Microchemical Journal 123 (2015) 914

Contents lists available at ScienceDirect

Microchemical Journal

journal homepage: www.elsevier.com/locate/microc

Reliability in standardization of sodium thiosulfate with

potassium dichromate
Toshiaki Asakai , Akiharu Hioki
National Metrology Institute of Japan, National Institute of Advanced Industrial Science Technology, 1-1-1 Umezono, Tsukuba, Ibaraki 305-8563, Japan

a r t i c l e i n f o a b s t r a c t

Article history: Sodium thiosulfate is often standardized with potassium dichromate. In the standardization, iodine (triiodide)
Received 23 March 2015 liberated by potassium dichromate in an acidic potassium iodide solution is titrated with a sodium thiosulfate
Received in revised form 12 May 2015 solution. The iodine liberation process signicantly affects the titration results. In the present study, the accuracy
Accepted 12 May 2015
of the liberation process was examined by assaying potassium dichromate through two different paths: assaying
Available online 16 May 2015
directly by coulometric titration with electrogenerated Fe(II), and assaying by gravimetric titration through the
Keywords:
iodine liberation reaction with a sodium thiosulfate solution of which concentration was standardized by coulo-
Coulometric titration metric titration with electrogenerated iodine. The accuracy of the standardization of a sodium thiosulfate solution
Oxidation-reduction titration by potassium dichromate was discussed from the apparent assays of potassium dichromate under different
Certied reference material measurement conditions.
Potassium dichromate 2015 Elsevier B.V. All rights reserved.
Sodium thiosulfate
Potassium iodate

1. Introduction The iodine liberation process is signicantly affected by the amounts

Sodium thiosulfate is a useful reducing agent for a titrant solution in
volumetric analyses such as iodometry and the Winkler method to
measure dissolved oxygen [1,2]. Oceanography community requires
the accuracy and the reproducibility of the measurements of dissolved
oxygen; the accurate standardization procedure of sodium thiosulfate
is needed. The agent is often standardized with potassium dichromate
and, on the other hand, is usually standardized with potassium iodate
in Japan [36]. Iodine (triiodide ions) liberated by potassium dichro-
mate or potassium iodate in an acidic potassium iodide solution is titrat-
ed with a sodium thiosulfate solution to standardize the thiosulfate
[reactions (1), (2) and (3)].
K2 Cr2 O7 6KI 7H2 SO4 Cr2 SO4 3 4K2 SO4 7H2 O 3I2
KIO3 5KI 3H2 SO4 3K2 SO4 3H2 O 3I2
2Na2 S2 O3 I2 2NaI Na2 S4 O6
Corresponding author. Tel.: +81 29 861 6881; fax: +81 29 861 6890.
E-mail address: t-asakai@aist.go.jp (T. Asakai).
of acid and potassium iodide added, the waiting time for the liberation,
and light; therefore, the process plays a key role for the accuracy of stan-
dardization of sodium thiosulfate. The authors reported an accurate
standardization procedure for a sodium thiosulfate solution with potas-
sium iodate [7].
It has been pointed out that there are some biases in the standardi-
zation of sodium thiosulfate with potassium dichromate [3,812].
Some researchers reported that excess consumption of thiosulfate
occurred in an acidic medium due to air oxidation of iodide ions [8,9].
Hahn described that the main reason of the excess consumption of thio-
sulfate in a weak acidic medium was not air oxidation of iodide, but a
complex between Cr(III) and thiosulfate, which reacted with iodine
slowly [3,12]. Kolthoff and Belcher reviewed the studies on the titration
1 procedure [3] and mentioned the appropriate conditions to ensure
complete and stoichiometric reaction with dichromate: the hydrogen
ions concentration 0.2 mol L1 to 0.4 mol L1, 2% of potassium iodide,
and standing for 10 min to 15 min before starting titration with
2 thiosulfate.
The impact of the biases has remained obscure because the accuracy
of the standardization of thiosulfate has been discussed by using relative
methods for conventional titration. Thiosulfate is usually standardized
with sublimed iodine whose purity is hypothetically regarded as 100%.
3
In the present study, the accuracy of the standardization was absolutely
evaluated by coulometric titration (Fig. 1). Coulometric titration abso-
lutely yields an accurate oxidimetricreductometric assay based on
Faraday's Law. The accuracy of the iodine liberation process was exam-
ined by comparing two assays of potassium dichromate using different

http://dx.doi.org/10.1016/j.microc.2015.05.012
0026-265X/ 2015 Elsevier B.V. All rights reserved.
10 T. Asakai, A. Hioki / Microchemical Journal 123 (2015) 914

1 g, and potassium iodide and solutions were weighed with resolution

of 10 g.

3. Results and discussion

3.1. Determination of sodium thiosulfate by coulometric titration and

uncertainty estimation

The concentration of a sodium thiosulfate solution (3 g to 8 g) stan-

dardized by coulometric titration was, for example, 208.653 mmol kg1
(RSD 0.007%, n = 5) (RSD means relative standard deviation, and n is
the number of measurements under a repeating condition). The uncer-
tainty sources for the concentration of the solution (ca. 3 g of ca.
200 mmol kg1) were as follows (the number in each parenthesis indi-
cates the relative standard uncertainty): repeatability (0.0032%, RSD of
the mean, n = 5), Faraday constant (0.000 002 2%), a standard resistor
(0.000 65%, calibration and the inuence of temperature), a frequency
counter (0.000 05%), a voltmeter (0.000 15%) and weighing (0.0019%,
sensitivity and linearity of the balance). Finally, the combined standard
uncertainty was 0.0038% relative and the expanded uncertainty with a
coverage factor 2 was 0.0076% relative. The coulometric titration
Fig. 1. Experimental design for the assay of potassium dichromate to investigate the iodine allowed the authors to discuss the accuracy of the iodine liberation reac-
liberation reaction. tion by potassium dichromate in the order of magnitude of 0.01%. A so-
dium thiosulfate solution is stable enough to be required in the present
study [16,17].
titrations based on coulometric titration. Firstly, potassium dichromate
was directly assayed by coulometric titration with electrogenerated 3.2. Inuence of the amount of potassium iodide added
Fe(II) [13]. Secondly, potassium dichromate was assayed by gravimetric
titration through the iodine liberation reaction with a sodium thiosul- Approximately 0.2 g of potassium dichromate were dissolved in
fate solution, of which concentration had been standardized by coulo- 100 mL of non-deaerated water followed by adding 1 g to 6 g of potas-
metric titration with electrogenerated iodine. The accuracy of the sium iodide and 1 mL of 9 mol L1 sulfuric acid in turn; after the acid
standardization of a sodium thiosulfate solution by potassium dichro- addition, iodine liberation started. The iodine liberation process moni-
mate was discussed from the apparent assays of potassium dichromate tored by constant-voltage biamperometry using a dual platinum-chip
under different measurement conditions. electrode (applied constant-voltage 500 mV) is shown in Fig. 2; the
indicator current corresponds to the amount of iodine liberation.
Under the condition using 1 g to 2 g of potassium iodide and 1 mL of
2. Material and methods 9 mol L1 sulfuric acid, lacking of iodide ions in the solution led to the de-
position of solid iodine, and the indicator current drastically decreased in
Constant-voltage biamperometry (a dead stop method) was utilized unforeseeable timing after the initial increase. Under the conditions using
to measure the amount of liberated iodine under several conditions as a 3 g to 6 g of potassium iodide and the same amount of sulfuric acid, the
preliminary experiment. A DC source and a digital multimeter with a
dual platinum-chip electrode (each platinum chip: 0.5 mm 0.5 mm
long) were used for constant-voltage biamperometry. Constant voltage
was applied between the platinum pair, and the current was monitored.
The current had good proportionality to the amount of liberated iodine
[7]. A dual platinum-plate electrode (each 5 mm 5 mm) was used in
gravimetric titration to increase sensitivity as described previously [7,
1315].
The experimental set-ups and the procedures of coulometric titrations
for potassium dichromate and a sodium thiosulfate solution were de-
scribed previously [7,1315]. As described in the previous report [13],
high purity potassium dichromate as a certied reference material
(CRM) was established by coulometric titration with electrogenerated
Fe(II). The certied value (mass fraction) of the CRM, NMIJ CRM 3002-a,
is 99.974% 0.011%, where the value following indicates the expanded
uncertainty with the coverage factor 2.
Gravimetric titration was carried out to assay potassium dichromate
with a sodium thiosulfate solution through the iodine liberation reac-
tion in the following procedure: approximately 0.2 g of potassium di-
chromate were placed in a 200 mL tall beaker, it was dissolved in
100 mL of water, potassium iodide and 9 mol L1 sulfuric acid were
added, and the solution was titrated with a sodium thiosulfate solution.
The concentration of the sodium thiosulfate solution was standardized Fig. 2. The iodine liberation process monitored by constant-voltage biamperometry for
0.2 g of potassium dichromate using a dual platinum-chip electrode (applied constant-
in advance by coulometric titration with electrogenerated iodine [7]. A voltage 500 mV), stirring at 900 revolutions per minute (rpm). The sample was dissolved
plastic syringe with a PEEK needle was used to weigh the sodium thio- in 100 mL of non-deaerated water, and 1 g to 6 g of potassium iodide and 1 mL of
sulfate solution. Potassium dichromate was weighed with resolution of 9 mol L1 sulfuric acid were added.
 T. Asakai, A. Hioki / Microchemical Journal 123 (2015) 914
decrease of the indicator current did not occur; these conditions seemed
to be enough to completely generate dissolved iodine. In the previous
study [7], potassium iodate (0.15 g), whose redox equivalent is almost
the same as that of potassium dichromate (0.2 g), liberated iodine
completely in a couple of seconds after acid addition under the condi-
tion using 3 g to 5 g of potassium iodide and 1 mL of 9 mol L1 sulfuric
acid; consequently, no waiting time for the iodine liberation was neces-
sary. Potassium dichromate conversely required more than 5 min for
the complete iodine liberation according to Fig. 2.
Gravimetric titration was performed to assay potassium dichromate
in a darkroom using a sodium thiosulfate standardized by coulometric
titration. The dependency of the apparent potassium dichromate assay
on the amount of potassium iodide added is shown in Fig. 3. Approxi-
mately 0.2 g of potassium dichromate in a beaker was dissolved in
100 mL of deaerated or non-deaerated water followed by adding 1 g
to 5 g of potassium iodide and 1 mL of 9 mol L1 sulfuric acid in turn;
liberated iodine was titrated at 10 min after the acid addition with a so-
dium thiosulfate solution, of which concentration was standardized by
coulometric titration. The beaker was stoppered with Paralm during
the liberation to avoid the volatilization of iodine. Fig. 3 includes the cer-
tied value of potassium dichromate determined by coulometric titra-
tion with electrogenerated Fe(II) [13] and also the dependency of the
apparent potassium iodate assay on the amount of potassium iodide
added [7]. Liberated iodine by potassium iodate was titrated immedi-
ately after iodine liberation starting. The similarity of the assay values
of potassium dichromate and potassium iodate was a coincidence
because of high purity materials. The apparent assay of potassium di-
chromate with 1 g of potassium iodide and 1 mL of 9 mol L1 sulfuric
acid using deaerated or non-deaerated water was more than 101%.
The apparent assays of potassium dichromate had a minimum for
both the results using deaerated water and those using non-deaerated
water; the minimum value with 2 g of potassium iodide was close to
the certied value. The titration with less than 2 g of potassium iodide
was not able to be accurately completed since the indicator current
drifted upward during the titration. In this iodometry, the indicator cur-
rent decreased with progress of titration, that is, the liberated iodine
was reduced with a sodium thiosulfate solution. The end point of this
type of reaction was at 0 A of the indicator current where the iodine
just disappeared. The apparent assay of potassium dichromate
corresponded to the mass of the sodium thiosulfate solution consumed
until the end point. The upward drift of the indicator current indicated
Fig. 3. Dependencies of the apparent potassium dichromate and potassium iodate assays
by gravimetric titration with thiosulfate on the amount of potassium iodide added. Iodine
liberation times after the acid addition were 10 min for potassium dichromate and 0 min
for potassium iodate. : 0.2 g of potassium dichromate dissolved in 100 mL of non-
deaerated water; 1 mL of 9 mol L1 sulfuric acid. : 0.2 g of potassium dichromate dis-
solved in 100 mL of deaerated water; 1 mL of 9 mol L1 sulfuric acid. : 0.15 g of potassi-
um iodate (NMIJ CRM 3006-a) dissolved in 100 mL of non-deaerated water; 1 mL of
9 mol L1 sulfuric acid [7]. : Certied value (99.974%) of potassium dichromate by
coulometric titration with electrogenerated Fe(II) (not related to the abscissa axis) [13].
11
the production of iodine or oxidants with progress of time. The drift
made it difcult to draw the titration curve and to determine the end
point accurately. The apparent assays were also signicantly affected
by the speed of titration. Less amount of potassium iodide and slow ti-
tration led to more evaporation of iodine and more inuence of light;
accurate titration results were not expected.
The iodine liberations were obviously not completed under the
condition using 1 g of potassium iodide and 1 mL of 9 mol L1 sulfuric
acid for either 0.2 g potassium dichromate or 0.15 g of potassium iodate
[7] as shown in the preliminary examination by constant-voltage
biamperometry (Fig. 2). Consequently, unreacted potassium dichro-
mate and potassium iodate remained in the solutions before starting a
titration with thiosulfate. The authors reported that unexpected good
results for potassium iodate under incomplete iodine liberation condi-
tions were caused by resupplied iodide ions with progress of the reac-
tion [7]. Such good results did not happen for potassium dichromate
as shown in Fig. 3. Judging from Fig. 3, it was inferred that excess sodium
thiosulfate was consumed in the presence of the unreacted potassium
dichromate. There could be the following three possibilities: (1) sodium
thiosulfate was insufciently oxidized, (2) it was decomposed under the
incomplete iodine liberation in an acidic medium, and (3) it interacted
with potassium dichromate or generated Cr(III) and lost the reducing
capability. If the remaining dichromate caused insufcient oxidation of
thiosulfate, it would result in excessive consumption of thiosulfate.
Rao and Sarma reported the direct standardization of thiosulfate by po-
tassium dichromate with copper(II) sulfate as a catalyst [18] and men-
tioned that dichromate directly oxidized thiosulfate to tetrathionate
state as usual. Therefore, excessive consumption of thiosulfate was not
caused by its insufcient reduction. Though the possibility of its decom-
position before reacting with remaining dichromate or iodine cannot be
excluded, as mentioned above Hahn reported a phenomenon relevant
to the third possibility [12]. It was the generation of the complex
between thiosulfate and Cr(III) in the presence of dichromate; conse-
quently, excess thiosulfate was apparently consumed. According to his
report, Cr(III) once formed does not react with thiosulfate, but the com-
plex between Cr(III) and thiosulfate is formed when the reduction of
Cr(VI) takes place in the presence of thiosulfate.
The reason of slight increase in the apparent assays of potassium di-
chromate under the condition using more than 3 g of potassium iodide
could be the signicant oxidation of potassium iodide during iodine lib-
eration (10 min) in an acidic medium depending on the amount of po-
tassium iodide added. Signicant air-oxidation of iodide ions in an
acidic medium without oxidants such as potassium iodate was not ob-
served [7,15]; however, Cr(III) could accelerate the oxidation reaction
of them (details are described in section Inuence of background
oxidation of iodide ions and vaporization of liberated iodine).
3.3. Inuence of the amount of acid added
Approximately 0.2 g of potassium dichromate was dissolved in
100 mL of non-deaerated water followed by adding 3 g of potassium io-
dide and 0.5 mL to 5 mL of 9 mol L1 sulfuric acid in turn; after acid ad-
dition iodine liberation started. The iodine liberation process was
monitored in the same manner as done for the examination of inuence
of the amount of potassium iodide added (vide supra); the results are
shown in Fig. 4. Under the condition using 3 g of potassium iodide and
0.5 mL of 9 mol L1 sulfuric acid, the iodine liberation was not complet-
ed. Under the condition using 3 g of potassium iodide and 1 mL to 5 mL
of 9 mol L1 sulfuric acid, the iodine liberation was apparently complet-
ed. It seems that more acids accelerate the iodine liberation reaction.
Gravimetric titration with thiosulfate was performed to evaluate the
dependency of apparent assay of potassium dichromate on the amount
of sulfuric acid used. A dual platinum-plate electrode was used for the
end-point determination. Fig. 5 shows the results of the gravimetric
titration together with the apparent assays of potassium iodate [7]. Ap-
proximately 0.2 g of potassium dichromate were dissolved in 100 mL of
12 T. Asakai, A. Hioki / Microchemical Journal 123 (2015) 914
Fig. 4. The iodine liberation process monitored by constant-voltage biamperometry for
0.2 g of potassium dichromate using a dual platinum-chip electrode (applied constant-
voltage 500 mV), stirring at 900 rpm. The sample was dissolved in 100 mL of non-
deaerated water, and 3 g of potassium iodide and 0.5 mL to 5 mL of 9 mol L1 sulfuric
acid were added.
deaerated or non-deaerated water followed by adding 3 g of potassium
iodide and 0.5 mL to 4 mL of 9 mol L1 sulfuric acid in turn; liberated
iodine was titrated with a thiosulfate solution at 10 min after acid addi-
tion. The results by gravimetric titration using 100 mL of deaerated
water and 1 mL or more of 9 mol L1 sulfuric acid were consistent
with those by coulometric titration with electrogenerated Fe(II); on
the other hand, those by gravimetric titration using non-deaerated
water increased with the increase in the amount of acid used. This fact
suggested that dissolved oxygen oxidized iodide ions in an acidic
media described in detail below [7]. The assay of potassium dichromate
dissolved in 100 mL of deaerated or non-deaerated water, under the
condition using 3 g of potassium iodide and 0.5 mL of 9 mol L1 sulfuric
acid, was more than 106%. The indicator current drifted upward during
the titration under the conditions of being short on acids as under those
of being short on potassium iodide (Fig. 3); accurate titration results
were not ensured. To conclude, an appropriate minimum amount of
Fig. 5. Dependencies of the apparent potassium dichromate and potassium iodate assays
by gravimetric titration with thiosulfate on the amount of sulfuric acid used. Iodine liber-
ation times after the acid addition were 10 min for potassium dichromate and 0 min for
potassium iodate. : 0.2 g of potassium dichromate dissolved in 100 mL of non-deaerated
water; 3 g of potassium iodide. : 0.2 g of potassium dichromate dissolved in 100 mL of
deaerated water; 3 g of potassium iodide. : 0.15 g of potassium iodate (NMIJ CRM
3006-a) dissolved in 100 mL of non-deaerated water; 3 g of potassium iodide [7]. : Cer-
tied value (99.974%) of potassium dichromate by coulometric titration with
electrogenerated Fe(II) (not related to the abscissa axis) [13].
acid is required in order to accurately carry out the standardization of
sodium thiosulfate with potassium dichromate. In addition, deaeration
is desirable to avoid the excess oxidation of iodide by dissolved oxygen.
3.4. Inuence of the iodine liberation time
Gravimetric titration with thiosulfate was carried out to evaluate the
dependency of apparent assay of potassium dichromate on the iodine
liberation time. Fig. 6 shows the results of the gravimetric titration to-
gether with the apparent assay of potassium iodate (NMIJ CRM 3006-
a) [7]. Approximately 0.2 g of potassium dichromate was dissolved in
100 mL of deaerated or non-deaerated water followed by adding 3 g
of potassium iodide and 1 mL of 9 mol L1 sulfuric acid in turn; liberated
iodine was titrated with a thiosulfate solution at 0 min to 20 min after
the acid addition. The apparent assays by gravimetric titration at
10 min or more after the acid addition using deaerated or non-
deaerated water were consistent with the certied value by coulometric
titration with electrogenerated Fe(II); however, those by gravimetric
titration at less than 10 min became higher. The apparent assay of potas-
sium dichromate dissolved in deaerated or non-deaerated water by
gravimetric titration at 0 min after the acid addition was more than
100.6%. The indicator current during titration also drifted upward
under the conditions of being short on liberation time. It was also sug-
gested that the positive bias at shorter iodine liberation time resulted
from complex formation between Cr(III) and sodium thiosulfate in-
duced by unreacted potassium dichromate [12] as mentioned above.
On the other hand, potassium iodate oxidized iodide ions in a couple
of seconds; therefore, even shorter iodine liberation time caused no
bias.
3.5. Inuence of background oxidation of iodide ions and vaporization of
liberated iodine
The background oxidation of iodide ions gives positive bias to the
potassium dichromate assay and the vaporization of liberated iodine
gives negative bias to it. Constant-voltage biamperometry was employed
to estimate the background oxidation of iodide ions. Gravimetric titra-
tion with thiosulfate was performed to evaluate the vaporization of
liberated iodine under a stoppered beaker and an unstoppered one
during the libetaion.
The authors previously reported the impacts of the oxidation of
iodide ions in an acidic media [7,15]. In the previous report [7], indicator
Fig. 6. Dependency of the apparent potassium dichromate and potassium iodate assays by
gravimetric titration with thiosulfate on the iodine liberation time in a stoppered beaker.
The condition of 3 g of potassium iodide and 1 mL of 9 mol L1 sulfuric acid was used. :
0.2 g of potassium dichromate dissolved in 100 mL of non-deaerated water. : 0.2 g of po-
tassium dichromate dissolved in 100 mL of deaerated water. : 0.15 g of potassium iodate
(NMIJ CRM 3006-a) dissolved in 100 mL of non-deaerated water [7]. : Certied value
(99.974%) of potassium dichromate by coulometric titration with electrogenerated Fe(II)
(not related to the abscissa axis) [13].
 T. Asakai, A. Hioki / Microchemical Journal 123 (2015) 914
current of constant-voltage biamperometry was monitored for 100 mL
of non-deaerated water containing 1 g to 5 g of potassium iodide and
1 mL to 5 mL of 9 mol L1 sulfuric acid without potassium dichromate
nor potassium iodate in both the presence and the absence of room
light. The current always increased acceleratively with time in a bright
room and proportionally with time in a darkroom. The larger amount
of sulfuric acid (within 1 mL to 5 mL of 9 mol L1 in 100 mL of water)
or surprisingly the smaller amount of potassium iodide (within 2 g to
5 g in 100 mL of water), the larger increasing rate of the current under
each condition of a bright room and a dark one. Less than 2 g of potassi-
um iodide caused smaller increasing rate of the current compared with
2 g of potassium iodide. The cause of the strange dependency of the cur-
rent on iodide ion concentrations remains unclear. The impact of the
air-oxidation to the assay of potassium dichromate under the used con-
dition of acid concentrations was larger than that under the used condi-
tion of iodide ion concentrations; however, the impacts were always
less than 0.01% at 10 min with below 5 g of potassium iodide and
below 2 mL of 9 mol L 1 sulfuric acid in a darkroom, and less than
0.03% at 10 min with below 5 g of potassium iodide and 3 mL to 4 mL
of 9 mol L1 sulfuric acid in a darkroom. In the examination of the de-
pendency of the apparent potassium dichromate assay shown in
Fig. 3, the assays slightly increased under the conditions using 3 g or
more of potassium iodide. The impact of the background oxidation
was less than 0.001% of the potassium dichromate assay under the con-
dition using 5 g of potassium iodide and 1 mL of 9 mol L1 sulfuric acid
in non-deaerated water at 10 min after acid addition in a darkroom;
therefore, the positive bias for higher concentrations of potassium
iodide in Fig. 3 was not caused by the simple background oxidation of
iodide ions without light. It is known that Cr(III) accelerates air-
oxidation of iodide ions [19]. Fig. 7 illustrates some examples of changes
in background currents resulting from the oxidation of iodide ions. The
changes in the indicator currents were monitored under the conditions
of different amounts of 9 mol L1 sulfuric acid and of potassium iodide
in a darkroom. The indicator currents could be converted to the impact
on the assays of potassium dichromate based on the titration curve. The
effects of the oxidation of iodide ions under the conditions with 1 mL to
5 mL of 9 mol L1 sulfuric acid and 1 g to 5 g of potassium iodide were
Fig. 7. The iodine liberation process monitored by constant-voltage biamperometry with-
out of potassium dichromate using a dual platinum-plate electrode (applied constant-
voltage 150 mV), stirring at approximately 400 rpm. Increase of background current due
to air-oxidation of iodide ions was monitored every 1 min after adding 1 mL of
9 mol L1 sulfuric acid following the addition of 3 g or 5 g of potassium iodide to
100 mL of water with/without Cr(III) and Fe(II). Cr(III): chromium(III) potassium sulfate
dodecahydrate, equivalent number of Cr(VI) in 0.2 g potassium dichromate; Fe(II): less
than one-tenth number of the Cr(VI).
13
negligible for the assays of potassium iodate or potassium bromate as
the authors previously mentioned [7,15]. However, the effects under
the presence of Cr(III) were signicant in fact. The indicator currents
signicantly increased for the solutions containing both chromium(III)
potassium sulfate dodecahydrate, equivalent number of Cr(VI) in 0.2 g
potassium dichromate, and iron(II) sulfate, less than one-tenth number
of the Cr(VI), which was used to reduce Cr(VI) of a possible impurity.
Added Fe(II) might have consumed a part of dissolved oxygen, and
even if the Cr(III) compounds contained any impurity oxidants such as
Cr(VI) which could oxidize iodide ions, Fe(II) could have reducing effect
on the impurities. Therefore, it was obviously conrmed that air-
oxidation of iodide ions were accelerated by the Cr(III) compound itself.
If any solid Cr(III) compound is used, no dissolved oxygen is additionally
introduced with Cr(III) and on the other hand there is a possibility of in-
troducing any impurity oxidants. As a matter of fact, the similar results
were obtained by directly using solid chromium(III) potassium sulfate
dodecahydrate. In the case of 3 g of potassium iodide and 1 mL of
9 mol L1 sulfuric acid with Cr(III) in Fig. 7, the impact of generated io-
dine on the assay of 0.2 g of potassium dichromate was over 0.3% at
10 min of the iodine liberation time. Chromium(III) had a role to accel-
erate air-oxidization of iodide ions; consequently, it led inaccurate stan-
dardization of sodium thiosulfate to some degree. Though these positive
impacts were too large, it was concluded that the air-oxidation of iodide
ions accelerated by Cr(III) was the main cause of the bias for 3 g to 5 g of
potassium iodide shown in Fig. 3, because the actual experimental
condition under which Cr(III) was generated with the progress of the
titration was different from the condition of Fig. 3 under which Cr(III)
always existed since the initial point.
Gravimetric titration of potassium dichromate with thiosulfate in an
unstoppered beaker was performed to evaluate the degree of iodine va-
porization during 5 min to 20 min after acid addition under the condi-
tion using 3 g of potassium iodide and 1 mL of 9 mol L1 sulfuric acid.
The inuence of the iodine vaporization to the potassium dichromate
assay was ca. 0.02% per minute in an unstoppered beaker after the
acid addition. Most of the liberated iodine was titrated within 30 s
after removing the stopper for the usual titrations; therefore, the bias
from iodine vaporization loss fell much smaller than half of 0.02%. Ac-
cording to our previous paper [7], the degree of iodine vaporization
under the condition of 3 g of potassium iodide and 1 mL of 9 mol L1
sulfuric acid in 100 mL of non-deaerated water containing 0.15 g of potas-
sium iodate were 0.017% min1 and 0.0006% min1 in an unstoppered
beaker and a stoppered one, respectively. Therefore, degree of iodine va-
porization in a stoppered beaker is negligible; roughly similar degree is
also seen from the results between 10 min and 20 min in Fig. 6.
More acids and more potassium iodide could accelerate the iodine
liberation; consequently, it could allow us to titrate potassium dichro-
mate with a thiosulfate solution at 10 min or earlier after acid addition.
Besides shorter liberation duration could be of advantage to avoid oxi-
dation of potassium iodide. However, more acids and more potassium
iodide increase the background and give the positive error to the
assay of potassium dichromate.
3.6. Estimation of accuracy of standardization with potassium dichromate
The authors developed NMIJ CRM 3006-a potassium iodate accord-
ing to the article reported [7]; the certied value was 99.973%
0.018% (the value following indicates the expanded uncertainty
with the coverage factor 2). The uncertainty value includes all compo-
nents from the homogeneity, the stability, etc. The titration used for po-
tassium iodate was very robust with respect to changes in the amounts
of acid and potassium iodide and also the liraration time.
The accuracy of the standardization of sodium thiosulfate with po-
tassium dichromate was roughly estimated. An appropriate condition
for the standardization with potassium dichromate was assumed as fol-
lows: 3 g of potassium iodide, 1 mL to 2 mL of 9 mol L1 sulfuric acid in
100 mL of deaerated water, and 10 min liberation time. The uncertainty
14 T. Asakai, A. Hioki / Microchemical Journal 123 (2015) 914
was calculated when conditions were changed within 2 g to 4 g of po-
tassium iodide, 1 mL to 2 mL of 9 mol L1 sulfuric acid and 5 min to
15 min of the liberation time in 100 mL of deaerated water. The possible
biases of the assays at most were 0.024% for potassium iodide, 0.005%
for acid and 0.026% for time. The bias of the standardization could be
often up to 0.1% judging from this estimation. In the case of being
short on acid and inadequate deaeration, the bias might exceed even
1%. These were all positive biases from the assays under the appropriate
condition. Compared with potassium iodate, the visual endpoint detec-
tion (e.g., with starch) for potassium dichromate has a small disadvan-
tage of hard color judgment.
4. Conclusions
Potassium dichromate assays through two different paths were
compared: direct coulometric titration with electrogenerated Fe(II)
and gravimetric titration through the iodine liberation reaction with a
sodium thiosulfate solution of which concentration was determined
by coulometric titration with electrogenerated iodine. The accuracy of
the latter gravimetric titration was discussed. When analysts utilize
the reaction to standardize a sodium thiosulfate solution with potassi-
um dichromate, the bias from the accurate concentration of the sodium
thiosulfate solution might exceed even 1% unless an appropriate condi-
tion is chosen. The results of the gravimetric titration under an appro-
priate condition of iodine liberation were in good agreement with
those of the former direct coulometric titration, though allowance of
the appropriate condition for iodine liberation was very narrow. Impu-
rities in reagents, sulfuric acid and potassium iodide, involved in
oxidation-reduction reaction could affect the results of the standardiza-
tion in principle. Analysts could need blank testing to estimate the effect
of reagents used on the standardization. The testing could also give im-
portant information of the inuence of light; the effect on the standard-
ization was negligible because there was no dependency of the assays
on the amount of sulfuric acid and potassium iodide with potassium io-
date in the present study. Though potassium dichromate as a reference
to standardize sodium thiosulfate is not superior to potassium iodate
from the viewpoint of robustness against condition changes, it was con-
rmed in the present study that potassium dichromate was also
reliably available.
References
[1] L.W. Winkler, Die bestimmung des im wasser gelsten sauerstoffes, Ber. Dtsch.
Chem. Ges. 21 (1888) 28432855.
[2] ISO 5813, Water Quality Determination of Dissolved Oxygen Iodometric
Method, International Organization for Standardization, Geneva, 1983.
[3] I.M. Kolthoff, R. Belcher, Volumetric Analysis, Volume III Titration Methods:
Oxidation-Reduction Reactions, Interscience publications, Inc., New York, 1957.
[4] American Chemical Society, American Chemical Society Specications Reagent
Chemicals, 9th ed. Oxford University Press, New York, 2000.
[5] JIS K 8001, General Rule for Test Methods of Reagents, Japanese Industrial Standards
Committee, Tokyo, 2009.
[6] T. Asakai, M. Murayama, Scheme and studies of reference materials for volumetric
analysis in Japan, Accred. Qual. Assur. 13 (2008) 351360.
[7] T. Asakai, A. Hioki, Investigation of iodine liberation process in redox titration of
potassium iodate with sodium thiosulfate, Anal. Chim. Acta 689 (2011) 3438.
[8] C.R. McCrosky, The oxidizing action of potassium dichromate as compared with that
of pure iodine, J. Am. Chem. Soc. 40 (1918) 16621674.
[9] W.C. Vosburgh, Potassium dichromate as a standard in iodimetry and the
determination of chromates by the iodide method, J. Am. Chem. Soc. 44
(1922) 21202130.
[10] W.C. Bray, H.E. Miller, The standardization of thiosulfate solution by the permanga-
nateiodide methods, J. Am. Chem. Soc. 46 (1924) 22042211.
[11] S. Popoff, J.L. Whitman, Standardization of solutions used in iodimetry II, J. Am.
Chem. Soc. 47 (1925) 22592275.
[12] F.L. Hahn, The reactions taking place in the iodimetric determination of chromates, J.
Am. Chem. Soc. 57 (1935) 614616.
[13] T. Asakai, A. Hioki, Comparison of three electrochemical end-point detection
methods to assay potassium dichromate by coulometric titration, Accred. Qual.
Assur. 17 (2012) 4552.
[14] T. Asakai, A. Hioki, Precise coulometric titration of cerium(IV) as an oxidising agent
with electrogenerated iron(II) and reliability in cerium(IV) standardisation with
sodium thiosulfate, Anal. Methods 4 (2012) 34783483.
[15] T. Asakai, A. Hioki, Potassium bromate assay by primary methods through iodine
liberation reaction, Anal. Methods 5 (2013) 62406245.
[16] T. Asakai, M. Murayama, T. Tanaka, Precise coulometric titration of sodium thiosul-
fate and development of potassium iodate as a redox standard, Talanta 73 (2007)
346351.
[17] T. Tanaka, H. Hayashi, Y. Komiya, H. Nabekawa, H. Hayashi, Stability of sodium thio-
sulfate solution and constant-current coulometric iodometric titration of potassium
iodate, Bunseki Kagaku 56 (2007) 327332.
[18] V.P.R. Rao, B.V.S. Sarma, Standardization of thiosulfate using potassium dichromate
by direct titration, Anal. Chem. 37 (1965) 13731375.
[19] A.I. Vogel, Vogel's Textbook of Quantitative Inorganic Analysis, 4th ed. Longman,
London, 1978. 376.