DETERMINATION OF THIAMINE IN PREMIXTURE AND COMPOUND FEED BY LIQUID CHROMATOGRAPHY METHOD

Bull Vet Inst Pulawy 52, 435-440, 2008



DETERMINATION OF THIAMINE IN PREMIXTURE AND
COMPOUND FEED BY LIQUID CHROMATOGRAPHY METHOD

JOLANTA RUBAJ, WALDEMAR KOROL,
GRA?YNA BIELECKA, AND KRZYSZTOF KWIATEK1

National Feed Laboratory, National Research Institute of Animal Production,
20-079 Lublin, Poland
1National Veterinary Research Institute, 24-100 Pulawy, Poland
jrubaj@clpp.lublin.pl

Received for publication July 08, 2008



Abstract
mg/kg dry matter), and 3-4 mg/kg is found in powdered

milk (17). From many years, Polish industrial feed

The analytical procedure of thiamine quantification
production has been based on recipes, which contain
in premixtures and compound feedingstuffs by HPLC method
defined amounts of vitamin B1, according to the
is presented. Thiamine was extracted from feed samples with
requirements of the animal species. A recommended
0.1M hydrochloric acid at 100°C for 30 min. In the case of
thiamine doses to compound feedingstuffs for poultry,
compound feedingstuffs, 10% Taka-diastase solution was
varies from 1-2 mg/kg in case of compound
added to the samples, and then the samples were incubated at
feedingstuffs for laying hens to 3-4 mg/kg for turkeys
37°C for 17 h. Afterwards, thiamine was oxidised to
(16). The main thiamine sources are its synthetic form
thiochrom, using 1% alkaline potassium ferricyanide (III). The
filtrated solution was analysed by HPLC with isocratic flow of
like hydrochloride and nitrate salts, available in the
eluent chloroform-methanol 90:10 (v/v). The measurement
form of preparations and vitamin premixtures. To
was done by a fluorescence detector. This method has been
compound feedingstuffs, thiamine is added as a
applied for the quantification of the total content of thiamine in
premixture for the supplementation of its level,
compound feedingstuffs, as well as added thiamine in the form
according to the animal’s requirements. Thiamine is
of hydrochloride or nitrate salt. The limit of the quantification
characterised by low stability, especially in neutral or
of this method was determined on the level of 1 mg/kg. The
alkali environment. The quantitative losses of thiamine
coefficient variation of thiamine quantification results in
can occur particularly during thermal and pressure
premixtures samples was 3.7%, and in compound feedingstuffs
processing of compound feedingstuffs (6). Therefore,
was 5.56%. The Horrat value for premixture samples reached a
value of 0.60, whereas in the case of compound feedingstuffs it
the possibility of quantification of vitamin B1 as an
was 0.63, thus confirming acceptable precision of the
additive to premixtures or supplementary feed, and of
procedure. The recovery rate for thiamine added in the form of
quantification of total thiamine content in feed, are both
hydrochloride was 102.3% in premixes and 98.9% in
important for monitoring analyses of feedingstuffs in
compound feedingstuffs. The recovery rate for thiamine added
the frames of the official control, as well as internal
in the form of nitrate salt to compound feedingstuffs was
control carried out by feed producers.
similar and reached a value equal to 98.7%.
Polish feed laws do not contain any standard

methodology for quantitative analyses of vitamin B
Key words
1 in
: thiamine, quantification, HPLC,
animal feedingstuffs. Likewise, international (ISO),
premixtures, compound feedingstuffs.
European (EN), or Polish (PN) norms do not contain

any methods for the quantification of this vitamin,

concerning the total amount or as additive in its
Thiamine (vitamin B1) plays main functions in
synthetic form. In the past, recommended official
tissue respiration, participating mainly in carbohydrate
methods were based on fluorometric measurement of
transformation processes. The bio-active form of this
thiochrom, which was received from vitamin B1
vitamin is thiamine pyrophosphate, which is a
oxidation after extraction, enzymatic hydrolysis, and
coenzyme of decarboxylases, transketolases, and
extract purification (2, 15). These methods were
enzymes, which catalyse oxidative decarboxylation of
sensitive to interference with some disturbing
ketoacids. The thiamine deficiency is harmful for
substances. However, the fluorometric method could
animal organisms. In feed materials of plant origin,
not be used in case of the presence of materials, which
vitamin B1 is present in small quantities, mainly in
absorb thiamine or influence the thiochrom
pericarp (3-5 mg/kg) and germs (10-20 mg/kg) of
fluorescence. Taking this into account, the fluorometric
cereal grains, yeasts (40-120 mg/kg), potatoes (5-10
method for vitamin B1 quantification was withdrawn by

---

436
a Commission Directive 98/54/EC from the registry of
Extraction. Because Vitamin B1 is sensitive to
standard methods for animal feedingstuff analysis.
light, all actions during this phase of analysis were
Classical microbiology methods for the quantification
conducted without exposure to daylight (by using amber
of thiamine in food and feed in spite of sufficient
glass flask or flask protected by aluminium foil).
sensitivity are characterised by small selectivity, long
1. Preparation of premixture sample:
execution time, high work consumption, and low
The 5 g of the premixture was weighed in
precision (3, 5).
Erlenmeyer flask, and then 98 ml of 0.1 M hydrochloric
Among instrumental methods, the high
acid was added and the flasks were agitated for 30 min.
performance liquid chromatography (HPLC) method is
Afterwards 2 ml of 50% trichloroacetic acid was added
sufficient as regards to its selectivity and sensitivity for
and that solution was boiled for 10 min in a water bath.
the thiamine quantification in pharmaceuticals, food,
Next, the samples were cooled to room temperature. The
and feedingstuffs. Analytical procedures, elaborated
extracts were then transferred to centrifuge tubes and
some years ago, base on electrochemical and UV
centrifuged for 10 min. If it was necessary, an additional
detection (14, 18) or fluorometric detection of the
dilution was done, in order to fit the concentration of the
vitamin (1, 18, 19). Each of the elaborated and
solution to the calibration curve.
validated procedures contains precisely described
2. Preparation of compound feedingstuff
extraction stage, which allows quantification of the
sample:
total thiamine amount in the analysed sample.
The 5 g of the compound feed was weighed in
Taking these into account, the study was
an Erlenmeyer flask, then 60 ml of 0.1 M hydrochloric
undertaken with the aim of applying the HPLC method
acid was added and the flasks were placed into an
with fluorometric detection for the quantification of
ultrasound bath at 50°C for 30 min. After that, they were
thiamine, which was added to premixture, or
agitated for 30 min, and then boiled for 30 min in a
quantification of total thiamine amount in compound
water bath. Afterwards, they were cooled to room
feedingstuffs. The validation parameters of the method
temperature and the pH was adjusted to 4.3-4.7 by 2.5 M
were evaluated for both premixture and compound
sodium acetate solution. Then, 5 ml of 10% Taka-
feedingstuff sample matrices.
diastase enzyme was added to the solutions, which were

then incubated at 37°C for at least 16 h, but no longer

than 17 h. After that time, the enzymatic reaction was
Material and Methods
stopped by an addition of 2 ml of 50% trichloroacetic

acid. The samples were once more boiled in a water bath
Materials. The study was carried out on
for 10 min, and cooled to the room temperature. The
premixture and compound feedingstuff samples, which
whole amount of the extract was then transferred into a
contained vitamin B
measuring flask, then completed with deionised water to
1. The parameters of this method
were also evaluated on certified reference materials:
100 ml volume, and mixed. The part of the extract was
CRM 121 (Wholemeal flour) and CRM 421 (Milk
then spun in the centrifuge.
powder) (3). The following reagents were used in the
Derivatisation. The 5 ml of the extract was
study: chloroform and methanol (HPLC grade),
transferred into a new flask and 5 ml of water-saturated
hydrochloric acid, trichloroacetic acid, sodium
isobutanol was added to the sample. The whole volume
hydroxide, isobutanol saturated with water, potassium
was agitated for 30 s. Afterwards 3 ml of 1% potassium
ferricyanide (III), and sodium acetate (POCH Gliwice,
ferricyanide (III) in 15% sodium hydroxide was added,
Poland). Enzyme Taka-diastase from Aspergillus oryzae
and the samples were once more agitated for 30 s. Then
was delivered by Fluka. Deionised water was prepared
the samples were left for the phase separation. Then, the
on Mili-Q system (Millipore, France).
upper phase was taken, filtered by a syringe filter, and
Standards and standard solutions. The
20 µl of the solution was dosed on a chromatographic
standard stock solutions of thiamine hydrochloric salt
column.
and thiamine nitrate salt (100 µg/mL in 0.1 M
Evaluation of the procedure. The precision of
hydrochloric acid) were stored at 4°C in darkness up to
this method was checked by the evaluation of the
one month. The working solutions from 0.1 to 1.0
repeatability and reproducibility. In case of precision,
µg/mL were prepared each time just before use.
the Horwitz coefficient called Horrat value (Hor) was
Liquid chromatography - fluorescence
calculated. The Hor is the ratio of the reproducibility of
detection. High pressure liquid chromatograph (Dionex
standard deviation SDr, calculated from the data, to the
P-680) with fluorescence detector (Dionex RF 2000),
target standard deviation ?p,, calculated from the Horwitz
extinction length ?=365 nm and emission wavelength
formula (9) ?p=0.02 C0.8495, where C means
?=435 nm was used for the analysis. The separation was
concentration expressed as denominated mass fraction
done on the column LiChrospher 100 NH
(e.g. 1 mg/kg = 10-6). Acceptable Hor, which describes
2 (250 mm x
4.6 mm x 5 µm), or a similar column. The solution of
the precision of measures, are 0.5<Hor<2 (10, 11). In
chloroform and methanol (90:10, v/v) was a mobile
order to adjust the results to the repeatability conditions,
phase for LC. The isocratic eluent flow was 2 mL/min,
the target standard deviation ?p was multiplied by 0.66
at a controlled temperature of 25ºC for the separation of
(SDr=0.66 SDR) (1). The accuracy of the method was
interfered elements, which could be in the sample
evaluated by the recovery rate calculation and analysis
extract.
of certified reference materials CRM 121 (wholemeal
flour) and CRM 421 (milk powder). The estimated

---

437
expanded uncertainty of the elaborated method was
102.3%. The limit of quantification was 1 mg/kg,
evaluated by within-laboratory approach, according to
combined standard uncertainty was 7.6%, and expanded
the technical report of Eurolab No. 1/2007 (7).
uncertainty was 15.2%.
Because the within-laboratory approach for
The bias of the method was evaluated on the
standard uncertainty (u) estimation was used, it was
basis of the results from CRM analyses and was
necessary to evaluate the within-laboratory expressed as the relative difference between the
reproducibility (s), as a measure of precision, and
achieved result and certified reference value. The bias
systematic bias (b), calculated from the CRM analyses.
was determined in the range of 4.8%-5.1%.
The standard uncertainty was estimated according to

formulas from technical report of Eurolab (7) and the

direction of Nordtest (8).
Discussion


2
2
u =
s
+ b
[1]
In the presented analytical procedure, a high
performance liquid chromatography (HPLC) was

applied. The procedure allowed the separation and
Within-laboratory reproducibility was obtained
quantification of vitamin B
from the range between the repeated analyses of feed
1 as an additive to premixture
or total amount of this vitamin in compound
samples. According to the Nordtest direction (8), for two
feedingstuffs.
individual thiamine analyses in each sample, the mean
The
achieved
coefficients of variation (3.7% in
value, difference between analyses (range), relative
case of premixture, 5.6% in case of compound
difference (%), and mean relative difference (%) for
feedingstuffs), and Horrat values, which amounted to
every samples of the same kind of feed were calculated.
0.60 and 0.63 respectively, confirmed that the applied
The mean range divided by d coefficient (for two
analytical procedure offers acceptable precision.
repetitions d=1.128) was equal to the within-laboratory
According to Horwitz (10, 11), acceptable Horrat values
reproducibility standard deviation.
should range between 0.5 and 2. The Horrat values
The contribution of statistic bias (b) in the
below one provided evidence of a high precision of the
uncertainty measure was calculated from the mean
results. The accuracy of measures was also excellent,
standard deviation of results CRM ?, the uncertainty
and was expressed by the thiamine recovery rate in the
CRM uref, and the standard deviation of analysis CRM
range from 98.7% to 102.3%.
sref, based on the following formula (7, 8):
The precision of results at the level of 3% (Hor

0.47) and recovery rate ranging from 91% to 103% were

obtained by Wollard and Indyk (18) in the analyses of
hydrophilic vitamins in baby-foods (6.7 mg/kg) by
2
2
s
2
b =
? + u
ref
+
[2]
ref
n
liquid chromatography combined with fluorometric
detection. In interlaboratory studies, where the HPLC

method with fluorimetric detection was applied for the
? (bias 2)
determination of total thiamine content in pet foods and
? =
i
[3]
compound feedingstuffs for turkey, the repeatability
n
values were achieved, ranging from 3.3% to 3.9% with

an acceptable Horrat value of 0.89 for pet foods and

0.44 for compound feedingstuffs for turkey (1). The
Results
recovery rate of thiamine hydrochloride in pet-foods

analyses, ranged from 77.0% to 98.8% (1).
The characteristic chromatograms of reference
The obtained chromatographic separation,
thiamine solution (a), premixture extract (b), and
precision parameters, and accuracy defined by the
compound feedingstuffs (c) are presented in Fig. 1. The
recovery rate, illustrated that the applied analytical
retention time was about 3 min. For the applied
procedure was correct. The accuracy of the elaborated
analytical procedure and chromatography conditions of
procedure was additionally confirmed in analyses
the separation, there were not observed any of the
performed on certified reference materials (CRM) and
interference substances.
reached a value of 95.0% (uncertainty 5.0%). This
Table 1 contains the validation parameters
method allows the determination of thiamine, which was
obtained during studies, such as precision of the method,
added to premixtures and compound feedingstuffs in the
which was based on repeatability and expressed as a
form of hydrochloride and nitrate salts, i.e. in the form
coefficient of variation (3.7-5.6%), intermediate
used by feed producers. Because of this, this method can
precision, expressed as within-laboratory reproducibility
be applied for official control, and allows verifying the
(3.6%), and Horrat value (Hor=0.60-0.63), which are the
producers’ declarations with regard to the amount of
objective measure of precision for the applied analytical
thiamine addition to premixtures and compound
procedure. The calibration curve possessed good
feedingstuffs.
linearity in the range of 0.1-1.0 µg/mL with the
In order to verify the applied analytical
correlation coefficient r=0.9999. The recovery rate of
procedure, the levels of thiamine in certified reference
thiamine addition to the analysed samples in the form of
materials were determined. The achieved results are
hydrochloride and nitrate salts ranged from 98.7% to
listed in Table 2.

---

438







25.0 Standard Emission
mV
EM:435 nm

1 - thiamine – 3.052




10.0





min

-5.0
0.00
1.00
2.00
3.00 a) 4.00
5.00
6.00
7.00


a)


25.0 Sample 1
Emission
mV
EM:435 nm

1 - thiamine – 3.138




10.0





-5.0
min

0.00
1.00
2.00
3.00 b) 4.00
5.00
6.00
7.00




70.0 Sample 2
Emission
c)
mV
EM:435 nm




1 - thiamine – 3.106

50.0



25.0

c)


2 – 4.315


-10.0
min

0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00



Fig. 1. Characteristic chromatograms of thiamine: a) chromatogram of standard extract (concentration 0.1449 µg/mL),
b) chromatogram of compound feedingstuff extract (concentration 0.1458 µg/mL), c) chromatogram of premixture
extract (concentration 0.3566 µg/mL).

---

439







Table 1
Validation parameters of the method for thiamine determination



Validation parameters


Limit of quantification
1 mg/kg
Standard calibration curve

r=0.9999


y=23.373x-0.1125
Linearity range, µg/mL

0.1–1
Precision CV, % (n=10)

(Repeatability)

Compound feed
3.12 mg/kg
5.6
Premixture
35.2 mg/kg
3.7
Hor (compound feed)
3.12 mg/kg
0.63
Hor (premixture)
35.2 mg/kg
0.60
Within-laboratory
7.7 – 800
reproducibility CV,% n=11
mg/kg 3.60
Recovery rate - thiamine

hydrochloride, % (n=6)

Compound feed
3.12 mg/kg
98.9
Premixture
35.2 mg/kg
102.3
Recovery rate-thiamine nitrate,
% (n=6)

Compound feed
3.12 mg/kg
98.7
Combined uncertainty u (%)
7.6
Expanded uncertainty

U = 2 x u (%) (k = 2)
15.2


Table 2
The concentration of thiamine in certified reference materials



Certified reference
Declared concentration of
Achieved thiamine
material
thiamine (uncertainty)
analyses results
mg/kg dry matter
mg/kg dry matter
CRM 421
6.51 (±0.48)
6.18
Milk powder
CRM 121
4.63 (±0.39)
4.41
Wholemeal flour

---

440
References
12. Keller H.E.: Analytical Methods for Vitamins and

Carotenoids in Feed. Department of Vitamin Research
1.
Anon.: Analytical Methods Committee. Determination of
and Development Roche, Basle, 1988.
thiamine and riboflavin in pet foods and animal
13. Lebiedzi?ska A., Marsza?? M.L., Kuta J., Szefer P.:
feedingstuffs. Analyst 2000, 125, 353-360.
Reversed-phase high-performance liquid chromatography
2.
Anon.: Commission Directive 73/46 establishing
method with coulometric electrochemical and ultraviolet
Community methods of analysis for the official control
detection for the quantification of vitamins B1 (thiamine),
of feedingstuffs. Annex II. Point 2: Determination of
B6 (pyridoxamine, pyridoxal and pyridoxine) and B12 in
thiamine, OJ L 83, 30/3/1973, pp. 0021-0034.
animal and plant foods. J Chromatography A 2007, 1173,
3.
Ball G.F.M.: Thiamine (vitamin B
71-80.
1) In: Bioavailability
and Analysis of Vitamins in Food. Chapman & Hall, New
14. Marsza?? M.L., Lebiedzi?ska A., Czarnowski W., Szefer
York, 1998 pp. 267-292.
P.: High-performance liquid chromatography method for
4.
Chhabra A., Dudhe A.: Simultaneous determination of
the simultaneous determination of thiamine
ruminal niacin, pyridoxine, thiamine and riboflavin levels
hydrochloride, pyridoxine hydrochloride and
by high performance liquid chromatography. J Animal
cyanocobalamin in pharmaceutical formulations using
Sci 1999, 69, 567-569.
coulometric electrochemical and ultraviolet detection. J
5.
De Vries J.W.: Vitamins Water-Soluble. In: Method of
Chromatography A 2005, 1094, 91-98.
Analysis for Nutrition and Labeling, D.M. Sullivan &
15. Methodenbuch Band III. Die chemische Untersuchung
D.E. Carpenter (Eds), AOAC International, Gaithersburg,
von Futtermitteln. VDLUFA, Verlag, Darmstadt, 1988.
MD, 1993, pp. 131-138.
16. Nutrient Recommendation and Nutritive Value of
6.
Eitenmiller R.R., Lin Ye, Landen W.O., Jr.: Thiamine.
Feedingstuffs for Poultry. Poultry Feeding Standards (in
In: Vitamin Analysis for the Health and Food Sciences.
Polish). Ed. S. Smulikowska and A. Rutkowski. Institute
CRC Press, Boca Raton, FL, 2008, pp. 291-324.
of Physiology and Animal Feeding, Jab?onna, 2005.
7.
Eurolab Technical Report 1/2007. Measurement
17.
Raw Material Compendium. A compilation of
uncertainty revisited: Alternative approaches to
worldwide data sources. Novus Europe S.A., Brussels,
uncertainty evaluation, Eurolab, Paris, 2007.
1996.
8.
Handbook for Calculation of Measurement Uncertainty
18. Woollard D., Indyk H.E.: Rapid determination of
in Environmental Laboratories. Nordtest Project 1589-02.
thiamine, riboflavin, pyridoxine and niacinamide in
2007. www.nordicinnovation.net
infant formulas by liquid chromatography. J AOAC
9.
Horwitz W.: Evaluation of analytical methods used for
International 2002, 85, 945-951.
regulation of food and drugs. Anal Chem 1982, 54, 67A-
19. Vinas P., Lopez-Erroz C., Balsalobre N., Hernandez-
76A.
Cordoba M.J.: Determination of thiamine and its esters in
10. Horwitz W.: The certainty of uncertainty. J AOAC
beers and raw materials used for their manufacture by
International 2003, 86, 109-111.
liquid chromatography with postcolumn derivatization.
11. Horwitz
Agric Food Chem 2003, 52, 3222-3227.
W., Albert R.: The Horwitz Ratio (HorRat): A
useful index of method performance with respect to
precision. J AOAC International 2006, 89, 1095-1109.

---