compounds present in waters as well as in sediments. Finally, component 3 revealed a correlation among the lythogenic elements present in high concentrations in sediments.

Cluster Analysis (CA)

The result of CA (cluster analysis) allowed to obtain data about the features of each sample of water and sediment, the different sampling stations included in this study being grouped in different conglomerates, based on their contamination extent, which is shown in a dendogram graph (Figure 2). The dendogram clearly revealed the formation of 2 groups:

A: Not impacted (stations N° 1, 4 and 9) and not directly impacted by the mining spills (stations N° 6, 7 and 8) sampling stations. In this latter group it was observed that sampling station N° 8 (April of 2004) just as sampling station N° 9 (August of the 2003), presented higher distances than the rest of the stations corresponding to group A.

In the case of sampling station N° 8 (April, 2004), the greater City-Block distance may be attributed to the decrease in the flow rate of the Cachapoal river in April, 2004, in contrast with the flow rate observed during the other sampling periods, which favored a slight increase in lythogenic element concentration. On the other hand, the greater distance of sampling station N° 9 (August of the 2003), corresponding to the Tinguiririca river (baseline), it attributed to the differences in mineral composition between the Tinguiririca and Cachapoal river basins.

B: Sampling stations directly impacted by copper mining activities of (stations N° 2, 3 and 5). In this case it was observed that sampling station N° 2 (August of the 2003) presented a higher City-Block distance than the other stations in group B, which was attributed to the greater concentrations of polluting elements added by mining spills as determined at this sampling station.

The statistical analyses were carried out considering the water and sediment samples independently, in order to make a comparison of the results with the system considered as a whole. Both PCA and CA provide similar results to the obtained when all experimental data were considered simultaneously. From an environmental point of view the later situation is more representative of an ecosystem because interactions between both compartments occur.

Information from the scores matrix PCA regarding from the components weights (Table 3) shows a similar grouping of the sampling sites obtained from the CA (Figure 2). As can be seen in Table 4, the sampling sites grouping as B in CA (Figure 2) showed the main weight in the component 3. Contrarily, the sampling sites grouping as A in CA (Figure 2) have the main weights in the component 1 and 2 in PCA.


The VI Region of Chile has been impacted by liquid and air emissions from the copper mining process. The distribution of pollutants in waters and sediments of the Coya and Cachapoal rivers is controlled mainly by the distance from the pollutant source (copper smelter) and by the flow rate of the rivers. PCA indicated association among all heavy metals determined in the environmental matrices analyzed, revealing that their high concentrations are due to the discharge of liquid wastes to the Coya River. Further, CA correlated all sampling stations impacted by contamination resulting from the mining processes.


The authors acknowledge the financial support provided by the Japanese International Cooperation Agency (JICA) through JICA-CENMA project FONDECYT (project 1030005), CONICYT (AT-403001), and Universidad de Chile PG-68-02.



1.     J. Jr. Cairns and P. V. McCormick, Environ. Prof. 14, 186 (1992).        [ Links ]

2.     D. Zabowski, C. L.Henry, Z. Zheng, and X. Zhang, Water, Air, Soil Pollut. 131, 261 (2001).        [ Links ]

3.     K. Harmsen, "Behavior of Heavy Metals in Soils", Centre for Agricultural Publ. And Doc, Agrie. Res. Rep. 866. Wageningen, 1997, The Netherlands.        [ Links ]

4.     P. D. Abel, 'Water Pollution Biology", Taylor and Francis, London, 2004, p. 189.        [ Links ]

5.     M. Wardas, L. Budek, and H. Rybicka, Appl. Geochem. 11, 197 (1996).        [ Links ]

6.    Z. Borovec, Sci. Total Environ. 177, 237 (1996).        [ Links ]

7.     M. Singh, G. Muller, and I. B. Singh, J. Geochem. Explor. 80, 1 (2003).        [ Links ]

8.     K. Anazawa, Y. Kaida, Y. Shinomura, T. Tomiyasu, and H. Sakamoto, Anal. Sci. 20, 79 (2004).        [ Links ]

9.     P. Richter, R. Seguel, I. Ahumada, R. Verdugo, J. Narváez, and Y. Shibata, J. Chil. Chem. Soc. 49, 333 (2004).        [ Links ]

10.   J. Narváez, P. Richter and M. Toral. Analytical and Bioanalytical Chemistry, 381 (7), 1483 (2005).        [ Links ]

11.        [ Links ]

12.   P. L. Smedley and D. G. Kinniburgh, AppI. Geochem. 17, 517 (2002).        [ Links ]

13.   US-EPA 6010B method, U.S. Environmental Protection Agency, Office of Solid Waste. "Test Methods for Evaluating Solid Waste Physical/Chemical Methods ", SW-846 on CD-ROM, National Technical Information Service (NTIS), U.S. Department of Commerce, Springfield, VA, USA, 1996.        [ Links ]

14.   US-EPA 3051 method, U.S. Environmental Protection Agency, Office of Solid Waste. "Test Methods for Evaluating Solid Waste Physical/Chemical Methods", SW-846 on CD-ROM, National Technical Information Service (NTIS), U.S. Department of Commerce, Springfield, VA, USA, 1996.        [ Links ]

15.   A. Walkley and I. A. Black, Soil Sci. 37, 29 (1934).        [ Links ]

16.   Standard Method for the Examination of Water and Wastewater, 19° Ed. A. Eaton, L. Clesci and A. Greenberg Eds. United Book Press, Inc., Baltimore, Maryland, 1995.        [ Links ]

17.   Principal Component Analysis, 2ntl Ed. I. T. Jolliffe. Springer Series in Statistics. 2002. P. 487.        [ Links ]

18.   Principal Components and Factor Analysis. Electronic Text Book StatSoft. ©Copyright StatSoft, Inc., 1984-2003.        [ Links ]

19.   Cluster Analysis. Electronic Text Book StatSoft. ©Copyright StatSoft, Inc., 1984-2004.        [ Links ]

20.   W. Huang, J. Martin, P. Seyler, J. Zhang, and X. Zhong, Mar. Chem. 25, 75 (1998).        [ Links ]

21.   R. Bentley and T.G. Chasteen, Mol. Biol. Rev. 66, 250 (2002).        [ Links ]

22.   N. Mazlum, A. Ózer, and S. Mazlum, Tr. J. Engin. Environ. Sci. 23, 19 (1993).        [ Links ]

23.   J. N. Miller and J. C. Miller, "Estadística y Quimiometria para Química Analítica", Pearson Educación, S.A., Prentice Hall, Madrid, 2000, p-221.        [ Links ]

24.   F. Peña, S. Cárdenas, M. Gallego, and M. Valcárcel, Anal. Chim. Acta, 526,11 (2004).        [ Links ]


(Received 3rd March 2007 - Accepted 15th June 2007)

Corresponding author: e-mail:


Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons