On Errors in the Iodometry of Bang’s Method for the Determination of Lipoids in the Blood.’

BY H. Heckscher and 0. Krarup M6ller.

(Communication from Medical Section I1 and the Dispensary at the Copenhagen Municipal Hospital).

In 8 work on some comparative investigations into the micro- methods of Bang and of Bing and Heckscher for the determination of the neutrul fut and free cholesterine in the blood, that is to say the measurement of the “p-fraction” or “primary ether extract” of the blood, one of us [Heckscher (l)] has stated as his experience that in blanks carried out with Bang’s method he has not succeeded in making free the calculated quantity of iodine; in all the blanks the titrated quantity of iodine was less than it should have been according t o the calculations.

Bang’s method is based.upon an oxydation of the lipoids by means of potassium dichromate and sulphuric acid and a subsequent iodo- metrical determination of the quantity of dichromate remaining. The details of the practising of the method are as follows:

The extract, of which the extraction medium has been evaporated, is boiled for 30-45 seconds with lccm. 1% NaOH and then cooled. 1-OOO ccm. n/10 potassium dichromate solution and 5 ccm. cone. sulphuric acid are added, the test-tube then being shaken vigorously. (By this means is developed a fairly strong heat, which is necessary for the oxydation of the lipoids.) After not less than 15 minutes the contents of the test-tube are rinsed with 100-150 (in our experiments always exactly 100) ccm. of tap-water into a beaker or an Er lenmeyer flask; 2 ccm. 5% potassium iodide solution are added, the fluid is shaken

l) Der Redaktion am 26. April 1927 zugegangen.

ON ERRORS IN THE IODOMETRY OF BANGS METHOD Erc. 223

a little, and after not less than 30 seconds titrated with n/lO sodium thiosulphate solution with starch solution as an indicator.

The blanks referred to were carried out in this manner: 1 ccm. of NaOH solution, measured into a test-tube, was boiled for

ll/z minutes in a water-bath and then cooled, after which there was added an accurately measured quantity of n/lO potassium dichromate solution (1/4-1 ccni.) as well as 5 ccm. concentrated sulphuric acid; the process was then carried on in accordance with the particulars given above. In those blanks into which, as in the case of lipoid analydes proper, 1-OOO ccm. dichromate solution were added, on the average 0.885 ccni. n/lO sod-thiosulphate solution was used for titrating, which, according t o equivalence, means that 0.115 ccm. (i. e. 11.50/,) n/10 potassium- dichromate solution was lost.

A loss of this dimension is unfortunate, as in lipoid analyses it will in certain cases be almost as great as the dichromate consumption of the lipoids, 0.116 ccm. n/lO dichromate solution being almost corre- sponding to 0.046 mg. of the lipoid content of the blood; the size of the fraction in question varies under normal conditions between about 0.06 and 0.12mg. (O/,) per 0.1 ccm. of blood [Bing and Heckscher (591. Especially is it unfortunate if the value varies appreciably in size, as under certain circumstances this may greatly compromise the accuracy of the method. In the tests by Heckscher (l), however, this loss of dichromate was of constant size and; by taking this into consideration when calculating the results of the analyses, the author succeeded in arriving at adequate results which were confirmed by means of analyses of chemically pure lipoids.

It has therefore seemed t o us to be of importance to obtain further enlightenment as to the causes of the appearance of this loss of di- chromate, so much the more as none of the investigators who have previously worked with the method have published anything that might serve as an explanation. Bang (3) states without comment that the blank value usually varies between 0.06 and 0.10 ccm. dichromate solution. Blix(4) states that the quantity of water used for diluting “consumes” 0.03-0-04 ccm. dichromate solution, but otherwise does not anywhere give any indication as t o the actual blind values. The other workers, Maas(5) and Szent-Gyorgyi(6) do not make any direct statement about this question.

The following experiments have served to enlighten us as to the causes of this loss of dichromate; through them we have reached a partial

224 H. HECKSCHER ASD 0. I<R.\ltCl~ JfOLLER

0.935 1 1.030 0.940 1 1,020 0.945 1-025 0.940 1.015

understanding of these rather coniplicated niatters. \Ye now publish oiir investigations liecause we believe that, even if they are not complete, they will he of practical significance to those who choose t o ~ o r k with Bang‘s method.

That the loss of dichromate in our experiments have not been the result of the presence of organic substances in the tubes is suffiently shown by the constant of the values in each experiment. The glasses nsed were, liy the way, always cleansed in the manner prescribed by Blix with dichromate and sulphuric acid.

The pipette and microburette for measuring the dichromate and thiosulphate solutions were accurately calibrated by weighing with quicksilver.

The reagents were stored 011 a shelf side by side, that is so say at the same temperature.

In all experiments the thiosulphate and dichromate solutions were exactly n/lO, adjusted in the usual manner t o iodate and thiosulphate respectively.

All titrations in each experiment were made on one and the same day and with the same reagents.

Experinient 1: On the influence of the quantity of water used for dilution.

A: To 1 ccm. tap-water in a test-tube was added 1 .OOO ccm. dichro- mate solution and 5 ccm. conc. sulphuric acid, the mixture being allowed to stand 15 minutes; then dilution with tap-water and titration as described.

B: To 25-200ccm. tap-water in a flask was added 1.000ccm. dichromate solution and 5 ccm. concentrated sulphuric acid; after 15 minutes titration as usually.

Sulphuric acid: Merck’s pro analysi I. The figures indicate the ccm. of n/10 thiosulphate solution consumed.

1.010 1.000 1.015 1.000 1.010 1.000 1.005 0.995

Experiment 1 .

200 ccm. water

0.990 0.985 0.985 0.980 0.980

0.984

ON ERRORS IN THE IODOMETRY OF BARTG'S METHOD ETC. 225

1.

9 -.

3.

4.

Froni this it will be seen: that even great variations in the quantity of water, when this exceeds 50 ccni., are of little significance; yet it must be regarded as being advisable to use 100 ccm. water, as by this means the calculated equivalence was found; the loss of dichromate must be caused by reductions which proceed prior t o the time when the dirhromate-sulphuric acid mixture is diluted ; after these experiments we can not join Blix in his regarding the water as the cause of the loss of dichromate.

f i z p e r i ~ j i ~ 7 t t 2: On the influence of NaOH upon the processes which

A: 1 ccm. lo/, NaOH in a test-tube, boiled in a water-bath 11/2 minutes and cooled.

B: 1 ccm. lo/, NaOH in a test-tube, not boiled. C': 1 ccm. tap-water in a test-tube, not boiled.

are the cause of the loss of dichromate.

Thereafter added to A, B and C 1.000 ccm. n/10 dichromate sol. and 5 ccm. conc. sulphuric acid. After standing for 15 minutes the con- tents of the tubes were rinsed by the help of 100ccm. tap-water into flasks and titrated as usual. The figures indicate the ccm. of n/10 thiosulphate solution used by the titration.

Ezperiment 2. Sulphuric acid: ordinary, pure, colourless (pharm. dan.).

A. I B. I C.

0.890 0.880 0.895 0.890 1 @I: 1 0.880 0-875 0.900 0.900 0.895 I 0.880

Sverage: 0.889 I 0.894 I 0.889

This shows that NaOH has no effect upon the processes in question. Therefore NaOH was never used in the subsequent experiments, but instead 1 ccm. tap-water.

Furthermore, after this it is clear that the loss of dichromate must be due to transformations between the dichromate and the sulphuric acid, transformations which, according to Experiment 1, only take place when the quantity of water present is less than a certain size, i. e., when the ratio of water to sulphuric acid is below a certain value. The reasons for this are examined in the following experiments.

Before we report on these we may insert the statement that different pure potassium diehromate preparations: two separate supplies

226 €1 HECKSCHER AXD 0. KRARUP MOLLER

by Kahlbauni pro analysi) and a thrice recrystallized preparatioii kindly given 11s by Professor Yald. Henr iques , 11. l)., gave coii- cording results. In all the experiments reported on below one of Kahlbauni’s preparations (pro analysi) was used. (This is of cou1‘zc not saying that one may not have occasion to use dichroniatc preparations of varying qualities.)

The sulphwic acid remains to be referred to. Experiments 1 and 2 (series A and A) have already shown that different sulphuric acid preparations give varying results, and this is fully confirnied by the following experiments.

Expe&e?zt 3: On the dependence of the loss of dichromate upon the sulphuric acid preparation.

A: To 1 ccm. tap-water in a test-tube was added 1.000 ccm. n! 10 potassium dichromate solution and then 5 can . conc. sulphuric acid. The tube was shaken vigorously for a moment. At the end of 15 minutes the mixture was rinsed with 100 ccm. tap-water into a flask and titrated as usual.

B: To 100 tap-water in a flask was added 1.000ccm. n/10 potassium dichromate solution and 5 ccm. conc. sulphuric acid. After standing 15 minutes, titrated as usual.

The figures indicate the ccm. of n/10 thiosulphate solution used for the titration.

(Table on next page.)

The two Rlerck’s sulphuric acids used were one and the same preparation from the manufacturer but supplied in two portions by the dealer. There is, however, reason for supposing that the second portion, marked “Mercks 11” has in some manner become contaminated (by an organic impurity). The two specially made preparations were given us by Mr. E s t r u p , D. Ph. of the University Chemical Laboratory, where they were made for our use.

From this experiment it will be seen that the reduction of dichromate by the Bang method, i. e. in series A, varied between 0.051 ccm. (5.lo/J and 0-129ccm. (12-90/,) all according to the sulphuric acid employed. In no case, despite the use of the purest obtainable sulphuric acid preparations, did we succeed in avoiding loss of dichromate.

Besides what has already been stated, it will be seen that the difference between the sulphuric acids can be noted too - although to a lesser degree - in series B, where three of the acids employed (those which resulted in the heaviest loss in series A), evidently procured a loss of dichromate.

ON ERRORS IN THE IODOMETRY OF BANG’S METHOD ETC. 227

Experinaent 3.

Sulph. acid preparation

Ord. pure (pharni. dan.).

Merck’s pro analysi I

Merck’s pro analysi I1

K a h 1 b a u m’s pro analysi

Specially made; about 50 ccm. distilled over

K,CrO,

Specially made; about 50 ccm. distilled over

CrO,

A

0.895 0.880 0.900 0.880 0.910 0.890 0.885 0.895 0.905

0.950 0.950

0.893

0.955 I 0.935 0.940 0’945 0.945 0.940 1 0.945

0-880 0.890 I 0.910 0.900 0,896

0.895 0.895

0.850 0.880 I 0.875

0-870 0.860 0.890 J 0.865

0.940 0.950

0.940

0.945 0 * 940

0.955 0.955

B

0.960 0.955 0.960 0.970 0.960 0.965 0.960 0.960 0.965 0.960

0.962

1.000

1.000 .0.999 0.995 1 1.000

1.000 1

0.990 1

0.980 0.975

0.965 0.965 I 0.980 0.970 1 0.970 0.970

0.960 0.975 0.975

0.995 l.m) 0.997

1.010 0.995

1.000

We do not know with certainty whether the reduction of dichro- mate in question is solely due to more or less “accidental” impurities in the sulphuric acid or whether they are, at any rate partly, “ordinary” transformations between pure dichromate in aqueous solution and con-

centrated sulphuric acid, which develop heat when mixed together; our continued experiments, however, seem to point mostly in the direction first indicated. These subsequent experiments deal with the dependence of the processes upon the quantity of dichroniatc present and upon the ratio of aqueous solution and conc. sulphuric acid. The first of these aspects has previously been dealt with by Heckscher (l), who calculated a correction-graph for Bang’s method on this basis; these earlier experiments are in positive conformity with ours.

Ezperitrtent 4: On the influence of the quantity of K,Cr,O,. In this experiment the volumes of aqueous solution and coiic. sulphuric acid have been kept constant, whilst the quantity of dichromate has varied.

Sulph. acid: Merck’s pro analysi I1 (see Exper. 3). A: 2.000 ccm. n/10 dichromate. B: 1 ccm. tap-water + 1.000ecm. dichromate. C: 1.5ccm. ,) + 0.500 ccm. ,, D: 1.75ccm. ,, + 0-250ccm ,> E: 1.85ccm. ) ) + 0.150ccm. > 3

A-E: To this dichromate in aqueous solution in a test-tube was added 5 ccm. conc. sulphuric acid and then continued as usual.

(Table on next page.)

This shows that the positive size of the loss diminishes with decreasing quantities of dichromate, although not proportionately to the quantity of dichromate, the percentage of loss increasing with decreasing quantities of dichromate. The same was the case in Heckscher’s earlier experiments, in which an ordinary, pure sulphuric acid (pharm. dan) was employed. Yet these experiments were performed with the sulphuric acid preparations that proved t o give the greatest loss. The following experiment was made with sulphuric acids (Merck’s 111) which gave a smaller loss.

Experiment 5: Principally the same as Experiment 4. A, B, C and D quite the same as in Experiment 4. I?: 100ccm. tap-water in a flask, to which was added 1.000ccm.

dichromate solution as well as 5 ccm. conc. sulphuric acid; thus F corre- sponds to series B in Experiment 3.

Sulphuric aoid: Merck’s pro analysi 111. (Table on page 230.)

As in the previous experiment there was here observed: a loss of dichromate, decreasing with the quantity of dichromate added. On the other hand, no definite increase was observed in the percentage of

ON ERRORS IX THE IODOMETRY OF BAKG'S METHOD ETC. 229

Experiment 4.

Consumed ccm. thiosulphate

solution

1.820 1.840 1.830 1 .845 1.825 1.816

0.870 0.866 0.870 0.890 0.875 0.870

0.415 0.415 0.410 0.410 0.415

0.190 0.190 0.190 0.180 0.190 0.190

0.090 0.090 0.090 0.085 0.095 0.090

1.820

0.873

0.413

0.188

0.090

Loss of dichromate

ccm.

.

0.171

0.127

0.087

0.062

0.060

Percentage of loss of dichromate -

Present experiment

- -

8 . 6

12.7

17.4

24.8

40.0

He c ks c h e r's ealier

experiment ~

11.5

15.8

23.2

loss. It must, however, be remarked that here we are working 011 the limit of the obtainable accuracy in the titremetric determination.

We do not know with certainty how these indications are to be interpreted. It may be that the explaaation is, that both the sulphuric acid and the dichromate contain impurities. Therefore, where we are dealing with a comparatively pure sulphuric acid (as here in Experi- ment 5) it may be, at any rate for the most part, an impurity of the dichromate that is the cause of the loss (the loss is fairly proportionate t o the quantity of dichromate added). With a more impure sulphuric

230

- _ _

.A

B

C

D

Loss of dichromate

ccm.

H. HECKSCHER AKD 0. KRARUP MOLLER:

Experi?,lent 5.

Percentage of loss of

dichromate

Consumed ccm. thiosulphate

solution - ..

1.950 1.940

1.950

0.950

0.965 0.960

0.475

0.470 0.460

0.230

0.240 0.245

1.005

0.995

0.050 1 2.5

0.044 1 4 .4 I

0.039 ’ 7 . 8

0.013 1 5 . 2

“0”

“O” 1 acid (Experiment 4 and Heckscher’s earlier ones) we have, on the other hand, two causes of loss, and this may possibly explain the observations made.

Yet the following experiment shows that other factors than the quantity of dichromate and the quantity of the sulphuric acid also have an effect upon the.processes. In this experiment the quantity of dichromate has been kept constant, whereas the volume of water and the quantity of sulphuric acid were varied.

and sulphuric acid. Experinzent 6: On the influence of variations in the quantities of water

A: B: 1 ccm. ,, + 1.000ccm. ,, + 5ccm. 1,

C: Occm. ,, + 1.000ccm. ), + 10ccm. 7,

D: lccm. ,, +1*000ccm. ,, f10ccm. ,, E: 3ccm. ,, + 1.000ccm. ,, + 10ccm. 2 1

F: 100ccm. ,, + 1.000ccm. ,) + 5ccm. Y Y

0 ccm. water + 1 *0o0 ccm. dichromate + 5 ccm. sulph. acid

Sulphuric acid: Merck’s pro analysi 11. Technique as usual. (Table on next page.)

ON ERRORS IX THE IODOMETRY OF BANG'S METHOD mc. 231

-

- - A

R

c

1)

E

F

Experiment 6. _. - - ~.

Temperature Thiosulphate Loss of of the liquids solution consumed 1 dichromate

I

850 c.

114O C .

64O C'.

85O C.

114O C.

32O C.

0.910 0.920 0.915 0.915 0.920 0.910

0.895 0.905 0.910 0 * 905 0.910 0.915

0.890 0.895 0.890 0.905 0 * 895

0.880 0.870 0.875 0.880 0.875 0.885

0 * 855 0.855 0.865 0.870 0.865 0-870

0.915

0.907

0.895

0.877

0.863

0.990 1

0'980 0.986 I 0.975

0.990 0.985 0.985 J

0.08.5

0.093

0.105

0.123

0.137

0.014

Percentage of loss of

dichromak

8.5

9.3

10.5

12.3

13.7

1.4

This shows that these variations in the quantities of water and sulphuric acid had an unmistakeable influence upon the loss of dichro- mate, this increasing with the ratio of water to sulphuric acid within the limits indicated and with the positive quantity of sulphuric acid.

The a,ssumption of an impurity in the sulphuric acid (and in the dichromate) is not sufficient to explain this; there must be still another factor that exerts its influence. And the probability is that it is a

question of the heat developed by the mixing of the water and the sulphuric acid; for where the quantity of sulphuric acid is constant (A and B: 5 ccm., C, D and E: 10 ccni.) the table shows an increctsivg i d u c i i o n of the diclwoimfe corresponding io the iiicrectsing del;elop)ne)if of heat. That such a variation between the quantity of water and tlie sulphuric acid brings about a variation in the heat developed is, by the way, confirmed by some measurements made by Blix (4).

In the above experiment i t was furthermore clear that series C, I) and E retained their high degree of heat much longer than series A and B (due to the greater total volume of the fluid in the first-named series); this, however, is hardly of any significance in this connection, as repeated experiments have shown us that the reduction of dichromate takes place approximately immediately after the adding of the sulphuric acid.

Naturally, the temperature of the mixtures was taken immediately after bringing the components together and by means of a thermometer previously heated in a similar mixture of water and sulphuric acid. Prior t o mixing, the various liquids had a temperature of 19OC.

These are the results we have obtained. If we summarise the observations made from all these experiments we arrive at the following conclusions :

1. Although in all our experiments we have employed reagents which complied with the conditions laid down by Bang and by Blix, we hace in no case succeeded in avoiding the loss of dichroninte shoic-n b?j

Heckscher. 2. This loss, this value of the blanks, is of fairly considerable

dimensions and in our experiments (the ordinary experimental method according to Bang) has varied between 0.051 and 0.129 ccm. of dichro- mate solution, i. e. between 5.1 and 12.90/0 of the dichromate added.

3. The loss is due to a l-eduction of dichromate in the greatly heated aqueons solution of dichromak urul cone. sulphul-ic acid prior to the dilution, for if we proceed in making the experiment so that the volume of water is greatly increased in proportion to the quantity of sulphuric acid, the sulphuric acid not being added until after the dichromate has been diluted with a large quantity of water (100 ccm.), the loss can be avoided.

4. The quality of the sulphuric acid was the principal governor of the size of the reduction of the dichromate, but even specially ptirified sulphuric acid preparations still gave a perceptible loss.

h. we cannot, h o ~ CWI’, deny that impuri/,t/ of /lw po/ / iwi iou-

6. The degrPP of hPaf of the nzi:rkiw of diehl.omnte-sull?hurie cicid is

7. There was no perceptible loss caused by the water itself. 8. With the foregoing in view it cannot he regarded as having

been determined whether the loss of dichromate is solely due to (more or less “accidentd”) impurities in the sulphuric acid (and in the potassium-dichroniate), or whether it is also partly brought about by “ordinary“ transformations between dichromate in aqueons solution and cone. sulphuric acid which, when mixed together, develop heat.

For purely practical purposes these observations lead to the following considerations which are of importance to anybody working with Bang’s method of lipoid determination:

The selection of sulphuric acid must be a very careful one - and maybe this also applies to the potassium dichromate; only those pre- parrations should he used which in the blinds prove to give only a slight loss.

At any rate, every time a fresh portion of sulphuric acid or dichromate solution is taken into use, blanks ought to be made in order to determine the size of the loss. Series of experiments in which this principle is not followed are hardly above criticism. An unobserved loss of dichromate of this origin will of course lead to the obtaining of analysis-results that are too high.

tlic1~ro)mte may have caused loss, although to a smaller extent.

probably of irnportancc.

Skandmav. Archiv. LII. 16

Liter a t u r e.

1 . Heckschrr, H., Biochem. Ztschr. 1927. Bd. CLXXXI. S. 444. 2. Bing. H. I., und Heckscher, H., Bioclwm. Zlschr. 1924. Bd. CXLIX.

3. Bang, I., Biochem. Ztschr. 1918. Bd. XCI. S. 86. 4. Bl ix , G., ‘‘Studies on diabelic lipemia.” Dissect. Lunrl. Sweden. 1928. 8. Maas, J., Biochem. Zfschr. 1924. Bd. CXLIV. S. 379. 6. von Szent-GyBrgyi, A., und Tominags, A., Biochem. Zlschr. 1924.

S. 83.

Bd. CXLVI. S. 226.