The analysis of heavy metals in digested samples depends on parameters such as pH, temperature, time and digestion procedures. These parameters were Investigated in order to obtain optimum experimental conditions for the determination of mercury, cadmium and lead in fish, unfiltered water, sediment, soil and plant samples from Naivasha area, Kenya. The aim was to determine the level of pollution, regarding these heavy elements in the area. The techniques used were Flame and Cold Vapour Atomic Absorption Spectrophotometry < FAAS and CVAAS), and Energy- Dispersive X-ray Fluorescence analysis (ED-XRF).
The AAS absorbance and XRF intensity for mercury, cadmium and lead were found to be pH dependent. Maximum values were found in the lower pH region, <1.5, with AAS and 1.6-2.7 with ED-XRF for mercury and lead. The detection limit of mercury with CVAAS was found to be lower than with XRF. Inter-comparison of mercury and lead, in a cross-section of the digested samples, between AAS and XRF techniques showed excellent correlation coefficients of 0.9982 and 0.9999 respectively.
In this project, it was found that by using a 3:1:1 ( HNO3:HC104:H2S04) acid mixture (vol/vol), higher mercury sensitivity was obtained than with a 3:1 <HN03:HC104> acid mixture, using Cold Vapour AAS as the analytical technique. The former digestion method was tested on two International Atomic Energy Agency standard reference materials, MA-M-2/TM and MA-A-2CTM), which gave high mercury recoveries of 77.7% and 100% respectively.
Comparison between stannous chloride and sodium borohydride as reducing agents in CVAAS technique was also carried out. Stannous chloride exhibited higher sensitivity, stability and lower limit of detection than sodium borohydride. Prolonged oxidation time (at least overnight) of the digested samples gave high correlation coefficient (0.977) and regression coefficient (0.991), between the two reducing agents. The detection limits of mercury with SnCl2 and NaBH4 as reducing agents were found to be 0.243 and 2.303 ng/ml respectively. It was found that mercury concentration reading was negatively correlated with the digestion temperature. A digestion temperature of 50 ± 3°C was found to be preferable.
For the determination of cadmium and lead with flame AAS, a digestion mixture of 3:1 nitric and perchloric acids was used. Different models of AAS equipment, located in different labs, were compared and the detection limit and sensitivity values found to be close to those given in the literature.
The sample concentration levels for the three heavy elements investigated (mercury, cadmium and lead) were found to range, respectively, as follows: fish: 0.052 - 1.521 mg/kg, 0.008 - 0.562 mg/kg, 0.514 - 5.111 mg/kg (wet weight); water and condensed steam: 0.019 - 0.143 mg/1, 0.045 - 0.200 mg/1 , 0.970
- 5.525 mg/1; sediments: 0.209 - 0.269 mg/kg, 0.530 - 2.129 mg/kg, 13.710 - 28.084 mg/kg (dry weight); soils: 0.104 - 0.256 mg/kg, 0.479 - 2.668 mg/kg, 12.133 - 37.497 mg/kg (dry weight) and plants: 0.097
- 0.153 mg/kg, 0.305 - 2.961 mg/kg, 4.465 - 27.926 mg/kg (dry weight).
The concentration of the three heavy elements in fish and sediments compare well with the recommended and literature values. There was little indication of pollution with respect to these elements. However, the levels of the three elements in the water samples indicated probable pollution. On the other hand, soils and plants had elevated cadmium levels, whereas mercury and lead were within the expected range.