Justificación del Problema

The initial results of Li-AAS suggested that « 10 ppm lithium solutions (known

from the PE composition) were actually « 0.4 ppm in lithium, with hardly any

colouration of the AAS flame occurring when the solution was introduced. This

suggested that the presence of the polymer was somehow suppressing the registration of the lithium. In order to examine this a series of aqueous solutions of PEs of

Each Hittorf compartment is « 60 mg, 130 jimol of which is lithium. A current of 50 jiA flowing for 10 hours corresponds to a charge of 1.8 C, or 19 jimol of

electrons. If all of the current is carried by the anions this is the change in the number

composition PEOxLiClO^ (x = 8,18, and 50) were prepared such that they were each «

10 ppm in lithium. Separate solutions of roughly the same composition were prepared

by dissolving PEO and LiClO^ together in distilled water, to determine if the PE

production procedure gave rise to the detection problem. Finally, two solutions of «

8:1 composition were prepared: one being PEO and LiClO^ acidified with 10 cm^ of 5 M H2SO4 with a total volume of 100 cm3, in an attempt to depolymerise the polymer; and the other being LiC1 0 4 and CH3 0(CH2CH2 0 )4CH3, in an attempt to model a

depolymerised electrolyte.

Analysis of these solutions (whose results are in Table 4-A) confirmed that the

presence of PEO suppressed the detection of lithium, with the higher PEO- concentration solutions producing a lower response. There was no significant difference between the electrolyte solutions and the mixed solutions, suggesting that the

procedure for producing the electrolyte did not account for the signal suppression. The

acidified sample, which had been stirred at room 'temperature overnight, gave a

proportionately greater response than the other 8 : 1 samples; but at an apparent lithium

concentration of 0.85 ppm rather than 10.26 ppm there was no significant resolution of

the problem. The solution prepared using the tetramer gave a strong response, being

out of the calibration range of the spectrometer (5 ppm), but the colour of the flame (bright crimson) suggested that little, if any, suppression of the lithium signal took

place.

The problem with the lithium detection was an initially surprising occurrence,

because the actual PEO concentration was only « 10 mM, and in an aqueous solution little salt-polymer interaction was anticipated. Such interferences in AAS may arise

from physical or chemical processes, and they are briefly described below.

4,5.1.1 Physical Interferences in Atomic Absorption Spectrometry.

Physical interferences may arise due to the viscosity of the sample solutions; if

the solution is too viscous the nébulisation efficiency of the spectrometer is affected, and this influences the number of free atoms in the flame. The 8:1 samples did not

seem unduly viscous, so this is not likely to be the source of the twenty-fold reduction in the detected lithium level. Solutions of 50:1 electrolyte were noticeably more viscous than those of the 8 : 1 electrolyte, which may account for the greater reduction in the

lithium detection for solutions of higher PEO content. The use of an electrothermal

atomiser, in which atoms from the sample are produced in a furnace rather than from

aspiration into a flame, circumvents this problem, and this method (which is more

sensitive) has been used to detect copper, iron, and chromium in PEO [89]. Metals which were added to the polymer were recovered at an efficiency of greater than 95% using this technique. The method of standard addition [90] may have circumvented this

problem, but it would have been time-consuming, and a sample signal of the order of

0.40 ppm would probably not have permitted the accurate differentiation of electrode

compartment and reference compartment results.

4.5.1.2 Chemical Inteiferences in Atomic Absorption Spectrometry.

Chemical interference may arise from processes occurring in the sample

solution or the flame, such as the formation of thermally stable oxides, carbides, or nitrides. Any ionisation of the metal atoms in the flame will reduce the number of atoms available for excitation, and this may be circumvented by adding to the solution

elements which are more ionisable (e.g. potassium). Interactions with cations or

anions from the matrix may occur, enhancing or suppressing the absorption of metals

in the flame, depending if the associated species produced are easy or hard to dissociate

in the flame: it is reported [91] that glucose present at a concentration of < 1 0 - 6 M

causes a marked decrease in the absorption of calcium present at a concentration of 10-4

M. In this case small amounts of glucose may complex or form compounds with the

calcium salt present, suppressing the calcium absorption. It is also known [92] that surfactants with 8 - ^ 0 ethoxy units form ion-association complexes with cations (in this

case zinc), and interfere with their detection. It is, therefore likely that chemical interference between PEO (or fragments from the decomposition of PEO in the flame) and the lithium was the source of the signal supression. An experiment which may

confirm this would be the introduction into the spectrometer of an aqueous solution of

lithium perchlorate in the presence of a lithium-specific crown ether. Suppression of

this signal would support the idea that chemical interactions between PEO fragments

and lithium atoms occurred.

4.5.2 Quantitative Lithium Analysis Using Atomic Absorption