RESEARCH· ·1970·

Procedure
Sa1nple sol~ttion.-Weigh
1.000 g of sample and
transfer to a 30-ml zirconium crucible. Add 5 g of
sodium peroxide and mix;. then cover with an additionn.l
2 g of sodium peroxide. Place .a· zirconium lid
on the crucible and heat gently over an open flame
until the contents melt. Increase the heat and maintain
the crucible at bright-red heat for 2 minutes.
Cool. Cautiously ~dd 20 ml of water to the crucible,
and heat to dissolve the melt. Transfer the solution to
a 250-ml beaker, and wash the crucible with 2.5-percent
v/v hydrochloric acid. Combine wash and solution,
add 20 ml of 50-percent v/v hydrochloric acid,
and evaporate to dryness on the steam bath to dehydrate
the silica. Add 5 ml of 50-percent v /v hydrochloric
acid and 20 ml of water, and heat to dissolve
soluble salts. Filter the solution through a retentive
filter paper, and collect the filtrate in a 50-ml volumetric
flask. vVash the paper and residue three times
with 2- to 3-ml portions of 1-percent v /v hydrochloric
acid solution. Reserve the filtrate. Place the filter
paper in a 20-ml platinum crucible, and ash the paper
inn muffle furnace at 500°C. Cool. Add 5 ml of hydrofluoric
acid and 1 ml of perchloric acid, and evaporate
to dryness. Add 10 ml of 2.5-percent v/v hydrochloric
acid to dissolve the salts. Add this solution to the reserved
filtrate solution, and dilute to volume with
water.
Ext1·aotion.-Transfer an aliquot of the diluted
solution, containing 5-20 mg Co, to a 125-ml separatory
funnel. Adjust the volume of the solution· to
about 50 ml with· 2.5-percent v /v hydrochloric acid.
Similarly prepa.re standard solutions as directed in
the section, "Standardization." If the cobalt content
of the sample is expected to be less than 10 ppm,
transfer the entire sample solution to the separatory
funnel. Add 10 ml of sodium citrate solution, and mix.
Adjust the· pfi to between 3 and 4 with 2N hydrochloric
acid or 2N sodium hydroxide by using pH
paper. Add 1 ml of 30-percent hydrogen peroxide, and
allow to stand 2 minutes. Add 2 ml of the 2-nitroso-
1-naphthol solution, mix, and allow to stand 30 minutes.
Add 10 ml of MIBI{, stopper, and shake for 1
minute. Drain off and discard the aqueous layer. Add
5 ml of 2N hydrochloric acid solution and shake gently
for 20 seconds. Drain off and discard the hydrochloric
acid solution. Add 5 ml of 2N sodium hydroxide solution,
shake gently for 20 seconds, and discard the
aqueous phase. Repeat this operation with an additional
5 ml of sodium hydroxide solution. Wash the
organic layer twice by shaking with 3-I?J.l portions of
water, and discard the aqueous phase each time. The
MOUNTJOY
B175
organic phase should be a reddish brown at this point,
which indicates the removal of excess sodium hydroxide
and chelating reagent. Drain the organic phase
into a 15-ml centrifuge tube, stopper, and centrifuge
at about 1,500 rpm for 5 minutes. Measure the abso~ption
of the MIBI{ phase along with the reference
standards on an atomic absorption spectrophotometer.
Use a water-saturated solution of MIBK for a reference
blank.
Standardization.-Prepare reference standards containing
2, 5, 10, and 20 p.g Co in the organic phase for
a· working curve by adding appropriate aliquots of
the 10-p.g/ml Co standard solution to a series of separatory
funnels and process through the procedure
starting at the section, "Extraction." Prepare the
standard solutions concurrently with the sample solutions.
RESULTS AN'D DISCUSSION
Angino and Billings (1967) reported spectral interference
on the 2407 A resonance line of cobalt by Ca,
Mg, Fe, Na, and 1{ when they used an air-propane
flame. The solutions containing lmown amounts of
cobalt and either Mg, N a, K, Fe or Ca in 2.5-percent
hydrochloric acid were prepared, and the .interferences
of these elements in an air-acetylene flame were determined.
The ·analyses showed that (1) Mg, Na, and K
at concentrations as high as 6,000 ppm can be tolerated,
and ( 2) Ca and Fe interfered at both levels tested
(table 1). The 6,000-ppm Fe solutions gave readings
equal to about 0.5 ppm Co. The interference due to
iron seems to be independent of the amount of cobalt
up to at least 5 ppm, but the calcium interference is
greatest when no cobalt is present and is very small
i~ the presence of 5 ppm Co. The reason for t4is is
not known. The proposed extraction procedure removes
these interferences and, i~ addition, concentrates
the cobalt and increases the sensitivity of the
atomic absorption determination. The sensitivity of
cobalt in MIBI{ was found to be about 0.05 ppm for
1-percent absorption, an enhancement of about 10 times
that found for aqueous solutions.
A tracer technique was used to determine the efficiency
of the extraction of the cobalt complex into
MIBIC Co 60 was added to a solution containing 250
mg of U.S. Geological Survey sample W -1, then
processed through the extraction procedure. The test
indicated that approximately 98 percent of the cobalt
was extracted into the MIBI{ with a single 10-ml extraction.
Therefore, MIBI{ was chosen in preference
to chloroform or other organic solvents because of its
excellent flame characteristics for atomic absorption
work and because the stripping operation of the ex-