| dc.description.abstract | Optimisation of the Total Reflection X-ray fluorescence spectrometer (TXRF) used in this work
was accomplished by instrumental adjustments of relevant parameters, sample preparation
and appropriate choice of excitation conditions. The TXRF module used was optimised for Mo
x-ray tube at 40 kV and the Bremsstrahlung cut-off energy at 20.0 keY. Sample preparation
method was chosen according to the nature and origin of the sample. Sample excitation
efficiency depended largely on the homogeneity of the sample material on the carrier surface
and the x-ray tube operating conditions. Sample preparation by preconcentration technique
improved the detection limits of certain elements by a factor of over 50, in comparison to direct
analysis method in which homogeneity and controlled spreading of the sample on the carrier
surface rarely exceed a factor of two. In this study, spreading of 10 ul sample was limited to 5-6
mm diameter by completely drying the quartz sample carrier in a low pressure (350 mbars)
oven at temperature of 70° C for 10 hours.
Trace element analysis of various water samples that are sold in Nairobi as mineral drinking
water and tap water was done to assess the environmental pollution from heavy metal
contamination. For most samples analysed, the levels of potassium (K) ranged from detection
limit (0.2 ug/ml) to 28.9 ug/ml, calcium (Ca) 2.2 to 120 ug/ml, titanium (Ti) detection limit
(11 ug/l) to 60 ug/l, manganese (Mn) detection limit (8 Ilg/l) to 670 Ilg/l, iron (Fe) 31 to 540
ug/l, zinc (Zn) detection limit (Sug/I) to '4730 Ilg/l, bromine (Br) detection limit (8 ug/l) to 248
ug/I, rubidium (Rb) detection limit (10 Ilg/l) to 40 ug/l, and strontium (Sr) detection limit (8
ug/l) to 1000 ug/l, The experimental concentration values for potassium (K) and calcium (Ca)
for most of the samples were in agreement to predetermined values (label). In general, most of
the local mineral water samples contained higher levels of trace elements compared to the
imported brands. _Concentration levels of iron in 50 % of the imported, and 80 % of the local
samples exceed the WHO maximum limits. Principal Component Analysis of the results for
water samples revealed three factors of pollution sources. The highest component loading
clustering include rubidium, strontium and calcium in the first eigenvalue; titanium, iron,
bromine and zinc, in the second; zinc, manganese and potassium in the third. The percentage of
total variance explained by the component was 31.4 %, 27.3 %, and 14.8% respectively.
For biological samples, three local food samples consumed by most Kenyans, and some
certified reference materials were analysed. The samples were: cereal based; raw maize (Zea
mays) and processed flour; vegetable based; kale (Brassica olaracea ) or locally referred to as
sukuma wiki, and beverage based; instant and cured tea leaves (Camelia sinensis). The cereal
based samples contain low levels of trace element compared to vegetable samples. The
distributions of trace elements in these samples, to a large extent, are influenced by the type of
- IX -
the food sample or environmental factors. Trace element intake in the Kenyan diets heavily
relies on the consumption of kale that has significant proportion of strontium (l07±59) ug/g and
rubidium (89±25) ug/g, tea leaves with manganese (l060±169) ug/g and rubidium
(116±26) ug/g, Consumption of maize meal contributes largely to Fe (27±9 Ilg/g) intake given
that it is the stable diet. The study also indicated that the concentration levels of the elements in
the food stuff is modified by processing methods. Principal Component Analyses on the maize
sample results indicated, 64 % of the observed total elemental concentration variance, was
explained by a single factor while for kale and tea samples, the most significant factor
explained only 45% and 42 % respectively. | en |
