Cerate-Arsenite
Measurement of Iodine
in the Subnanogram Range
Hans Hoch and Charles G. Lewallen
A method for the measurement of subnanogram quantities of iodine is describeo.
Procedural detail and precision of results are presented for application of the method
to protein-bound or total iodine in duplicate on 12.3 ii.l serum. Following wet ashing
with chloric reagent, the iodine-catalyzed cerate-arsenite reaction was carried out
at temperatures between 23 and 27#{176},
and the reaction rate coefficients were corrected to refer to 25.0#{176}
by an empirical equation. The recovery of iodine standards,
added to the precipitated protein before ashing, ranged between 96 and 106%. One
contributing cause of this variation is the formation of an inhibitor of unknown nature
during heating. Samples were brought to a final volume of 180 .l prior to fading.
The detection limit was of the order of 0.04 ng iodine. For 62 serum samples
having a mean iodine content of 0.558 ng (range 0.15-1.31 ng), the standard
deviation calculated from duplicates was 0.019 ng/180 d. For 31 samples having
a mean iodine content of 2.24 ng (range 1.61-2.96 ng), the standard deviation was
0.039 ng/180 l. Expressed as micrograms of iodine per 100 ml of serum, the mean ±
standard deviation for the group of 62 samples was 4.54 ± 0.15.
IN
and research
situations
there exists a distinct
need for an iodine method
permitting
measurements
in the subnanogram range. This need is encountered,
for example,
in paper chromatographic
fractionation
of the iodine compounds
in the necessarily
small
amounts
of biologic
nlaterials
which can be subjected
to paper
chromatography.
The technic
reported
here was developed
to meet this
lleed. It will be shown that by scaling down to subnanogram
quantities,
A VALU
l:Tv of clinical
precision
can be maintained,
if not improved.
By way of demonstrating
the applicability
of the method,
we present
the procedural
details
and
precision
of results
for measurements
of protein-bound
iodine
(PBJ)
or total iodine oil microiiter
quantities
of serum.
Of tile acid (1) and alkaline
(2) methods
of ashing,
the former
has
the advantage
that ashing tubes which have proven free of interfering
time Physical
Cimemistry
Laboratory,
Veteraims
Admmiimiistration
Center,
Martitmsburg,
25401 ; amid the National Tnstitut-e
of Arthritis and
Metabolic
Diseases,
National
institutes
of Health,
Bethesda,
Md.
Received
for psiblieatiomm Jumie 10, 1968; accepted
for publication
July 25, 1968.
Frommi
W.
Va.
204
Vol. 15, No. 3, 1969
MEASUREMENT
OF IODINE
205
materials
giveii off by the glass can be reused
practically
indefinitely.
This cannot be done with alkaline
ashing
because
of loss of iodine by
sorption
on etched
tube surfaces
(3).
The high spectrophotometric
absorptivity
of cerate ion at 317 urn (4) was utilized,
and advantage
was -taken of favorable
properties
of polyethylene
nilcropipets,
such as
reproducibility
of delivery
and ease of handling.’
Tile conditions
for
increased
sensitivity
through
optimal
HC1O4 and FT0504 concentrations
ill the cerate-arsenite
system
(:) have been retained,
but -the chloric
acid wet-ashing
reagent
has been changed
to contain
more chromic,
and less perchloric,
acid. As is often the case when a method is adapted
to a smaller
scale, an appreciable
economy
in time and reagent
requirements
has been achieved.
Materials and Methods
Reagents
HC1O1, 17%
(w/v)
rfIlis
was pi’epaid
from 73-74%
(w/w)
HC1O4, density 1.71 (‘‘70%’’ ItCH)4, (L Frederick
Smith Chemical
Co.,
Columbus,
Ohio), by diluting
10 ml to 73 ml with HO. rIlliS solution
previously
was designated
erroneously
as ‘‘10%’’ (5-7).
2. Chioric flCi(l reagent
KClO., 29 g (Baker Analyzed
Reagent),
four times recrystallyzed
froni TTOO, was dissolved
in 65 ml boiling
H.,O; 23 ml ‘‘70%’’ (w/w) HCIO4 was added slowly. After storage
at 4#{176}
for at least 1 day, the supernate
was pipetted
off and stored
at -14#{176}.
3. Chromic
acid sointion
(6)
Cr0:1, 200-300
mg (Merck),
was
twice recrystallyzed
by dissolving
in a small amount
of 1120, heating
with 2-3 ml “70%”
HC1O4, and removing
the supernate
after cooling.
The residue
was dissolved
in TIO and then diluted
such that 0.050 ml
of this solution,
when
added
to 1.35 ml IT0O, gave an absorbance
of
0.30 at 430 nm (approximately
0.6%, w/v).
4.
Wet-as/tiny
reagent
This consisted
of 9.0 ml chloric
acid
reagent,
2.0 ml chromic
acid solution,
0.5 ml “70%”
IIC1O4, and,
optionally,
about 1 ng iodine per milliliter
of mixture.
5. H2S01,
10 N
Prepared
from concentrated
IT0SO4 which had
been purified
by heating
with 1/
of its volume of concentrated
HCI
(8, 9) to remove iodine and mercury.
6.
Arsenite
solutions
Stock solution
was prepared
by dissolving
in H20 19 g AsoOa (Fisher
Certified
Reagent),
7.8 g NaOH,
and 2.0 g
NaCl, and then diluting
to 500 ml. Working
solution
was prepared
by
diluting
stock solution
1 :10 with TTOO.
1.
Pointed
out by Dr. M. C. Sanz
(Central
Laboratory,
University
Hospital,
Oeneva,
land) in December 1966 in a lecture given at Georgetown
University
Medical
School,
ton, D. C., under
the auspices
of the American
Association
of Clinical
Chtemmsists.
Switzer-
Washming.
206
HOCH
& LEWALLEN
Clinical
Chemistry
7. (Jeric sulfate
solution
(‘e(804)0,
3.0 g, aitirydrous
and purified
(Fisher),
was suspended
in 4.8 ml 10 N 112804, diluted
to 100 ml with
1120, and stored at least 1 week to let tile precipitate
settle.
8. Blank solution
A mixture
of 0.80 ml “70%”
HC1O4 and 0.25 ml
0.6% chromic
acid solution
was heated until -the excess HO had evaporated;
0.134 ml 10 N 112804 and 2.66 ml arsellite
working
solution
were added.
9.
lodate standard
KH(lOq)o,
2.7 rug (U. Frederick
Smith Co.),
was dissolved
in 10.8 ml of tile blank solution
and further
diluted
approximately
1 :100 by weight
with
the blank
solution.
Apparatus
Wet -ashing tubes
Pyrex test tubes, 76-78 mm long and 13 mm in
external
diameter,
were weighed
to the nearest
0.1 mg and the weights
permanently
inscribed
on the tubes.
Heating
bath
This consisted
of a beaker half full of glass beads
as described
previously
(5). Heat was provided
from above by a 250-w
infrared
lamp directed
at an angle of about 45#{176}
to the horizontal
and
placed at a distance
of about 14 cm from the far edge of the beaker.
This arrangement
kept the beads on the surface
at 140-150#{176}.
Spectrophotometer
and accessories
A Beckman
DB spectrophotometer
with a tungsten
light source was adapted
to hold a semimicrocuvet
made of Herasil
and having
an optical
path of 5 mm, an
inside width of 4 mm, and outside dimensions
of 7.5 X 12.5 X 45 mm.*
Black plastic
insulating
tape was placed on the light-entry
side of the
cuvet. A window 2.0 mm high and 3.6 mm wide was cut out so that light
could reach neither meniscus
nor walls. Water from a reservoir
exposed
to room temperature
without
further
regulation
was circulated
through
the cell holder. Absorhance
was recorded
by a Beckman
potentiome-tric
recorder’ spanning
5 in. for an absorhance
range of 1.0. The temperature
of the cuvet contents
at the end of each run was measured
to the nearest
0.03#{176}
with a plastic
thermistor
microprohe
of a Telethermometer.t
Polyethylene
micro pipets
These were fabricated
in the laboratory
and calibrated
to deliver
ill the following
sizes: 0.5, 0.632, 1.09, 6.67,
and 12.3 l. One pipet was calibrated
for variable
delivery
between
10.0 and 25.0 i in increment-s
of 1 .l.
Procedure
Precpifafion
-
of Protein
into the bottom
60 l 110, 12.3 l
*PreCiSioIl
tYellow
(‘elis,
Springs
Tue.,
of a wet-ashing
serum, and, with
New
imistruniemit
You’k,
tube were pipetted
in succession:
mixing,
0.30 ml 17% HC1O4. These
N. Y.
(‘o., Yellow
Sprimigs,
Ohio,
Vol. 15, No. 3, 1969
MEASUREMENT
OF IODINE
207
proportions
are similar
to those used by Fischl
(10).
After 3 mm. the
tubes were centrifuged
for 5 mm. at 1800 rpm in an angle centrifuge.
The bottoms
of the tubes were 9 cm from the axis. The supernate
was
aspirated
with a narrow-tipped
polyethylene
l)ipet.
The precipitate
was not washed.
When labeled
iodide was added -to
serum, the precipitate,
after removal
of tile supernate
(and not further
washed),
retained,
on the average,
6.3% of the label. In the absence of
dietary
iodine supplements,
serum iodide levels after an overnight
fast
averaged
about 0.2 g/100
ml (ii).
IJirder these conditions
iodide
retamed in the unwashed
precipitate
voulcl result in a i)ositisre error in
PBI measurement
of only 0.013 g/l00
ml. For appreciably
higher
iodide levels the correction
can be computed
from the total iodine-PB
I
difference.
Washing
the precipitate
by swirling
it in the wash liquid
has the disadvantage
that portions
of the precipitate,
on centrifuging,
may float, and when spread
over the tube wails may yield low results
on subsequent
ashing.
Wet-Ashing
To all samples
0.072 ml of wet-ashing
reagent
was added. Tile ashing
tubes were placed on top of the beads of the heating bath, leaning almost
vertically
against
tile edge of -the beaker.
It proved to be easier to ash
the precipitated
protein
of serum than to ash whole serum as such. The
amount
of supplementary
chioric acid reagent
(Reagent
2) that had to
be added during ashing was less for precipitated
protein
than for whole
serum. Routinely,
three portions
(or for whole serum four portions)
of
0.02 ml each of chloric acid reagent
were added at the following
times:
1-2 mm., at 3-6 miit. (when the precipitate
was almost completely
dissolved),
arid at 7-10 mm. (or for whole serum, at 1, 3, 6, and 10 mm.).
After a total ashing time of 25-30
the digest had an amber color
and was nonwetting
to the glass. Sometimes,
at lower temperatures,
red crystals
appeared
during
the ashing.
Preparation of Samples for Cerate Fading
All samples
were next brought
to uniform,
optimal
concentrations
of
HC1O4, H2S04, and arsenite.
Tube contents
of HC1O4 remaining
after
ashing
were somewhat
variable
and were measured
individually
by
weight. The contribu-tion
by weight of materials
other than HC1O4 was
negligibly
small and was ignored.
The tubes were cooled and weighed.
Tile weigilts
of the contents
(2535 mg) divided
by 1.71 (the density
of constant-boiling
HC1O4) were
equated
to the residual
volume
of TIC1O4. in succession
were added
6.67 pJ 10 N 112804, 133 pJ arsenite
working
solution,
and sufficient
208
HOCH
& LEWALLEN
Clinical
Chemistry
“70%”
HC1O4 to bring the total volume of HC1O4 to 40 I (total volume
180 l).
The tubes were tipped and rotated
until all of the wall except
for a few millimeters
at the top was wetted, then they were capped with
Parafilm
and centrifuged
to collect the fluid at the bottom.
The total
volume of 180 jLl sufficed for duplicate
fading-rate
measurements,
each
performed
on 78 j.i.
Determination
of Cerat.
Fading Rates
Measurements
were begun
20-30 mm. after
addition
of arsenite
Heating
to 50 or 60#{176}
following
arsenite
addition
(5) was found
to be
unnecessary
and has been omitted.
Of the contents
of the ashing
tube, 78 In were transferred
-to the cuvet. After temperature
equilibration of the sample,
0.5 pi eerie sulfate
solution
was pipetted
onto the
bent bottom end of a polyethylene
stick; the latter was then dipped into
the cuvet with slight rotation
for #{189}-i
see. to mix. Care was taken that
the position
of the cuvet remained
unchanged
from that used to adjust
the 100% transmission
reading.
The volume
of eerie sulfate
solution
was adjusted
so that the initial absorbance,
as registered
4-6 sec. after
addition
to 78 JLl of a slowly fading
solution,
was 0.77 -to within about
1%. The elapsed time, s, in units of 15 sec., between absorbance
readings
of 0.50 and 0.26 was measured
on the recorder
chart to within 1.5 sec.
or better.
The temperature
of the cuvet contents
at the end of the run
was recorded.
After each test the cuvet was rinsed with 1120 and airdried by suction. Drying required
about 30 sec.
The values for the reciprocal
time, a measure
of the fading
rate
coefficient,
were corrected
to refer to 25.0#{176}
by the empirical
equation
(100/s
=
(1
+
-
0.1014
t)
0.0646L.t)
where
r is a temperature-corrected
value proportional
to the fading
rate coefficient,
s is the time in 15-sec. units, and t = 25.0
t, where t
is the temperature
in degrees
centigrade.
The equation
used formerly
(5) contains
an error:
“s”
should
read “s/100.”
This equation
was
of the form
-
r
=
100/s
+
(100A/s
+
B)t
The factor 100/s inside the parentheses
allowed
for the iodine-concentration
dependence
of -the temperature
correction.
A better
approximation
is obtained
by substituting
r for this factor. When this is done
and the resulting
equation
is solved for r, an equation
of the form of
Equation
1 is obtained.
The coefficients
0.1014 and 0.0646 were obtained
from calibration
experiments
at four concentrations
of iodine over the
temperature
range of 23-27#{176}.
The sample value for r was converted
to total iodine content
by ref-
Vol. 15, No. 3, 1969
MEASUREMENT
209
OF IODINE
erence to the calibration
graph of Fig. 1. This value was corrected
by
subtraction
of the iodine conten-t of an appropriate
reagent
blank run
under
similar
conditions.
The time units employed
here and the ratio of absorbances
over
20
0
0
0
I
ng iodine
Fig.
minutes
1.
Calibration
for
absorbance
2
3
per 0.18 ml
of nonashed standards. Ordinate, r, is the
to decrease
from 0.50 to 0.26 at 25.0g.
reciprocal
time
in
quarter
which s and r have been measured
are admittedly
nonconventional
and
were adopted
for operational
convenience.
To facili-tate
comparison
with the data of others,
the following
conversion
formulas
are presented.
The half-time
of disappearance
of cerate absorbance
is given
by t112 = (100/4) (1.060) (1/r) mm., where the first factor converts
the
units of r to reciprocal
minutes
and the second
factor
corrects
the
absorbance
ratio from 0.50:0.26
to 0.50:0.25.
Similarly,
the first-order
reaction-rate
constant
for cerate decolorization
is given by k = (0.654)
(4/100)r
per minute.
Results and Discussion
Calibration
For the purpose
of reference,
unashed
iodine standards
were faded
in the blank solution
(Reagent
8), in the preparation
of which HCIO3
had been avoided.
A polyethylene
pipet delivering
1.09 .l was used for
adding
undiluted
or diluted
standard
solution
to 180 il of the blank
solution.
Alternatively,
the blank solution,
to which 1.09 j.tl undiluted
standard
solution
(containing
1.785 ng iodine)
had been added
per
180 l, was diluted
1 :1, 1 :2, 1:3, etc., with the blank solution.
210
HOCH
& LEWALLEN
Clinical
Chemistry
F1igume 1 shows tile plot of r against
the nanograms
of iodine coIltamed in 180 l, which sufficed for two fading tests. As can be seen, this
calibration
curve
is linear
at the higher
iodine
concentrations
butdeviates
from linearity
for tile lower concentrations.
Taking the slope iii
Fig. 1 as a measure
of sensitivity,
tile latter was 9.6 at iligh concentration, decreasing
ill the range
below 0.2 ng/180 l to 6.0 at tile origin. The
detection
limit was of the order of 0.04 ng iodine (or 0.02 ng iodine per
fading test volume)
and the upper limit-i.e.,
where the error of reading
the time began to interfere-was
3 ng iodine per 180 l. The curvature
in the low range is caused by the reduction
of cerate
ion by 3-valent
chromic
ion (6). This interference
decreases
with higher
iodide concentration,
the reduction
by Cr(III)
possibly
being catalyzed
by iodide
as is the reduction
by arsenite.
In order to avoid reading
in the lower,
curved portion
of the calibration
line, iodide has been added to the wetashing reagent.
Recovery Experiments
In any method for iodine analysis
based on the cerate-arsenite
reaction, the recovery
must either be quantitative
or be corrected
for. In
the event it is not quantitative,
an appropriate
correction
requires
that
the recovery
1)e sufficiently
constant
for a measured
recovery
factor to
1)0 applied
properly
to unknown
samples.
Tnitially in our hands, it was difficult to render recovelies
sufficiently
constant.
Experiments
yet to be reported
(12)
have shown that this
variability
is caused primarily
by the formation
during
digestion
of art
agent
which inhibits
the cerate-arsenite
reaction.
This finding
confirms the observation
of O’Neal
and Simms
(13)
that “an inhibitor
of catalysis
is present
in the evaporated
digestion
mixture.”
The
nature of the inhibitor
was stated to he unknown,
and no further
studies
were made by these authors
to elucidate
its source.
The digestion
reagent
and conditions
recommended
in the present
paper
have been selected
so as to minimize
formation
of inhibitor,
thereby
circumventing
the necessity
for recovery
corrections.
The choice of 140-150#{176}
as the -temperature
range
for ashing
was
based on the following
observations.
Appreciable
amounts
of inhibitor
were found if (1) the total volume of HC1O5 used exceeded
0.135 ml or
(2) the ashing were performed
at temperatures
lower than 130#{176}.
The
larger
volumes
of 11C103 prolonged
tile ashing
time because
of the
larger
volumes
of 1120 and IIC1O3 that had to be evaporated.
At
temperatures
below 130#{176}
the oxidation
of protein
was so slow that appreciable
11(1103 was lost by evaporation.
This had to 1)e replaced
(luring
211
MEASUREMENT OF IODINE
Vol. 15, No. 3. 1969
digestion
to prevent
reduction
of the hexavalent
chromium,
the amber
color of the latter being used as a guide. These findings
suggested
that
formation
of inhibitor
might be the result of longer exposure
to heat
rather
than the larger volumes
of 11C1O3 as such. Accordingly,
an attempt
was made to shorten
digestion
time. With
the heating
bath
initially
at 140#{176},
the temperature
dropped
to 130#{176}
when the tubes were
inserted
and then rose to 140-150#{176}.
Supplementary
HC1O3 was added in
0.02-ml portions
at 1, 3, and 7 mm. or at 1, 3, 6, and 10 mill. after the
start. Under
these conditions
ashing
was complete
at 23-30 mm., as
judged by the amber color and nonwetting
property
of the digest and
by the negligible
weight loss on continued
heating.
Details on formation
of inhibitor
at temperatures
higher
than 150#{176}
will be reported
in the
future
(12).
Quantitative
recovery
of added iodine requires
not only that iodine
not be lost during ashing, but also that inhibitor
formation
be negligible.
The ashing technic described
was adopted
for routine use. The following
2
Fig.
2.
Recovery
of
ashed
standards.
Amuounts
contained
in 0.180 ml suffice for duplicate
fading
tests.
Open
circles,
3
times 0.02 ml additional
UC1O,,
“
ashed
25 mm.
at 140-150#{176};
dots, 4 times 0.02 ml additional
.E
HClOa, ashed 25 mm. at 140155#{176};
triangles,
4 times 0.02 ml .2
additional
HClO,, ashed 33 miii.
at
120-140#{176}; crosses,
ashed
c
multiples
of 4.3 I serum.
I
0
reagent
blank
2
0
ng
experimental
quantitative
results
recovery
demonstrate
of added
of serun
or IIClO4-precipitated
I)ifferent
amounts
of iodine
that
iodine
either
the
iii
serum protein.
standards
were
iodine
added
technic
gives
virtually
the absence or Pl50flCC
ashed
under
conditions
212
HOCH
& LEWALLEN
Clinical Chemistry
used routinely
for ashing whole serum (0.072 ml of wet-ashing
reagent
followed
by 0.02-mi additions
of supplementary
HC1O3 at 1, 3, 6, and
10 mm.)
and HClO4-precipitated
protein
(0.072 ml of wet-ashing
reagent
followed
by 0.02-ml additions
of supplementary
HC1O3 at 1, 3, and 7
mm.). Figure
2 shows the recovery
of iodine in these ashed standards.
The principal
finding is that recovery
is not complete.
Tile deficit increases
between
0 and 0.8 ng of added iodine and thereafter
remains
constant.
This effect is compatible
wi-th the presence
in the digest of
an inhibitor
which inactivates
catalytic
iodine by reversibly
complexing
it. If the deficit had been caused by evaporation
of HI or T, one would
expect the same proportionate
loss at all levels, resulting
in an absolute
loss increasing
with total iodine. The quantitative
effect of the deficit
on iodine measurements
will be further
considered
below.
Perchloric
acid protein
precipitates
were prepared
in duplicate
from
l2.3-j.l aliquo-ts of serum as described
under
Procedure.
To one precipitate was added 1.785 ng iodine, and the samples were ashed together
with a reagent
blank. The results
are given in Table 1. The recovery
of iodine added before ashing ranged
from 96 to 106%.
In all the fading
tests for the data presented
here, care has been
taken to add the eerie sulfate
rapidly
and with a minimum
of stirring.
It has been observed
repeatedly
tha-t either rapid twirling
of the polyethylene
stick that is used for introducing
the eerie sulfate,
or stirring
longer than 1 sec., caused a lowering
of the fading rate. Since solutions
of iodide in the acid medium,
as used in the fasting
tests but in the
Table
1. PRoTEIN-BouND
IODIN
PBI#{176}
Serum
(g/1OO
Present
RECOVERY
EXPERIMENTS
I recover-sit
(%)
in!)
PBI
(pg/lOOm!)
RE
8.1
104
8.1
MA
MA
MA
6.5
-
6.5
6.4
6.0
103
HO
BE
MA
ST
FL
P0
EN
5.6
3.4
4.8
4.6
4.1
3.5
2.9
103
99
96
99
106
98
5.6
5.1
RE
RE
2.6
1.7
96
103
2.6
1.5
0.8
101
-
FL
*
E AND
study,
uncorrected
101
-
6.3
5.6
-
-
3.9
4.0
for recovery.
$ Percent recovery
of 1.785 ng iodine
By alkaline ashung method, performed
added
t-o precipitated
protein.
by Bio-Scienee
Laboratories,
Van Nuys,
Calif.(
14).
Vol. 15. No. 3, 1969
MEASUREMENT OF IODINE
213
absence of cerate, have been kept at room temperature
for over 1 year
without
change
in iodine concentration,
the iodine lost on exposing
a
large
surface
of the solution
by stirring
after
cerate
addition
is
probably
in the elemental
state.
Chromic acid has been used as an indicator
of HC1O3 being in excess,
while oxidizable
material
is still present,
to ensure that iodine remains
in the pentavalent
state as 11103. Chromate
has also been thought
to be
capable
of preventing
losses of iodine from fuming
mixtures
of iodic
and perchloric
acids (1).
However,
by itself hot Cr03 does not oxidize
HI to 11103 rapidly
enough, and HI is lost by evaporation
as shown by
the following
experiment.
To 125 l 0.6% Cr03 was added 10 l iodide
solution
containing
16.4 ng iodine; four 13.5-l
aliquots
of this solution
were digested
for 10 mm. at 260-265#{176}
without
and with 2, 10, and 20
“70%”
HC1O4. With 10 and 20 /Ll of HClO4 present,
0.53 and 0.76 ng of
iodine (32 and 46%, respectively)
were recovered.
The iodine content
of the other two digests
was negligibly
small. Chromic
acid causes the
digest to become nonwet-ting
to the glass when all organic
material
has
been eliminated.
With relatively
small amounts
of Cr03 present,
residues of organic
material
or excess 1120 appear
to lower the surface
tension,
so that -the digest
is prevented
from becoming
nonwetting.
With higher concentration
of Cr03, the digest may become nonwetting
before all excess 1120 has evaporated,
thus diminishing
its value as an
indicator
of completion
of ashing.
In principle
it is difficult to establish
rigorously
the accuracy
of a
method
for serum
iodine.
Catalysis
of the cerate-arsenite
reaction,
while highly
sensitive
to iodine,
is not entirely
specific.
Of greater
practical
importance
is the interference
caused by noncatalytic
reduction of cerate by a variety
of organic
substances.
In addition,
inhibition
of the iodine-catalyzed
reaction
has been demonstrated
for a variety
of
biologic materials
(3).
Thus, elimination
of organic material
by ashing
generally
has been considered
mandatory.
Destruction
of organic
materials
is not, however,
necessarily
a guarantee
of elimination
of interference.
For example,
inhibitory
were the metals mercury,
silver
(15),
gold (16),
and platinum
(6),
which may remain
after ashing.
In the
present
work yet another
kind of interference,
an inhibitor
formed
during ashing,
emphasizes
the necessity
of establishing
criteria
for accuracy.
Several
tests of accuracy
indicate
the reliability
of the present
technic.
A known amount
of iodine ashed together
with the precipitated
protein (internal
standard)
was recovered
completely
(Table
1). Had this
not been the case, a recovery
correction
would have been indicated.
however,
for such a correction
to be meaningful
required
that the re-
214
HOCH
& LEWALLEN
Clinical
Chemistry
covery
fraction
be the same both for sample
and for sample
plus
internal
standard.
It may be argued -that if an abnormal
recovery
(high
or low) occurred
as a consequence
of a faulty
value for the sample
digested
without
added standard,
a correction
by dividing
by the fraction of standard
recovered
would make the corrected
sample value still
more
erroneous.
If the ei’ror derived
from
a faulty
sample-plusstandard
value, a correction
of the sample value would not be justified.
Routinely,
then, tile recovery
was checked,
but no correction
was applied to the sample
value, instead,
those samples
of an internal
standard
added before
ashing
was
factory
were repeated
until recovery
values of 100
for which
considered
± 6% were
recovery
unsatisobtained.
The recovery
of iodine from ashed
standards
closely
approached
theoretical
values
(Fig. 2). The small negative
deviation
affects
PB1
measurements
as follows.
A sample value of 0.74 ng/12.3
l (PBI
=
6 g/100
ml) should be considered
too low by 0.09 ng/12.3
.l (read
from the graph
as the vertical
distance
between
the theoretical
and
experimental
lines).
For smaller
sample
values
the absolute
error
would be less-e.g.,
0.02 ng in 0.12 ng/12.3 /Ll (PBI = 1). These underestimates
of PBI by 12-16%
could have been, but were. not, corrected
for in Table 2.
As a further
check on accuracy
a serum was ashed in amounts
of
ix, 2X, and 3x 4.3 ,a1. The following
values
for the nanograms
of
iodine per sample
were found:
0.289, 0.628, 0.945. The highest
found
value was located on tile plot for ashed standards
(Fig. 2). The corresponding
abscissa
was partitioned
into three
equal
portions.
The
ordinate
values
given above agreed
closely
with the ordinates
read
from the curve (0.283 and 0.605), thus reproducing
the curvature
withill
the limits of precision.
This can be interpreted
as indicating
that ashing
of serum produces
no greater
interference
than ashing of iodine standards alone.
References
1. Zak, B., willard,
11. II., Myers,
G. B., amid Boyle,
A. J., Cliloric
acid method
for determinalion of protein-hound
iodine.
Anal. Chein. 24, 1345 (1952’).
2. Barker, S. B., llunmplirey,M. J., and Soley, M. H., The clinical
determination
of proteinbound iodine. J. din. Invest.
30, 55 (1951).
3. Foss, 0. P., Hankes,
L. V., amid van Slyke,
D. P., A study
of the alkalimie ashing
method
for determination
of protein-bound
iodine
imi serum.
dim. C/jim. Acta 5, 301 (1960).
4. Sanz,
M. C., Brechhuhler,
T., and Green,
I. J., The ultrarnicro-determunation
of total and
protein-bound
iodine.
C/in. C/jim. Aeta 1, 570 (1956).
5. Hoch, II., Sinneft,
S. L., Miller,
P. 0., and Mahady,
1. B., Dialyzahle
thyroid
hormonebinding
material
iii Imuman scm-urn. Bioe1iemi.try
4, 931 (1965).
6. Hodi,
H., Sintiett,
S. L., and MeGavack,
T. II., The (letermnination
of iodine
in biologic
material.
C/in. Chenm. 10, 799 (1964).
7. loch,
H., Simmnett, S. L., amid McGavack,
T. H., A critical
review of the factors
involved
in the determination
of protein-hound
iodine.
Joycc-Loebl
Review
2, 13 (1965).
Vol. IS. No. 3, 1969
S.
9.
10.
]1.
12.
215
OF IODINE
B., Technique
de dosage
de l’iode proteique
applicable
cmi clinique.
Rev. Franc.
Etudes
C/in. Biol. 1, 706 (1936).
Piette,
C., amid Piette,
M., Dosage
de l’iode proteique
du serum. Ann. Bioi. C/in. (Paris)
7-9,
1 (1960).
Fischi,
.J., Determination of protein-bound
iodine
in microanioumits of serum
or plasma.
Gun. C/mi-rn. Acta 1, 462 (1956).
Rail, J. E., Bobbins,
J., and Lewallen,
C. G., “Time Thyroid.”
In Time hormones
(Vol. 5),
Pincus, G., Thimana,
K. V., and Astwood,
E. B., Eds. Acad. Press, New York, 1964, p. 280.
Hoch,
H., audi Lewallen,
C. U., Inhibition
of time cerate-arsenite
reaction
following
cimlom-ic
Besson,
acid
13.
MEASUREMENT
digestion.
Unpublished
data.
14.
O’Neal,
L. \V., amid Sirnms,
E. S., Determination
of protein-bouuid
iodine
in plasma
or
serum. Anm. J. C/in. Pat/mo!. 23, 493 (1953.
Gaffney,
G. W., Gregerman,
B. I., Yiemmgst, M. J.,amid
Shock, N. W., Serum
protein-hound
iodinc concentration
iii blood
of euthyroid
nmen aged
18 to 94 years.
J. Geronto!.
15,
234 (1960).
15.
Sandeil,
E.
B.,
Mikroclmim.
16.
Fisher,
bound
A.
amid Kolthoff,
Acta
B.,
Levy,
iodine.
New
I.
M.,
Micro
determimmatiomi
of
iodine
by
a catalytic
nietliod.
of
low
protein-
1,
9 (1937).
B. P., amid Pm-ice, VT., Gold-au
Engi. J. Med. 273, 812 (1965).
occult
cause
serum