DETERMINATION OF CHLORIDES IN BIOLOGICAL FLUIDS BY THE USE OF ADSORPTION INDICATORS

Transcription

1 DETERMINATION OF CHLORIDES IN BIOLOGICAL FLUIDS BY THE USE OF ADSORPTION INDICATORS THE USE OF DICHLOROFLUORESCEIN FOR THE VOLUMETRIC MICRODETERMINATION OF CHLORIDES IN ZINC FILTRATES OF BIOLOGICAL FLUIDS BY ABRAHAM SAIFER AND JAMES HUGHES (From the Chemistry Division of the Department of Pathology, Queens General Hospital, Jamaica, New York) (Received for publication, April 11, 1939) This paper presents evidence that zinc filtrates (particularly the HagedornJensen (1923) filtrates) of urine (after treatment with hydrogen peroxide), whole blood, plasma, or serum are suitable for the accurate determination of chlorides with dichlorofluorescein as an indicator. The precision of this method for urine chlorides has been shown by comparison with established procedures such as the CaldwellMoyer (1935) modificat.ion of the VolhardHarvey (Harvey, 1910) method and the Van Slyke (192324) nitric acid digestion method for both normal and albuminous urines as recommended by Sendroy (1937). The method requires only one standard solution (0.02 N AgN03) and one indicator (0.05 per cent dichlorofluorescein in 50 per cent alcohol) for the determination of chlorides in urine, cerebrospinal fluids, blood serum, and plasma. EXPERIMENTAL Xolutions Standard sodium chloride gm. of pure, dry, recrystallized C.P. NaCl was dissolved in a liter of solution. St,andard silver nitrate. A 0.1 N solution of AgN03 was prepared from the pure, dry, recrystallized C.P. salt. This solution was standardized by titration with the standard sodium chloride with dichlorofluorescein as the indicator. An exact.ly 0.02 N 273

2 274 Determination of Chlorides solution of the AgN03 was prepared by dilution from the 0.1 N solution. Dichlorofluorescein indicator. Eastman Kodak Company s No. 373, 0.05 per cent in 50 per cent alcohol. Zinc sulfate per cent solution of ZnS01.7Hz0 prepared by dilution from a 45 per cent solution. Make up at least once weekly. 0.1 N NaOH. Prepared by standardization against 0.1 N HCl. Analysis of Standard Sodium Chloride Pure NaCl samples were pipetted into Pyrex testtubes (150 mm. X 15 mm.). 3 drops of 3 per cent Hz02 (chloridefree) were added and the tube was placed in a boiling water bath for 2 min Determination of Pure NaCl in HagedornJensen Filtrates N&l present NaCl found EWX mg. ml. per cent 1.oo $ fo z!zo.o utes. 5.0 ml. of 0.45 per cent zinc sulfate and 1.0 ml. of 0.1 N NaOH were added. The tube was heated in the boiling water bath for 3 more minutes and cooled under running water. The contents were filtered through a 4 cm. funnel, with Whatman No. 5 filter paper (7.0 cm.), into a 25 ml. Erlenmeyer flask. The tube and precipitate were washed with three 1.0 ml. portions of distilled water. 2 drops of the dichlorofluorescein indicator were added and the solution was titrated with 0.02 N AgNOs to the first definite pink coloration, The results obtained are given in Table I. I Analysis of Urine Samples To 0.2 ml. of urine in a Pyrex testtube were added 3 drops of 3 per cent C.P. Hz02 (chloridefree). The tube and its contents

3 A. Saifer and J. Hughes 275 were then heated in a boiling water bath for 2 minutes. 5.0 ml. of 0.45 per cent ZnSOa. 7Hz0 and 1.0 ml. of 0.1 N NaOH were added and t,he tube was heated in the water bath for 3 more minutes. The contents were cooled, filtered or centrifuged at high speed, and then transferred with washing into a 25 ml. Erlenmeyer flask, as described above. 2 drops of the dichlorofluorescein indicator were added and the filtrate was titrated with the 0.02 N AgN03 Determination of Chlorides in HagedornJensen Filtrate of Normal Urine (0.2 Ml. Samples) Average Average. Sample No. Caldwell Moyer method N&l O 1321.O 1310.o found Diohlorofluorescein indicator II 935.o o ~ 1305.o 1310.o o O O O N&l added 500.o 500.o 1000.o NaCl recovered by adsorption indicator method o 1010.o EWX pm cent z!zo.o 0.77 Average O 0.52 to the first definit e pink coloration which appeared throughout the solution. A set of triplicate determinations was performed on each urine sample with the CaldwellMoyer (1935) modification of the VolhardHarvey procedure in the following manner. To 0.2 ml. samples of the urine were added 1.0 ml. of concentrated HN03, 3.0 ml. of 0.02 N AgN03, 1.O ml. of nitrobenzene, and 1.0 ml. of ferric alum indicator. The excess AgNOB was then titrated with 0.02 N KCNS which had been previously standard

4 276 Det,ermination of Chlorides ized against the AgN03 solution. The first faint yellowish orange tint was taken as the endpoint. The results obtained for these urine determinations are given in Table II. Sendroy (1937) has pointed out that large errors occur in the usual Volhard chloride analyses when the urine is either low Comparison of Chloride Determinations in 7Jrine by Van Slyke and Adsorption Indicator Methods 4. Normal Sample No. Average o 5. Normal Average Albuminous. low chloride Average Albuminous. _. Average N&l an Slykc method %ny 1000.o III found :&F 1002.o 1005.o 1000.o 1090.O 1080.O 1090.O 1070.o 1080.o 1080.o O o o o o 690.o I N&l added o NaCl recovered by adsorption indioator method _ per cent in chloride content or contains large amounts of albumin. He used as a standard of comparison for his own iodate method Van Slyke s (192324) open Carius modification of the Volhard method with nitric acid ashing. To determine the effect of urines low in chloride and urines containing large amounts of f

5 A. Saifer and J. Hughes 277 albumin (4 + qualitative test) on the accuracy of the zinc filtrateadsorption indicator method, a large number of such urines were analyzed in the manner described above (except that 0.5 ml. samples of urine were used when low chloride values were expected), Chloride Determinations on Blood Serum or Plasma (0.2 Ml, Samples) IV Sample 1. Serum Average. 2. Serum Average. 3. Serum Average..~. 4. Plasma No. Adsorption indicstorzinc filtrate Van Slyke d$&dn mp. per cent o o Recoveries on added N&Cl by adsorption indicator Added mg. pm cent Recovered o o o Average O 5. Plasma Average o o o and t.he values obtained were checked against those obtained by triplicate analyses with the Van Slyke procedure as follows: To 1.0 ml. of urine (2.0 ml. when low chloride values are expected) in a large Pyrex testtube were added 5.0 ml. of 0.05 N. I

6 278 Determination of Chlorides AgNO, in concentrated HNOs and 1.0 ml. of distilled water. The tube and contents were heated in a boiling water bath for about 30 minutes and then cooled in cold water. 0.3 gm. of powdered ferric alum was added as an indicator and the excess AgNOa was titrated with st,andardized 0.02 N KCXS. The first orangeyellow tint lasting more than 30 seconds was taken as an endpoint. A t,ypical set of results for normal and albuminous urines is given in Table III. Analysis of Chlorides in Somogyi Whole Blood Zinc Filtrate (2.0 Ml. Samples Equal 022 Ml. of Whole Blood) Sample 1 No. V Adsorption indicator, CaldwellMoyer, Somogyi zinc filtrate Somogyi zinc filtrate 491.o 491.o o Average Average Average Recoveries were determined by analyzing samples of normal and albuminous urines prepared in the following manner. Known amounts of pure dry NaCl were weighed into 100 ml. volumetric flasks. The NaCl was dissolved and diluted to the mark with the previously analyzed samples of the normal or albuminous urine. Aliquot portions of these urines were then taken for analysis by the zinc filtrateadsorption indicator procedure. Resuhs of these analyses are shown in both Tables II a.nd III.

7 A. Saifer and J. Hughes 279 Analysis of Blood Serum or Plasma Chloride determinations were performed on zinc filtrates of blood serum or plasma with adsorption indicators in the following manner. 0.2 ml. of serum or plasma was pipetted into a Pyrex testtube (150 mm. X 15 mm.) and 1.0 ml. of 0.1 N NaOH was added. 5.0 ml. of 0.45 per cent ZnSOr.7Hz0 were then added drop by drop with constant stirring and the mixture was heated in a boiling water bath for 3 minutes and then cooled in cold water. The Determination of Chlorides on Whole Blood Samples (0.2 Ml.) Sample No. VI Adsorp$t;$ddicator Eisenman method Average Average... I Average I solution was then filtered, washed, and titrated in the same manner as given under the directions for known XaCl solutions. The results were compared against those obtained by the Van Slyke digestion procedure with 1.0 ml. samples of serum or plasma and titration of the excess AgN03 with a standard KCNS solution. These results are given in Table IV. Analysis of Whole Blood Chlorides Analysis of Somogyi s Zinc FiltrateSomogyi (1930) whole blood zinc filtrates were prepared in the usual manner. 2 or 3

8 Cetermination of Chlorides ml. aliquots of the filtrate were pipetted into a 25 ml. Erlenmeyer flask, 2 drops of the dichlorofluorescein indicator were added, and the filtrate titrated with 0.02 N AgNOs to the definite pink endpoint. The method was checked by running simultaneous determinations on the filtrates by the CaldwellMoyer (1935) modification of the Volhard procedure in which nitrobenzene is used to clarify and sharpen the endpoint. A typical set of results is given in Table V. Whole Blood ChloridesWhole blood chlorides were determined with adsorption indicators on 0.2 ml. samples of whole blood by the same procedure as given for serum or plasma. The results were checked against those obtained with Eisenman s (1929) method. A typical set of results is given in Table VI. DISCUSSION The endpoint change is from white to a pink coloration which is visible throughout the solution. However, a faint pink tinge appearing near the endpoint should not be mistaken for the endpoint. The endpoints obtained with zinc filtrates of biological fluids are not as sharp as those obtained with zinc filtrates of pure NaCl solutions or with adsorption indicators in organic media (Saifer and Kornblum, ), but are easily recognizable and distinct. A strong light shining over the shoulder of the investigator and the use of a white background for the titrations make the endpoint easy to detect. The zinc filtrate seems to aid in keeping the precipitated AgCl in colloidal form. It has also been found that slight amounts of oxidizing material, e.g. 0.2 ml. of a 3 per cent Hz02 solution, cause a decided increase in the amount of AgNO, used to obtain the endpoint with adsorption indicators. This error is easily avoided by driving off the excess Hz02 by heating the solution on a boiling water bath for several minutes. Values about 20 higher than the actual value are obtained when the adsorption indicator titrations are performed in the presence of the zinc hydroxide precipitate. The urine chloride values obt,ained by the adsorption indicator procedure check closely with those obtained by the CaldwellMoyer modification of the Volhard method and the Van Slyke nitric acid

9 A. Saifer and J. Hughes 281 digestion method. The method for urine takes less time than the Van Slyke procedure (as twenty to thirty determinations may be performed in an hour) and gives results which are just as reproducible. Table III shows that albuminous urine (giving a 4 + qualitative reaction with the Purdy heat test) gave results concordant with those by the Van Slyke digestion method. The maximum error obtained by the adsorption indicator method, with the Van Slyke method as a standard, was about 2 per cent and the average error was generally less than 1 per cent. The method as applied to whole blood (Somogyi) zinc filtrates and to HagedornJensen filtrates of plasma, serum, and whole blood was checked by comparison with established procedures. The average deviation of the method was about =t2 per cent. All filtrate and pipetting errors are eliminated by using the entire sample. The authors are grateful to Mr. Richard La Force for trying out the method previous to its publication and to Dr. Alfred Angrist, Pathologist and Director of Laboratories, for the interest he has shown in this work. SUMMARY Simple, accurate procedures for the determination of chlorides in zinc filtrates of urine, whole blood, plasma, and serum are described, with dichlorofluorescein as an adsorption indicator. Only one standard solution, 0.02 N AgKOa is used in the titration. The endpoint is sharp and the results obtained show an average error less than 2 per cent. BIBLIOGRAPHY Caldwell, J. FL, and Moyer, H. IT., Ind. and Eng. Chem., Anal. Ed., 1, 38 (1935). Eisenman, A. J., J. Biol. Chem., 82, 411 (1929). Hagedorn, H. C., and Jensen, B. N., Biochem. Z., 136, 46 (1923). Harvey, S. C., Arch. Int. Med., 6, 12 (1910). Saifer, A., and Kornblum, M., J. Biol. Chem., 112, 117 (193536). Sendroy, J., Jr., J. Biol. Chem., 120, 335, 441 (1937). Somogyi, M., J. Biol. Chem., 86, 655 (1930). Van Slyke, D. D., J. Biol. Chem.. 68, 523 (192324).

10 DETERMINATION OF CHLORIDES IN BIOLOGICAL FLUIDS BY THE USE OF ADSORPTION INDICATORS: THE USE OF DICHLOROFLUORESCEIN FOR THE VOLUMETRIC MICRODETERMINATION OF CHLORIDES IN ZINC FILTRATES OF BIOLOGICAL FLUIDS Abraham Saifer and James Hughes J. Biol. Chem. 1939, 129: Access the most updated version of this article at Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's alerts This article cites 0 references, 0 of which can be accessed free at tml#reflist1