Page 28 - An Introduction to Analytical Atomic Spectrometry - L. Ebdon

Page 13

Glassware may give rise to further errors. For trace level determinations contamination and losses can
occur through (a) surface desorption or leaching and (b) adsorption on surfaces. To avoid
contamination, all laboratory glassware should be washed with a detergent and thoroughly rinsed, acid
washed by soaking overnight in 10% v/v nitric acid (sp. gr. 1.41), then rinsed with deionized, distilled
water (DDW) and allowed to equilibrate in DDW overnight. Disposable plastic-ware such as
centrifuge tubes and pipette tips should be similarly treated unless they can be shown to be metal-free.
Contamination is a particular problem for analysis by ICP-MS owing to its high sensitivity. Some
elements (e.g. Al, Pb, Na, Mg, Ca) are ubiquitous in the environment so stringent precautions may be
necessary to avoid contamination, such as the use of clean rooms and laminar flow hoods and the
donning of special clothing before entering the laboratory. If ultra-low determinations are to be made
then these precautions may be necessary at all times.

Blanks should be run for all analyses as a matter of course. Even if high-purity reagents are used, the
level of the analyte in the blank may constitute the limiting factor in the analysis, and it may be
necessary to purify reagents used for dissolution.

Adsorption is also a problem at trace levels. Few solutions below a concentration of 10 µg cm can be
-3
considered to be stable for any length of time. Various preservatives to guard against adsorption of
metals on to glassware have been reported in the literature. Common precautionary steps are to keep the
acid concentration high and to use plastic laboratory ware.

Q. Under what conditions is sampling most problematic?
Q. How can we minimize the possibilities of (i) contamination and (ii) losses by adsorption?

1.5.2 Operator Errors

Experience tells us that no account of possible errors can ever be exhaustive, and some techniques will
be more prone to certain types of errors than others.

A frequent source of error is poor standards. Besides the obvious error of standards being wrongly
made, it should not be forgotten that trace metal standards are unstable. Concentrations of 10 µg cm
-3
and less usually need to be prepared daily. Even standards purchased from commercial suppliers will
age and this is especially true when chemical changes can be expected in the analyte (e.g. silicon).