Page 71 - Geochemistry of Oil Field Waters

FLAME SPECTROPHOTOMETRIC METHODS 59

Procedure. To determine the amount of potassium in the sample, transfer an
aliquot of sample to a 50-ml volumetric flask, add 20 ml of n-propanol, and
dilute to volume with distilled water. The specific gravity can be used to help
decide the aliquot size. For a brine with a specific gravity of 1.1, an aliquot
of 5 ml or less probably will be sufficient. Aspirate the sample into the flame
and record the emission intensity of the background at 750 mp and potas-
sium line at 766.5 mp. With this reading, use the preliminary calibration
curves and calculate approximately how much potassium is in the sample.
Determine an aliquot size that will contain about 0.05 mg of potassium.
Transfer equal aliquots to three 50-ml volumetric flasks. Add no potassium
standard to the first flask, 0.05 mg to the second flask, and 0.1 mg to the
third flask. Add 20 ml of n-propanol to each flask and dilute to volume with
distilled water. Aspirate and record the emission intensity of each sample at
766.5 mp and the background at 750 mp.
Optimum accuracy is attained by this method when the two standard
additions are respectively equal to and twice the amount of potassium in the
sample. Care should be taken that too much potassium is not present in the
final samples, because self-absorption will cause errors.

Calculation. The graph or formula illustrated in the lithium method can be
used. The value obtained in milligrams can be converted to milligrams per
liter by the following formula:

mg K x 1,000
ml sample = mg/l K+
The precision and accuracy of the method are approximately 2% and 4%
of the amount present. Several elements can interfere in the flame analysis of
potassium. Elements which ionize easily will lower the degree of ionization
of potassium, and elements which are difficult to ionize or have high ioniza-
tion energies will give the opposite effect. By using the Saha equation
(Herrmann and Alkemade, 1963), it is possible to estimate such interfer-
ences. Generally, the use of a standard addition compensates for inter-
ferences.
Rubidium and cesium

The flame spectrophotometer provides one of the most sensitive methods
available for determining rubidium and cesium. Cesium has a pair of emission
lines at 852.1 and 894.4 mp. Both lines are of about equal intensity, but
water produces a molecular band system at 900 mp which can interfere at
894.4 mp. Rubidium also has two strong lines in the red region at 780.0 and
794.8 mp.
It is necessary to use a photomultiplier with an S-1 response to detect
cesium and rubidium at the levels found in many waters. Examples of such
tubes are ITT type 6836/FW118, RCA types 1P22 and 7102, and DuMont