J. BIOMED. MATER. RES.
VOL. 9, PP. 4 7 9 4 8 5 (1975)
Determination of Fluorocarbon in Blood T. P. STEIN, WINSTON K. ROBBINS, HELENE B. BROOKS, and HERBERT W. WALLACE, Department of Surgery, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104, and Exxon Research and Engineering Company , Linden, New Jersey 07036
Summary Because fluorocarbons can dissolve relatively large quantities of oxygen and carbon dioxide, there is considerable interest in utilizing them to develop new methods of extracorporael circulation, artificial red blood cells, and liquid breathing techniques. A method for the assay of fluorocarbon in blood is presented. The fluorocarbon is extracted from the blood with toluene, and fluoride is released from the fluorocarbon in the toluene extract by reaction with sodium biphenyl. The inorganic fluoride is then extracted with aqueous sodium acetate, the pH of the extract is adjusted, and the activity of the fluoride ion is read with a fluoridespecific ion electode. The assay was effective for fluorocarbon concentrations in the range 1 to 30 ppm.
INTRODUCTION Since fluorocarbons can dissolve relatively large quantities of oxygen and carbon dioxide, there is considerable interest in utilizing them to develop a new method of extracorporeal blood oxygenation,' as well as artificial red blood cells2 and liquid breathing technique^.^ I n animal studies a gross bolus of fluorocarbon in the circulation has ~ by blocking major been found to harm the ~ r g a n i s m ,probably capillary beds. Since most proposed methods of oxygenation using fluorocarbon involve direct contact of the fluorocarbon with blood or other biological tissues, inadvertent contamination of the blood is always possible. It is, therefore, import)ant to have an easy method of determining the amount of fluorocarbon in blood. Such a n analytic technique should: l) detect parts per million, or lower levels, of fluorocarbon compounds in blood; 2) be applicable to a wide variety 479 @ 1975 by John Wiley & Sons, Inc.
480
STEIN ET AL.
of fluorocarbons; 3) utilize as small a sample as possible ( A 1 0 ml); 4) be reasonably rapid; and 5) be unaffected by inorganic fluoride in the blood. Most fluorocarbons are detected by measuring either the molecular species directly, or the fluoride obtained after decomposition of the compound. The level of natural inorganic fluoride in blood, although OW,^'^ could interfere with a procedure in which decomposition of the blood sample is followed by measurement of total fluoride. Most accepted methods’ are applicable only after the fluorocarbon has been separated from the aqueous matrix. Our initial attempts to determine the fluorocarbon in the extract by electron capture gas chromatography were unsatisfactory because with the model compounds we used, widely differing chromatographic conditions were required to obtain quantitative results. Holaday e t al. encountered this problem in the application of gas chromatography t o FX-80. Our interest in a general technique applicable to mixtures led us to examine another method. Several investigator^^-'^ have taken advantage of the high sensitivity of the fluoride-specific ion electrode for the analysis of fluorocarbons after their decomposition. All but oneLZused the Schoniger flask combustion technique to decompose the fluorocarbon. Although this technique works well n i t h small samples (-100 mg), we found that the model fluorocarbons were lost if the organic extracts of blood were concentrated to volumes less than 2 ml. As an alternative, sodium biphenyl decomposition was studied. This reagent, a-hich decomposes organohalogen compounds in a manner analogous t o sodium fusion, has been successfully applied to a variety of Our procedure for determining fluorocarbon in blood combines sodium biphenyl decomposition with the specific ion electrode measurement of fluoride. A perfluoro cyclic ether (“Freon” E-4, DuPont, Wilmington, Delaware), a perfluorotributyl amine (FC-47, 3i?I Corporation, St. Paul, Minnesota), and a polymer of (chlorotrifluoroethylene), where n = 4, 5, and 6 (“11-21”, Halocarbon Products Corporation, Hackettstown, New Jersey) were used as model compounds.
PROCEDURE A standard curve is prepared by dissolving fluorocarbon in ethanol (1 mg/ml) and adding suitable aliquots of the solution to 10 ml of
DETER,MINATION OF FLUOROCARBON I N BLOOD
481
blood to obtain standards ranging from 1 to 30 ppm of fluoride in blood. For most fluorocarbons a favorable partition between water and a n organic solvent can be obtained. The best solvent for the model compounds was toluene. Furthermore, extraction of fluorocarbon by a n organic, solvent separates inorganic from organic fluorocarbon. If glass vessels are used for the extraction of fluorocarbon from blood with toluene, a very stable emulsion forms which prevents reproducible results. The use of polyethylene rather than glass centrifuge tubes and the addition of 2-octanol during the initial extraction minimize the formation of a foam layer. We had no problem with the formation of an emulsion during the remaining extraction steps. Each 10 ml sample is placed in a 40 ml polyethylene centrifuge tube, and two drops of 2-octanol and 15 ml of toluene are added. The mixture is agitated for 1 min in a vortex mixer and spun a t 12,000 X g for 5 min a t 20°C. The toluene layer is removed to a separatory funnel. This extraction procedure is repeated twice with 10 ml of toluene, and the 3 toluene extracts are combined. Fifteen ml of 2 N sodium biphenyl reagent (Southwestern Analytical Chemicals, Inc.) is added carefully with gentle shaking, and the mixture is allowed to stand for 15 min. (The sodium biphenyl must be used before its expiration date.) Isopropyl alcohol, 1 to 2 ml, is added dropwise to hydrolyze excess reagent (indicated by disappearance of the blue-green color). Two drops of 2-octanol followed by 5 ml of 50% acetic acid are added; the mixture is shaken to prevent the formation of an emulsion, and the aqueous layer is removed. The acetic acid extraction is repeated with 2 ml of 5% acetic acid. Two extractions with 2 ml HzO are performed. The aqueous extracts are combined, the pH is adjusted to 5.6 by adding saturated sodium hydroxide dropwise, and the total volume is adjusted to 20 ml with distilled water. The fluoride ion activity of the solution a t 25OC is measured with a fluoride ion electrode (Model 96-09, Orion Research, Inc.). The concentration of fluorocarbon in the unknown is determined from a calibration curve constructed from the fluoride activities of the standards. Each fluorocarbon requires a different standard curve. A standard curve and 8 analyses can be done in a working day. Fluoride recoveries for each step were determined with E-4 as the model compound by taking a series of 10 ml blood samples and a t
STEIN ET AL.
482
various stages of the procedure adding known amounts of E-4 to the toluene extract or sodium fluoride to the product from the biphenyl reaction. A standard curve for fluoride ion was prepared by adding suitable aliquots of sodium fluoride to the aqueous acetate extract, and the percentage of fluoride recovered was calculated by comparison with this standard curve.
RESULTS AND DISCUSSION Iiigure 1 shows a standard curve for E-4 fluorocarbon. Each point was determined independently a t least 16 times. The data in Table I indicate that the incomplete recovery of fluoride is due solely to incomplete recovery of fluorocarbon by the toluene extraction procedure. Toluene extracts from blood may contain certain organic compounds which react with sodium biphenyl to giye a water-soluble product (possibly cyanide) which interferes with the fluoride ion electrode. Depending on the condition of the toluene extractions,
- -50
32ppm
1 0 ppm
3 2 PPm
0 5
I0
I5
I
1
I
DETERMINATION O F FLUOROCARBON I N BLOOD
483
TABLE I Recovery of Fluoride for Each Step in Procedure Theoretical Fluoride Fluorocarbon No. of Concn Condition Samples (ppm)
mV
Determined Fluoride Concn % (ppm) Recovery
16 20
8.4 14.0
+4.7 f 2.4a -5.3 f 2 . 5
6.7 10.0
80 71
E-4 in toluene pre Na
6 6
8.4 14.0
+0.5 =t-5.6 -14.3 f 3 . 6
7.9 14.5
94 103
NaF in toluene post Na
6 6
8.4 14.0
-5.1 f 4.7 -16.6 f 1 . 3
9.9 15.8
118 113
N a F in acetate extract
7 7
8.4 14.0
-1.6 7.0 -13.3 f 6 . 4
8.4 14.0
100 100
E-4 in blood
+
Values are mean f S.D.
blank readings ranged from +70 to +25 mV. I n order to obtain accurate and highly reproducible results, it is essential that the toluene extraction procedure be rigorously standardized, as previously described. Since we were interested in a general method applicable t o a wide variety of fluorocarbons, we tested the assay on 2 other fluorocarbons, FC-47 and 11-21, and on a 1: 1 mixture of E-4 and FC-47. Under the conditions described, fluorocarbon concentrations as low as 1 ppm for E-4, 2 pprn for FC-47, and 1 ppm for 11-21 were detected. While the percentage of recovery differs for each fluorocarbon (Table 11) because of differences in the solubility of these compounds in the organic solvent, the method does seem to be generally applicable. Possibly in cases where low yields are obtained, changing the extraction conditions would help. The procedure also worked well for a 1: 1 mixture of E-4 and FC-47 (Table 11). An understanding of the degree of biological “inertness’) of fluorocarbon has been hindered by the lack of a n adequate method of assaying fluorocarbon in biological tissues. We believe that the approach presented in this communication (organic extraction of fluorocarbon, conversion t o fluoride, and specific ion electrode measurement) can be extended to include most fluorocarbons of interest. I n addition, it may be adaptable to fluorocarbons in other biological tissues
STEIN E T AL.
484
TABLE I1 Recovery of Fluoride with Different Fluorocarbons in Blood ~
Fluorocarbon E-4 FC-47 E-4 and FC-47
Halocarbon, series 11-21 a
Theoretical Fluoride No. of Concn Samples (ppm) 16 20 8 8 8 8 8 8
8.4 14.0 8.4 14.0 8.4 14.0 6.1 10.1
mV t4.7 -5.3 $23.8 +lO.l +14.1 -7.4 +8.3 -5.7
f 2.4'
f 2.5 f3.2 f6.1 f 4.3 f 3.9 f 2.6 f 2.9
I)et,ermined Fluoride Concn
74
(ppm)
Recovery
6.7 10.0 3.0 5 .3 4 .5 8.2 :i.6 10.2
80 71 36 38 53 59 108 101
Values are mean f S.D.
by homogenization prior to the organic extraction. Such a method would supplement existing techniques for the determination of fluoro organic compounds in biological matrices. 9 , 1 9 This work was supported in part by U.S. Public Health Service Grant HL 16595-01, and by American Heart Association Grant 71-1040.
References 1. 2. 3. 4.
5. 6. 7. 8.
P. S. Malchesky and Y. Nos&,J . Surg. Res., 10, 559 (1970). H. A. Sloviter and T. Kamimoto, Nature, 216, 458 (1967). L. C. Clark, Jr. and F. Gollan, Science, 152, 1755 (1966). Y. Nosh, T. Kon, D. Weber, G. Mrava, P. Malchesky, H. MacDermott, C. Williams, Jr., L. Lewis, G. Hoffman, C. Willis, S. Deodhar, G. Harris, and R. Anderson, Fed. Proc., 29, 1789 (1970). F. W. Barnes and J. Runcie, J . Clin. Pathol., 21, 668 (1968). H. H. Cox and 0 . B. Dirks, Caries Res., 1, 69 (1968). T. S. Ma, in Treatise of Analytical Chemistry, Vol. 12, Section B-1, I. M. Kolthoff and P. J. Elving, Eds., Interscience, New York, 1965. D. A. Holaday, V. Fiserova-Bergerova, and J. H. Modell, Anesthesiology, 37,
387 (1972). 9. M. Hanocq, Mikrochim. Acta, 1972, 707 (1972). 10. D. A. Levaggi, W. Oyung, and M. Feldstein, J. A i r Pollut. Contr. Ass., 21, 277 (1971). 11. J. Pavel, R. Kuebler, and H. Wagner, Microchem. J., 15, 192 (1970). 12. R. C. Rittner and T. S. Ma, Mikrochim. Acta, 1972, 404 (1972).
1)ETEHMINATION O F FLUOROCARBON I N BLOOD
485
13. W. Selig, Micro- and Semimicro-Determination of Fluorine in Organic Compounds, U.S. Atomic Energy Commission, UCRL-71697, 1969. 14. I). A. Shearer and G. F. Morris, Microchem. J., 15, 199 (1970). 15. M. B. Terry and F. Kasler, hfikrochim. Acta, 1971, 569 (1971). 16. R. I). Chambers, W. K. R. Musgrave, and J. Savoy, Analyst, 86,3j6 (1961). 17. P. Johncock, W. K. It. Musgrave, and A. Wiper, Analyst, 84, 245 (1939). 18 P. P. Wheeler and M. I. Fauth, Anal. Chem., 38, 1970 (1966). 19. R. J. Hall, Analyst, 93, 464 (1968).
Received November 12, 1974 Revised January 1, 1975
VOL. 9, PP. 4 7 9 4 8 5 (1975)
Determination of Fluorocarbon in Blood T. P. STEIN, WINSTON K. ROBBINS, HELENE B. BROOKS, and HERBERT W. WALLACE, Department of Surgery, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104, and Exxon Research and Engineering Company , Linden, New Jersey 07036
Summary Because fluorocarbons can dissolve relatively large quantities of oxygen and carbon dioxide, there is considerable interest in utilizing them to develop new methods of extracorporael circulation, artificial red blood cells, and liquid breathing techniques. A method for the assay of fluorocarbon in blood is presented. The fluorocarbon is extracted from the blood with toluene, and fluoride is released from the fluorocarbon in the toluene extract by reaction with sodium biphenyl. The inorganic fluoride is then extracted with aqueous sodium acetate, the pH of the extract is adjusted, and the activity of the fluoride ion is read with a fluoridespecific ion electode. The assay was effective for fluorocarbon concentrations in the range 1 to 30 ppm.
INTRODUCTION Since fluorocarbons can dissolve relatively large quantities of oxygen and carbon dioxide, there is considerable interest in utilizing them to develop a new method of extracorporeal blood oxygenation,' as well as artificial red blood cells2 and liquid breathing technique^.^ I n animal studies a gross bolus of fluorocarbon in the circulation has ~ by blocking major been found to harm the ~ r g a n i s m ,probably capillary beds. Since most proposed methods of oxygenation using fluorocarbon involve direct contact of the fluorocarbon with blood or other biological tissues, inadvertent contamination of the blood is always possible. It is, therefore, import)ant to have an easy method of determining the amount of fluorocarbon in blood. Such a n analytic technique should: l) detect parts per million, or lower levels, of fluorocarbon compounds in blood; 2) be applicable to a wide variety 479 @ 1975 by John Wiley & Sons, Inc.
480
STEIN ET AL.
of fluorocarbons; 3) utilize as small a sample as possible ( A 1 0 ml); 4) be reasonably rapid; and 5) be unaffected by inorganic fluoride in the blood. Most fluorocarbons are detected by measuring either the molecular species directly, or the fluoride obtained after decomposition of the compound. The level of natural inorganic fluoride in blood, although OW,^'^ could interfere with a procedure in which decomposition of the blood sample is followed by measurement of total fluoride. Most accepted methods’ are applicable only after the fluorocarbon has been separated from the aqueous matrix. Our initial attempts to determine the fluorocarbon in the extract by electron capture gas chromatography were unsatisfactory because with the model compounds we used, widely differing chromatographic conditions were required to obtain quantitative results. Holaday e t al. encountered this problem in the application of gas chromatography t o FX-80. Our interest in a general technique applicable to mixtures led us to examine another method. Several investigator^^-'^ have taken advantage of the high sensitivity of the fluoride-specific ion electrode for the analysis of fluorocarbons after their decomposition. All but oneLZused the Schoniger flask combustion technique to decompose the fluorocarbon. Although this technique works well n i t h small samples (-100 mg), we found that the model fluorocarbons were lost if the organic extracts of blood were concentrated to volumes less than 2 ml. As an alternative, sodium biphenyl decomposition was studied. This reagent, a-hich decomposes organohalogen compounds in a manner analogous t o sodium fusion, has been successfully applied to a variety of Our procedure for determining fluorocarbon in blood combines sodium biphenyl decomposition with the specific ion electrode measurement of fluoride. A perfluoro cyclic ether (“Freon” E-4, DuPont, Wilmington, Delaware), a perfluorotributyl amine (FC-47, 3i?I Corporation, St. Paul, Minnesota), and a polymer of (chlorotrifluoroethylene), where n = 4, 5, and 6 (“11-21”, Halocarbon Products Corporation, Hackettstown, New Jersey) were used as model compounds.
PROCEDURE A standard curve is prepared by dissolving fluorocarbon in ethanol (1 mg/ml) and adding suitable aliquots of the solution to 10 ml of
DETER,MINATION OF FLUOROCARBON I N BLOOD
481
blood to obtain standards ranging from 1 to 30 ppm of fluoride in blood. For most fluorocarbons a favorable partition between water and a n organic solvent can be obtained. The best solvent for the model compounds was toluene. Furthermore, extraction of fluorocarbon by a n organic, solvent separates inorganic from organic fluorocarbon. If glass vessels are used for the extraction of fluorocarbon from blood with toluene, a very stable emulsion forms which prevents reproducible results. The use of polyethylene rather than glass centrifuge tubes and the addition of 2-octanol during the initial extraction minimize the formation of a foam layer. We had no problem with the formation of an emulsion during the remaining extraction steps. Each 10 ml sample is placed in a 40 ml polyethylene centrifuge tube, and two drops of 2-octanol and 15 ml of toluene are added. The mixture is agitated for 1 min in a vortex mixer and spun a t 12,000 X g for 5 min a t 20°C. The toluene layer is removed to a separatory funnel. This extraction procedure is repeated twice with 10 ml of toluene, and the 3 toluene extracts are combined. Fifteen ml of 2 N sodium biphenyl reagent (Southwestern Analytical Chemicals, Inc.) is added carefully with gentle shaking, and the mixture is allowed to stand for 15 min. (The sodium biphenyl must be used before its expiration date.) Isopropyl alcohol, 1 to 2 ml, is added dropwise to hydrolyze excess reagent (indicated by disappearance of the blue-green color). Two drops of 2-octanol followed by 5 ml of 50% acetic acid are added; the mixture is shaken to prevent the formation of an emulsion, and the aqueous layer is removed. The acetic acid extraction is repeated with 2 ml of 5% acetic acid. Two extractions with 2 ml HzO are performed. The aqueous extracts are combined, the pH is adjusted to 5.6 by adding saturated sodium hydroxide dropwise, and the total volume is adjusted to 20 ml with distilled water. The fluoride ion activity of the solution a t 25OC is measured with a fluoride ion electrode (Model 96-09, Orion Research, Inc.). The concentration of fluorocarbon in the unknown is determined from a calibration curve constructed from the fluoride activities of the standards. Each fluorocarbon requires a different standard curve. A standard curve and 8 analyses can be done in a working day. Fluoride recoveries for each step were determined with E-4 as the model compound by taking a series of 10 ml blood samples and a t
STEIN ET AL.
482
various stages of the procedure adding known amounts of E-4 to the toluene extract or sodium fluoride to the product from the biphenyl reaction. A standard curve for fluoride ion was prepared by adding suitable aliquots of sodium fluoride to the aqueous acetate extract, and the percentage of fluoride recovered was calculated by comparison with this standard curve.
RESULTS AND DISCUSSION Iiigure 1 shows a standard curve for E-4 fluorocarbon. Each point was determined independently a t least 16 times. The data in Table I indicate that the incomplete recovery of fluoride is due solely to incomplete recovery of fluorocarbon by the toluene extraction procedure. Toluene extracts from blood may contain certain organic compounds which react with sodium biphenyl to giye a water-soluble product (possibly cyanide) which interferes with the fluoride ion electrode. Depending on the condition of the toluene extractions,
- -50
32ppm
1 0 ppm
3 2 PPm
0 5
I0
I5
I
1
I
DETERMINATION O F FLUOROCARBON I N BLOOD
483
TABLE I Recovery of Fluoride for Each Step in Procedure Theoretical Fluoride Fluorocarbon No. of Concn Condition Samples (ppm)
mV
Determined Fluoride Concn % (ppm) Recovery
16 20
8.4 14.0
+4.7 f 2.4a -5.3 f 2 . 5
6.7 10.0
80 71
E-4 in toluene pre Na
6 6
8.4 14.0
+0.5 =t-5.6 -14.3 f 3 . 6
7.9 14.5
94 103
NaF in toluene post Na
6 6
8.4 14.0
-5.1 f 4.7 -16.6 f 1 . 3
9.9 15.8
118 113
N a F in acetate extract
7 7
8.4 14.0
-1.6 7.0 -13.3 f 6 . 4
8.4 14.0
100 100
E-4 in blood
+
Values are mean f S.D.
blank readings ranged from +70 to +25 mV. I n order to obtain accurate and highly reproducible results, it is essential that the toluene extraction procedure be rigorously standardized, as previously described. Since we were interested in a general method applicable t o a wide variety of fluorocarbons, we tested the assay on 2 other fluorocarbons, FC-47 and 11-21, and on a 1: 1 mixture of E-4 and FC-47. Under the conditions described, fluorocarbon concentrations as low as 1 ppm for E-4, 2 pprn for FC-47, and 1 ppm for 11-21 were detected. While the percentage of recovery differs for each fluorocarbon (Table 11) because of differences in the solubility of these compounds in the organic solvent, the method does seem to be generally applicable. Possibly in cases where low yields are obtained, changing the extraction conditions would help. The procedure also worked well for a 1: 1 mixture of E-4 and FC-47 (Table 11). An understanding of the degree of biological “inertness’) of fluorocarbon has been hindered by the lack of a n adequate method of assaying fluorocarbon in biological tissues. We believe that the approach presented in this communication (organic extraction of fluorocarbon, conversion t o fluoride, and specific ion electrode measurement) can be extended to include most fluorocarbons of interest. I n addition, it may be adaptable to fluorocarbons in other biological tissues
STEIN E T AL.
484
TABLE I1 Recovery of Fluoride with Different Fluorocarbons in Blood ~
Fluorocarbon E-4 FC-47 E-4 and FC-47
Halocarbon, series 11-21 a
Theoretical Fluoride No. of Concn Samples (ppm) 16 20 8 8 8 8 8 8
8.4 14.0 8.4 14.0 8.4 14.0 6.1 10.1
mV t4.7 -5.3 $23.8 +lO.l +14.1 -7.4 +8.3 -5.7
f 2.4'
f 2.5 f3.2 f6.1 f 4.3 f 3.9 f 2.6 f 2.9
I)et,ermined Fluoride Concn
74
(ppm)
Recovery
6.7 10.0 3.0 5 .3 4 .5 8.2 :i.6 10.2
80 71 36 38 53 59 108 101
Values are mean f S.D.
by homogenization prior to the organic extraction. Such a method would supplement existing techniques for the determination of fluoro organic compounds in biological matrices. 9 , 1 9 This work was supported in part by U.S. Public Health Service Grant HL 16595-01, and by American Heart Association Grant 71-1040.
References 1. 2. 3. 4.
5. 6. 7. 8.
P. S. Malchesky and Y. Nos&,J . Surg. Res., 10, 559 (1970). H. A. Sloviter and T. Kamimoto, Nature, 216, 458 (1967). L. C. Clark, Jr. and F. Gollan, Science, 152, 1755 (1966). Y. Nosh, T. Kon, D. Weber, G. Mrava, P. Malchesky, H. MacDermott, C. Williams, Jr., L. Lewis, G. Hoffman, C. Willis, S. Deodhar, G. Harris, and R. Anderson, Fed. Proc., 29, 1789 (1970). F. W. Barnes and J. Runcie, J . Clin. Pathol., 21, 668 (1968). H. H. Cox and 0 . B. Dirks, Caries Res., 1, 69 (1968). T. S. Ma, in Treatise of Analytical Chemistry, Vol. 12, Section B-1, I. M. Kolthoff and P. J. Elving, Eds., Interscience, New York, 1965. D. A. Holaday, V. Fiserova-Bergerova, and J. H. Modell, Anesthesiology, 37,
387 (1972). 9. M. Hanocq, Mikrochim. Acta, 1972, 707 (1972). 10. D. A. Levaggi, W. Oyung, and M. Feldstein, J. A i r Pollut. Contr. Ass., 21, 277 (1971). 11. J. Pavel, R. Kuebler, and H. Wagner, Microchem. J., 15, 192 (1970). 12. R. C. Rittner and T. S. Ma, Mikrochim. Acta, 1972, 404 (1972).
1)ETEHMINATION O F FLUOROCARBON I N BLOOD
485
13. W. Selig, Micro- and Semimicro-Determination of Fluorine in Organic Compounds, U.S. Atomic Energy Commission, UCRL-71697, 1969. 14. I). A. Shearer and G. F. Morris, Microchem. J., 15, 199 (1970). 15. M. B. Terry and F. Kasler, hfikrochim. Acta, 1971, 569 (1971). 16. R. I). Chambers, W. K. R. Musgrave, and J. Savoy, Analyst, 86,3j6 (1961). 17. P. Johncock, W. K. It. Musgrave, and A. Wiper, Analyst, 84, 245 (1939). 18 P. P. Wheeler and M. I. Fauth, Anal. Chem., 38, 1970 (1966). 19. R. J. Hall, Analyst, 93, 464 (1968).
Received November 12, 1974 Revised January 1, 1975
