I
E

 

 

 

 

THE USE OF THE ANTJMONY
TRICHLOREDE METHOD FOR THE
DETERMINATION OF VITAMIN A IN

FEEDS. MILK AND BUTTER

Thesis for the Degree. of M. S.
MICHIGAN STATE COLLEGE
Neiscn W. Abernethy
1938

 

 

THE USE OF THE ARTE-JOEY TRICHLORIDE LETHOD FOR TIE DETFE-EINATION

OF VIM-{IN 'A III FEEDS, MILK AND BUTTER

THE USE OF m ANTIMONY TRICELCRIDE MHOD FOR TILE BEEF ZIITATION

OF VITAMIN A IN FEEDS, MILK MID BUTTER

Thesis for the Degree of Master of Science

Michigan State College

By
Nelson W. Abernathy

Department of Chemistry
Division of Applied Science

Michigan State College

1938

 

’1".

v

(9.7

I. rt‘rwvv“.

1’

Acknowledgement

The author of this thesis wishes to express his
indebtedness and gratitude to Dr. C. A. Hoppert,
Professor of Chemistry, for his assistance in making

this work possible.

Introduction

Since the year 1912, when HOpkins (l) called attention to the
fact that normal growth does not occur when purified food rations
are used as feed, the study of vitamin.A has been very important.
Much has been learned since that time regarding the nature of the
vitamin and the methods of assay.

Steenbock (2) has contributed a great deal to the Chemistry of
vitamin.A, and the apparent relation between the vitamin.A potency
of plant and animal substances and the yellow pigment carotene.

In the year 1926, Carr and Price (3) obtained a blue color when
vitamin.A was treated with antimony trichloride. The same color
was obtained when carotene was used in place of the vitamin. This
suggested a possible relation between the two. Euler has proven
that carotene is very effective in supplementing diets deficient
in vitamin.A. Recently it has been quite definitely established
that carotene is the precursor of vitamin.A. It is believed that
the conversion of the pigment into the vitamin takes place in the
liver and that any excess carotene may be stored in this organ
until needed. The following shows the possible structural relation
between beta carotene and vitamin.A. Only the beta form of
carotene is given below, as it is now used as a basis for the

international vitamin A unit.

2.

Beta Carotene (0,403.56)

3 c /cs
3 3 on on
a! an as
HZC‘J/ {Waaaaaa c.¢ dos
3
sec c/ coca3
32

Vitamin A (0203300)

It was pointed out above that a definite color is obtained when
vitamin.A is treated with antimony trichloride. While this reaction
is not specific for the vitamin it does serve as a means of chemical
assay. The intensity of the blue color is, in general. directly
proportional to the concentration of the vitamin. Even though this
reaction may be given by other substances, the latter are usually
not present in materials which are practical sources of vitamin‘A
so that the reaction is useful and convenient for the chemical deter-
mination of this vitamin. Wokes and Willimott (M) have suggested
the use of a saturated solution of antimony trichloride in the test
as this strength solution has been found to give the best results.
Weaker solutions have been used but seem to give less accurate results.
Norris and Church (6) have studied the effect of temperature upon
the saturation point of antimony trichloride in chloroform and have
shown that the concentration has a definite influence on the color
reaction with vitamin A. They also studied the effect of light upon
this reagent and found that strong light caused the reagent to
deteriorate. Norris and Danielson (5) claim the color test checks
within reasonable limits with the biological method of assay.
Besides the lack of specificity of the reaction there are several
other shortcomings. The blue compound is very unstable so that the
color fades within a very short period of time after it has been
formed. Consequently the color intensity must be measured very soon
after the reagent has been added to the solution to be tested.
Moisture also interferes with the test causing cloudiness which
makes accurate measurements impossible.

For a number of years cod liver oil has been used as a source

of vitamin.A. When the oil is stored under favorable conditions the
vitamin.A potency of the oil may be retained for a long time.
However, when the oil is mixed with feeds the destruction of vitamin
A proceeds at a more or less rapid rate depending on the material
with which the oil is associated, as well as on the storage tempera-
ture and accessibility of air. Consequently, it is necessary that
a determination of the vitamin A content of feed mixtures be made.
For the determination of vitamin.A two methods of assay have been

' used, the chemical and biological. In as much as the biological
method is costly and requires considerable time, it is being gradually
replaced by the chemical method wherever the latter is applicable.

Accordingly, the present study concerns itself with the use of
the CarréPrice reaction in the determination of the vitamin.A content
of various materials to which cod liver oil had been added, as well
as in the determination of the vitamin A potency of milk and butter.
The work also involved a study of the presence, in various foodstuffs,
of substances which might interfere with the CarréPrice reaction.

For the measurement of the intensity of the blue color imparted
by the treatment of vitamin.A with antimony trichloride, use was made
of the Lovibond tintometer, an instrument used internationally to
measure the intensity of colors of various substances. The readings
taken with this instrument. when determining the concentration of
vitamin A, are expressed in terms of Lovibond blue units.

Procedure

A known quantity of cod liver oil, which had been previously

standardized, was added to various feeds and mixed thoroughly.

Various methods of extracting the oil from the feeds were tried.

The direct method of extraction gave nearly quantitative results in
the case of cornmeal, but when oatmeal was used, considerable difficulty
was encountered and another method had to be tried. Use was then made
of the Soxhlet extraction apparatus, which proved to be more satisfac-
tory than the former and was therefore resorted to for the extractions.
with this apparatus it was found that approximately twenty extractions
were made per hour, thereby enhancing the efficiency of the extraction
and permitting the materials to be handled with less difficulty. Upon
completion of extraction, 25 ml of water, 30 ml of ethyl alcohol, and
MO ml of 20% potassium hydroxide were added to the extract. The
mixtures were then heated on a steam bath until saponification was
completed. then transferred to a separatory funnel and 15 ml of ethyl
alcohol added. Two extractions were made using 100-125 ml portions of
ethyl ether. The ether extracts were then washed free from alkali
with distilled water (phenolphthalein being used as an indicator),
transferred to a flask containing 10 gm of anhydrous sodium sulfate
and shaken vigorously to remove the water present. After decanting
the solution and washing the sodium sulfate residue with anhydrous
ether, the ether was removed by placing the flask in a bath at about
50° C and attaching it to a suction pump. The residue was taken up
in chloroform, the volume of the latter depending on the amount of
the vitamin.A expected to be present. For the test, .2 ml of the
chloroform solution was placed in the testing cell and 2 ml of the
antimony trichloride reagent added. Readings were taken as soon as
possible after the development of the color.

In the case of milk and butter no preliminary extraction with
ether was made, these foods being directly saponified and the ether

extractions then made as previously described.

Data

In order to determine the applicability of the CarréPrice reaction
and the possible presence of interferring materials, mixtures of cod
liver oil of known potency with cornmeal, oatmeal, linseed meal and
alfalfa meal were prepared on the basis of 1 ml of oil to 99 gm of the
meals and subjected to study.

To determine the potency of the cod liver oil used in these
experiments, 1 ml of the oil was placed in a 50 ml volumetric flask
and chloroform added to bring it up to volume. Several trials were
made to determine the concentration of vitamin A in the solution, the
average reading being 1.9 blue units. To show the effect of saponi-
fication upon the oil a similar sample was saponified according to the
directions given in the procedure. In this case, several trials gave
an average reading of 1.7 blue units, which is Just slightly more than
a 10% difference between the two readings. It might be, therefore,
expected that some of the vitamin would likewise be lost when the cod
liver oil-meal mixtures were saponified in a similar manner.

In the case of the cornmeal mixture, 30 gm samples were placed in
extraction thimbles and extracted in a Soxhlet apparatus for periods
of 2, h, and 6 hours, to determine the time necessary for maximal
recovery of the vitamin. The material extracted was saponified and
further treated as previously described, the final unsaponifiable

residue being taken up in 5 ml of chloroform.

 

 

Table l
Tample} Period of Blue units ‘Yellow units B.U;7gm Expected
extraction of oil added potency in
B.U./ gm of
oil added
1 2 hrs 14.3 1.1 376 1038
2 2 hrs M1 1.2 896 1038
3 h hrs ML 1.2 896 1038
h 11 hrs 14.5 1.2 901 1033
5 6 hrs 1+.5 1.1 901 1038
6 6 hrs an 1.1 896 1038

 

 

 

 

 

 

 

 

From the data given in Table 1 it may be seen that a 2 hr period
of extraction was sufficient fer maximal recovery of the vitamin‘L.
The lovibond blue unit values obtained, represent approximately 90%
of the expected recovery.

Reference should be made at this point to the inclusion of the
yellow values given in all of the tables. It was necessary, in
matching the color given by the samples with the standard colors of
the tintometer, to introduce a certain amount of yellowu These
yellow values may be of some future interest but no attempt is made
to explain their significance.

.An oatmeal mixture was prepared as in the case of the cornmeal.
.All samples were taken up in 5 ml of chloroform. The following data

were obtained:

 

 

Table 11
bemple} Period of Blue units 'Yellow units B. U./ gm Expected
extraction of oil added potency i
3.11. /gm of
oil added
1 2 hrs 3.6 1.3 733 1038
2 2 hrs 3.5 1.2 712 1038
3 h hrs 3.7 1.2 753 1038
h h hrs 3.5 1.2 712 1038

 

 

 

 

 

 

 

 

The above data indicate a failure to account for an appreciable

amount of the vitamin.L added but that a two hour period of extraction

was sufficient to remove whatever vitamin could be dissolved by the

ether.

It was observed that in the extraction of the oatmeal-cod liver

oil mixture a considerable amount of finely divided material appeared

in the ether.

There seemed to be a direct relationship between the

decrease in vitamin.i recovery and the amount of the fine material

appearing in the ether extract.

However, no attempt was made to

determine the manner of the interference.

Linseed meal was next treated in a similar way, maximal recovery

being obtained with a two hour extraction period as shown in Table 111.

,As in the case of the cornmeal, approximately 90% of the expected

potency was accounted for.

 

 

 

 

 

 

 

Table 111
famplefi Period of Blue units ‘Yellow units B. Ut7figm Expected

extraction of oil added potency in
B.U./gm of
oil added

1 2 hrs h.6 2.6 937 1038

2 2 hrs h.5 2.5 901 1038

3 2 hrs u.5 2.6 901 1038

 

 

 

It is known that alfalfa contains a considerable quantity of
the pigment carotene, which gives the CarréPrice reaction as in the
case of vitamin.A. Consequently, an initial test sample was prepared
in order to determine the blue unit value that would be obtained
without the addition of cod liver oil. .A 20 gm sample was placed in
each extraction thimble and treated as previously described, the final
residue being taken up in 10 ml of chloroform. The results are given

in Table IV.

 

 

Table IV
fiample# Period of Blue units ‘Yellow units Blue units7gm
Extraction of material
1 2 hrs 2.8 2.0 16
2 2 hrs 2.9 2. 0 16.1;

 

 

 

 

 

 

 

.An alfalfa mixture was now prepared as in the case of oatmeal.
The bulkiness of this material necessitated the use of 20 gm samples
and the presence of carotene made the dilution of the final residue
with chloroform differ from the previous cases. 20 m1 of chloroform

was added. Table V shows the results obtained.

 

 

Table V
Eamp1e# Period of Blue units ‘Yellow units B. U./ gm Expected

extraction of oil added potency in
B;U./gm of
oil added

1 2 hrs 2.1 1.3 788 1038

2 2 hrs 2.2 1.3 911 1038

3 2 hrs 2.2 1.3 911 1038

 

 

 

 

 

 

 

 

10.

From the above data it may have been observed that the recovery
of the vitamin from the alfalfa compares favorably with the results
obtained with other feeds. This mixture, however, presented an addi-
tional step which involved a preliminary determination of blue units
given by the alfalfa previous to the addition of the cod liver oil.
The average of the blue units per gram of oil added is 870.

The Determination of Vitamin.A in Milk and Butter

Milk and butter are valued, among other things, for their vitamin
A content. The vitamin A potency is, however, quite variable depending
0n the supply of provitamin A in the ration of the dairy cows. {oreover,
in recent years a practice has developed in which a cod liver concentrate
has been added, chiefly with the intention of fortifying the milk with
vitamin D. Attention is, however, now being called also to the increased
vitamin A content resulting from the incorporation of the concentrate
with milk. Consequently, it was considered of interest and importance
to apply a chemical method to the determination of vitamin A in both
milk and butter.

As a preliminary measure, several determinations were carried out
on ordinary market milk, using 200 ml samples for saponification. A
known amount of cod liver oil was added to a similar source of milk to
determine the extent of recovery of vitamin A. Approximately the
expected recovery was made and in addition, the natural occurring
vitamin of the milk could be accounted for. In addition, several
samples of market milk fortified with Vitex, a cod liver concentrate,
were analyzed for vitamin.A. In the case of the latter type of milk,
100 ml was found suitable for analysis. The vitamin.A content of

butter was determined by saponifying 5 and 10 gm samples, the remainder

ll.

of the procedure being similar to that used in all of the previous

tests. The results of these analyses are summarized in the following

 

 

 

table.
Table VI
Sample # Blue units Yellow units Blue units/ml

1 2.9 1.9 .80
Market 2 3.0 2.0 .83
Milk

3 2.9 2.0 .80

1 3.3 1.5 1.81

2 3.5 1.5 1.92

3 3.1 1.2 1.70
Vitex
Milk h 3.0 1.1 1.65

5 3.3 1.1 1.81

6 3.2 1.1 1.76

7 3.3 1.2 1.81

1 2.5 .8 27.5 B.U./gm

2 2.6 .7 28.6 ..

3 2.1.L .7 26.11 n

h 25 .7 27.5 N
Butter 5 2.5 .7 27.5 t

6 2.6 7 28.6 n

7 2.11 .7 26A .,

8 2.3 .5 25-3 "

 

12

The samples of market milk showed the presence of considerable
natural occurring vitamin. The residues from the 200 ml samples were
diluted with 5 ml of chloroform as were the samples of Vitex milk.
From the data it is observed that the market milk shows the presence
of .81 blue units per ml of milk, while the Vitex samples show an
average reading of 1.78 blue units per ml of milk. The addition of
Vitex to the milk therefore is responsible for .98 units per ml, which
represents about 936 blue units per quart of milk. If this number is
multiplied by 1%, which is the conversion factor usually used when
converting blue units into international units, it is found that INCH
international units of vitamin.A are present in each quart of milk
due to addition of Vitex.

The butter samples of 5 and 10 gm were treated according to the
procedure, the final residues being taken up in 5 and 10 ml of chloroform
respectively. The vitamin A potency found in these samples is an
approximate agreement with values generally given for butter as deter-

mined by the biological method.

13.

Discussion

Altho the method employing the CarréPrice reaction for the
determination of vitamin.A is not yet officially recognized, it is
becoming more and more popular because of the obvious advantages in
time and cost as compared with the biological method of assay.
Various factors such as non-specificity of the reagent, loss of
vitamin.A incidental to the various steps employed in carrying out
the determination and a fading color, account for a relatively low
degree of accuracy. Nevertheless, the method in question does give
results that compare favorably with those obtained by the biological
method and it may therefore be expected that the former will replace
the latter in many cases.

In general, it may be said that the results which were obtained
in this study indicate that the method used is fairly reliable in
determining the vitamin A that has been added to feeds through the
use of cod liver oil. There is, however, a loss of the vitamin encoun-
tered, other than that explained by saponification. This additional
loss was found in most cases. The fact that foodstuffs may contain
some interfering materials may account at least in part for this
additional loss. The expected recovery of the vitamin in each case
was approximately 90% of the original amount added to the feeds. Of
this 907;, there was another loss of 10% in the majority of the feeds
tested. In the case of oatmeal, the recovery was considerably less
than expected, although no emplanation can be given.

In analyzing milk it was found that the percentage of added
vitamin.A, which could be accounted for was somewhat greater than in
the case of the other foodstuffs. From this fact, it would appear

that the use of the CarréPrice reaction in the determination of vitamin

1h.

A in milk, might become very useful in commercial fields as the demand
arises for the testing of milk for this vitamin. When definite claims
are made regarding the vitamin content of foods, it then becomes neces-
sary to make a check on such materials. From the milk samples, to
which Vitex had been added, a very high degree of recovery of the
vitamin was obtained, the data showing this to be about 95%, using the
initial saponified sample of cod liver oil as a basis.

As an indication of the reliability of the chemical method employed
in this study the following information may be cited. The distributors
of Vitex make the claim that when this product is used to fortify milk
with vitamin D, it also adds 1600 international units of vitamin A.per
quart of milk, as determined by the biological method. In making a
chemical assay of this milk it was found that by multiplying the number
of blue units per quart of milk by the factor 1.5 that approximately
IMOO international units could be accounted for. In so far as the
relation between blue units and international units is not definitely
established the factor 1.5 might cause some error in the conversion.
Since the biological method is costly and time consuming, there is
reason to believe that the chemical method will supplant the former in
such work as indicated above.

.Conclusions

The CarréPrioe reaction may be used with a fair degree of accuracy
to determine the vitamin.A in foods that have been fortified by the
addition of cod liver oil.

The degree of recovery of vitamin.A by the method used is relatively
high, the range generally lying between 80-95 .

This method is also applicable in the determination of vitamin.A, in

plain and fortified milk as well as in butter.

15.

Bibliography

Physiol, J, he, 173 (1913)

Steenbock, H. , and Boutwell, P.W. ,
J. Biol, Chem; M1, 81 (1920)

Carr, F.H., and Price E. A.,
Biochem. J. 20, M97 (1926)

Wokes, F., and Willimott, S.G.,
Analyst, 52, 515 (1927)

Norris, E. R., and Danielson, I, 5,,
J. Biochem,, 83, 1469 (1939)

Norris, E. R., and Church, A. E..
J. Biol. Chem., 89, he (1930)

 

 

III I!
‘I‘n‘l’ll‘
ll. .1 l4}.l.l

I. .4." ‘lll‘li‘ ‘\ .V'ill!

 

-§.|.

 

 

M-..i 4-71

A146 l18093
Abernathy

. *' I? ,7.
A146 118093

Abernathy

The use of the antimony tri-
chloride method for the
n of vitamin

\U