B. A. Manion et al. / Carbohydrate Research 345 (2010) 2698–2704
2699
should not be influenced by the type of starch tested, that is, it
should quantify similar iodine loads regardless of the starch matrix
to which the iodine was initially bound.
iodine solution added to the starch to 400
l
L. Control iodine stan-
dard samples where no starch was added were also carried
through all sample preparation steps. Table 1 shows the various
amounts of both the commercial and laboratory-prepared iodine
solutions used to prepare the standard curves with the total
amounts of iodine (g) and the hypothetical percentage of iodine
in the starch if 30 mg of starch is used for the sample preparation.
To enable future research, an improved iodine quantification
method has been developed based on the recently reported chemi-
cal alkylation and gas chromatographic (GC) procedure for the mea-
surement of various anions reported by D’Ulivo et al.27 Herein, the
authors utilized alkyloxonium tetrafluoroborates to derivatize inor-
ganic anions in an aqueous medium to first create volatile alkylated
derivatives of the anions, which are then quantified by headspace
GC. D’Ulivo et al. also performed the method specifically for iodine
quantification in iodized salt where 1 M triethyloxonium tetrafluo-
roborate (TEOT) was dissolved in dichloromethane, and a simple sol-
vent extraction of the derivatized ethyl iodide from the organic layer
could be performed. The major benefit of this procedure is in allow-
ing for longer storage of the derivatizing agent (since water con-
sumes the derivatizing agent by reacting to ethanol and ethyl
ether) and thus greater consistency in sample testing over time.
For iodate, these authors first ensured reduction to iodide using so-
The aqueous samples (400
lL) containing iodine in the form of an
2
I :KI mixture, with or without added starch, (typically a 30-mg
starch sample was used), were placed into 2-mL polypropylene
Eppendorf microcentrifuge tubes. For samples containing both
iodine and starch, binding of the iodine to starch was first ensured
by vortexing for 10 s and incubating for a minimum of 10 min.
Rapid binding of iodine to starch was evident by immediate color
formation, which was detectable by eye at added iodine concentra-
tions as low as 0.2% iodine in starch. Samples were then treated
with a 400-
lL addition of 0.5 M NaBH
4
in 0.1 M NaOH to bring
the total solution volume to 800
lL. The tubes were capped and
mixed by a vortex mixer until all starch had returned to its original
white appearance (typically 5 s). The tubes were then uncapped
and left to stand open for 15 min in a fume hood to allow excess
hydrogen gas to evolve and dissipate. The samples were again
capped and let stand for a further 24 h to ensure complete reduc-
tion of all iodine species to iodide and to allow the excess hydrogen
to dissipate. The sample tubes were then opened briefly to vent
any pressure build-up from residual hydrogen gas evolution and
then recapped. They were then centrifuged for 10 min at
13,000 rpm to separate the starch from the reaction mixture. A
4
dium borohydride (NaBH ) such that the resultant iodide would be
alkylated using TEOT with collection into dichloromethane for sub-
sequent GC analysis. Thus, we have adapted and further developed
the procedure by D’Ulivo et al.27 for the determination of total iodine
in both native granular starch and in modified starch. We report a
sensitive and reliable chemical procedure for determining total io-
dine in starch where the total bound iodine content, in the form of
polyiodide complexes, is first driven to iodide in solution by reduc-
tion with sodium borohydride, alkylated with TEOT, separated and
collected from the aqueous milieu into a recoverable dichlorometh-
ane collector phase and then quantified as ethyl iodide by sensitive
GC–MS and GC–FID instruments.
200-lL subsample of the supernatant obtained from above the
centrifuged starch pellet (i.e., if the sample contained starch) was
then pipetted into a new 1.5-mL Eppendorf microcentrifuge tube.
This subsample was then alkylated with 150
lL of 1 M TEOT in
dichloromethane containing 2 L/mL toluene added as the GC
l
2
2
. Materials and methods
internal reference standard (IS). After a 10-min purge time to allow
for the initial derivatization reaction to occur and any gasses to
evolve, the sample tubes were capped for a total of 4 h for the
.1. Materials
derivatization to take place. Finally, a 70
underlying dichloromethane phase was withdrawn and placed into
a 200- L glass microvial insert and placed into the 8-mm neck of a
standard GC vial, ready for injection into the GC instrument. Stan-
dards used to quantify iodine in starch were always prepared with
added starch (typically 30 mg) as opposed to iodine standards that
were prepared without starch in order to achieve most accurate
standard curve slope (see Section 3.3 below).
lL subsample of the
Melojel (Common Corn Starch; CS), Amioca (Waxy Corn Starch;
WCS) and Hylon VII (High Amylose Corn; HAC), Perfectamyl D6
Potato Starch; PS), Eliane 100 (Waxy Potato Starch; WPS), Tapioca
Starch (TS), starch samples were all obtained from National Starch
and Chemical Company (Bridgewater, NJ, USA). Midsol 50 wheat
starch was obtained from MGP Ingredients (Atchinson, KS, USA).
Triethyloxonium tetrafluoroborate was purchased from VWR Sci-
entific Co. (Mississauga, ON, Canada), Dichloromethane, 0.1 M io-
l
(
2
dine–potassium iodide solution (I :KI 1.3%:2.6%), sodium
2
.4. Gas chromatographic analysis
borohydride and sodium hydroxide were purchased from Fisher
Scientific Co. (Ottawa, Canada). Iodine and potassium iodide crys-
tals (99.99% purity) were purchased from Sigma–Aldrich Chemical
Co., (Oakville, ON, Canada).
The analyses were performed using a Thermo Trace Ultra GC
(
Thermo Scientific, Italy). An AS-1 autosampler injected 1 lL of sam-
ple into the GC inlet set at 230 °C with a 1:10 split ratio. Carrier gas
2
.2. Preparation of different I
2
:KI solutions
Table 1
Standard curve for the preparation of commercial and laboratory-prepared iodine–
potassium iodide solutions
In addition to the 0.1 M standard solution of iodine–potassium
iodide (1.3%:2.6% w/v) as purchased from Fisher, three different
:KI solutions were prepared, each with a total of 5% (w/v) total io-
dine The three solutions were a 1:9 (10%:90%), 1:4 (20%:80%), and
:2 (33%:67%) I :KI solutions. Additionally, for testing upper limits
I
2
:KI solution
Total iodine (g)
% Iodine
amount (
l
L)
in starcha
I
2
1.3% I :2.6% KI
commercial solution
2
2
10
0.00007
0.00033
0.00165
0.00330
0.00495
0.2
1.1
5.5
11.0
16.5
1
2
50
of iodine signal linearity, a 2% I
was also prepared.
2
(w/v) and 20% KI (w/v) solution
1
1
00
50
2
% I
prepared
2
:20% KI laboratory
10
25
0.00173
0.00432
0.00864
0.01297
0.01729
5.8
14.4
28.8
43.2
57.6
2
.3. Preparation of standards
5
7
0
5
Standards spanning five different amounts of iodine were pre-
pared by pipetting varying amounts of the I :KI solutions into
0 mg of starch with addition of Milli-Q water to bring the total
100
2
a
3
Percentage of iodine in starch (w/w) for 30-mg starch samples.