plasma atomic emission spectrometry (DCPAES) and
ETAAAS; presentation in slurry form can also be
made with the last two techniques. Nondestructive
analysis can be performed using X-ray fluorescence
(XRF).
Analytical Techniques
0017 Traditional gravimetric and titrimetric methods have
much poorer levels of detection for cobalt than
modern instrumental techniques and are subject to
interferences. Two of the more commonly used pre-
cipitating reagents are 1-nitroso-2-naphthol (interfer-
ence from copper and iron) and anthranilic acid
(interference from iron, nickel and zinc). Typical
titrants are potassium cyanide (interference from
copper, mercury, and zinc), potassium hexacyanofer-
rate (interference from manganese), and ethylene-
diaminetetra acetic acid (EDTA) (interferences from
nickel, zinc, and copper).
0018 Prior to the rapid expansion of flame atomic
absorption spectrometry (FAAS) in the 1960s, the
analysis of cobalt at the microgram level was per-
formed routinely by colorimetric techniques using
spectrophotometry in the ultraviolet and visible
wavelength range (300–700 nm). However, this tech-
nique is also subject to chemical interference from a
wide range of other elements. These have to be elim-
inated either by the addition of masking agents or
buffers, or by solvent extraction of the cobalt com-
plex. The more widely used complexing agents are
ammonium thiocyanate (interferences from iron and
copper), nitroso-R-salt (interferences from iron, chro-
mium, nickel, vanadium, and copper), 1-nitroso-2-
naphthol (interference from copper and iron), and
sodium diethyldithiocarbamate (interferences from
iron, aluminum, chromium, titanium, manganese,
copper, and nickel). By applying modern chemo-
metric procedures to spectrophotometric measure-
ments, cobalt can be determined simultaneously
with other trace elements. Flow injection analysis
has automated many cobalt colorimetric methods.
Recently, sensitive spectrophotometric methods have
been developed that depend on the catalytic effect of
the cobalt iron on the oxidation reaction between
hydrogen peroxide and o-dihydroxybenzene deriva-
tives such as tiron (disodium 1,2-dihydroxybenzene-
3,6-disulfonate), catechol, quinazarin, and gallocya-
nine; detection limits at the picogram level are obtain-
able. A more recent novel utilization of catalysis
incorporates chemiluminescence sensors that meas-
ure the catalytic activity of cobalt on the reaction
between luminol and certain oxidizing agents such
as hydrogen peroxide or periodate salts.
0019 The following spectrographic techniques have been
used for the analysis of cobalt (approximate detection
limits have been quoted to an order of magnitude):
FAAS (detection limit, 0.01 mgg
1
), ETAAAS (detec-
tion limit, 0.0002 mgg
1
), inductively coupled plasma
atomic emission spectrometry ICPAES, detection
limit, 0.01 mgg
1
), inductively coupled plasma
mass spectrometry (ICPMS, detection limit, 0.0002
mgg
1
), DCarcES (detection limit, 0.3 mgg
1
),
spark source mass spectrometry (SSMS, detection
limit, 0.005 mgg
1
) and DCPAES (detection limit,
0.01 mgg
1
).
0020FAAS has been the most extensively used technique
for the routine determination of cobalt at the
microgram level in food, agricultural, and biological
materials. The technique is relatively free from
chemical and spectral interferences, and a large
number of preconcentration chemical procedures
have been developed to improve detection levels.
Ammonium pyrrolidinedithiocarbamate and sodium
diethyldithiocarbamate are the two principal che-
lating agents that have been used for multitrace elem-
ent complexation prior to solvent extraction,
although a more cobalt-specific chelate is 1-nitroso-
2-naphthol. A range of different organic solvents have
been used for extraction of the chelated cobalt com-
pounds, the two most common being methyl isobutyl
ketone and chloroform. Ion-exchange resins such as
chelex-100 or C-18 sorbant materials can be used as a
method of on-line preconcentration in FIA systems
for FAAS.
0021ETAAAS has proved to be sufficiently sensitive for
the determination of cobalt at the nanogram level in
clinical and biochemical samples. The technique is
more susceptible to sample matrix interferences than
FAAS, but these have been significantly reduced by
using FAAS-type preconcentration techniques as
separation procedures for the removal of major elem-
ents. Alternatively, matrix modifiers such as palla-
dium, magnesium, or certain inorganic salts such as
ammonium phosphate or thiocyanate have proved
effective. Graphite tubes are normally used for atom-
ization, but more recently, tungsten tubes have been
found to be equally effective. Automatic on-line
preconcentration systems have been developed that
utilize a microcolumn packed with Muromac A-1 che-
lating resin. Analysis of solid samples is now possible
through the incorporation of minature cups that
can hold powdered samples. The resonant absorption
wavelength for cobalt lies in the ultraviolet region at
240.7 nm. Measurements are therefore subject to
background interferences that occur to a greater
degree in electrothermal rather than flame atomiza-
tion. All modern AAS instruments have automatic
background correction; there are four different
types: Zeeman, deuterium continuum, Smith–
Hieftje, and xenon continuum in the simultaneous
1434 COBALT