pH. It has been previously shown that 2,9-dimethyl-1,10phenanthroline is a specific reagent for copper(1) when the resulting complex is extractable with n-hexyl alcohol. The complex forms in pH range of 3 to 10 is very stable and is extractable within this pH range (12). Diverse Ions. The effect of 10 common diverse ions was checked. Table I summarizes the results which showed that more than about 10 ppm of nitrate, ammonium, magnesium, and aluminum should not be present and that 15 ppm of acetate chloride, sulfate, phosphate, potassium, and iron(I1) can be tolerated. In the case of magnesium and aluminum, the error assdciated with higher concentrations may be in part due to the high chloride concentration concomitant to
the use of magnesium(I1) chloride and aluminum(I1) chloride in the diverse ion study. Precision. An estimate of the precision obtainable with each method was obtained by analyzing six solutions containing 75 pg of perchlorate per 25 ml of aqueous solution. The indirect spectrophotometric method gave a mean absorbance value of 0.413,a standard deviation of 0.001 absorbance unit, or a relative standard deviation of 0.24%. The six samples analyzed by the indirect atomic absorption spectrometric method gave a mean absorbance of 0.463,a standard deviation of 0.002,or a relative standard deviation of 0.43%. RECEIVED for review May 27, 1968. Accepted July 3, 1968.
Use of Sodium Cerium EDTA for Titrimetric Determination of Metallic Elements with Ethylenediaminetetraacetic Acid Stanley S. Yamamura Idaho Nuclear Corporation, Idaho Falls, Idaho
THEuse of EDTA in chemical analyses has been studied and reported very extensively in the last 25 years. This popularity has led to the generation of much useful information and many useful methods of analysis. Unfortunately, it also has created a dilemma for the analytical chemist. Table I, based on information presented in one table in the “Handbook of Analytical Chemistry” edited by Meites ( I ) , shows the abundance of available methods for some common metals. It is easy to see that an analyst, especially one unaccustomed to EDTA methodology, is inclined to ask the question, “Which of this bewildering array of methods should I try first?” The purpose of the present study was to establish one procedure which could be used to determine reliably a large number of metals without special procedural adaptations for different metals. The titrimetric procedure based on the use of sodium cerium(II1) EDTA octahydrate (NaCeEDTA 8Hz0) salt ( 2 ) satisfied these requirements. This method has been found to be applicable to 33 metals including Am, Bi, Cd, Co, Cu, Fe, Ga, Hg, In, the lanthanides, Mn, Ni, Pb, Sc, T1, V, Y ,Zn, and Zr. It probably i s applicable to Bk, Cf, Cm, and Hf as well. EXPERIMENTAL
Apparatus and Reagents. All titrations were performed with a standard 10-mI buret calibrated in 0.05-ml divisions. Analytical Reagent and spectrographically-analyzed chemicals were used throughout without purification. The 0.05M EDTA titrant was prepared from the disodium salt, filtered through a 0.45-micron membrane filter, and standardized against standard hydrochloric acid solutions of zinc metal and zinc oxide by direct titration at pH 5.5 to a xylenol orange, 3 ’,3"-his( bis(carboxymethy1)aminomethyl}5 ’,5 ”-dimethylphenolsulfonaphthalein,end point and against (1) L. Meites, “Handbook of Analytical Chemistry,” McGrawHill, New York, 1963, pp 3-164 to 3-200. (2) S. S. Yamamura, ANAL.CHEM., 36, 1858 (1964). 1898
ANALYTICAL CHEMISTRY
Table I. Availability of Visual EDTA Titration Methods Different modes Number of Metal ion of titrationo references 36 43 Bi 3+ 53 56 Cd2+ 69 63 cuz+ 50 63 Fe3+ 61 67 Ni2+ 66 58 PbZ+ 68 59 Zn2+ 0 In a number of cases, several variations are listed for a given mode of titration, hence the number of available methods exceeds the indicated number of modes of titration.
standard nitric acid solutions of bismuth metal, mercury metal, indium metal, nickel metal, and lead nitrate salt by the proposed method. The seven standardization values with a relative standard deviation of 0.08% were averaged. The nature and compositions of the metal solutions used in this study are summarized in Tables I1 and 111. Xylenol orange was used as a 0.2z (w/v) solution in water while arsenazo, o-( 1,8-dihydroxy-3,6-disulfo-2-naphthylazo) benzenearsonic acid, was used as a 1 (w/w) mixture with solid sodium chloride. Sodium cerium(II1) EDTA was prepared as described in a previous paper (2) and used as the solid. Recommended Titration Procedure. Pipet a sample aliquot containing 0.05 to 0.45 mmole of titratable metal into a 150-ml beaker. Add 0.5 ml of concentrated perchloric acid and dilute to about 120 ml with water. Add 0.50 gram of NaCeEDTA.8H20, stir for about 2 minutes, then add 5 drops of 0.2z (w/v) xylenol orange or, if copper is being determined, sufficient arsenazo-sodium chloride mixture to give a rich red-orange coloration. Noting the number of drops, add pyridine dropwise to the appearance of a red or violet coloration, then add two times as much additional pyridine. If the red or violet coloration does not appear
z
Table 11. Titration Results of Metal Determinable by NaCeEDTA “Replacement” EDTA Titrimetry Log Kj of metal-EDTA Remarks Titration result“ complex Metal ion and solution used Cm, Bk, and Cf also are believed to be 0.02485 f 0.00012M Am(III), 0.02500M Am2(SO& 18.2 determinable. 0.05199 & 0.OOOOOM Bi(III), 0.05183M nitrate solution pre26.5 pared from the pure metal 0.04906 =!= 0.00005M Cd(II), 0.04915Msolution prepared from 16.5 Cd(N03)2‘4Hz0 At high cobalt levels, the pink color of 0.1696 zt 0.0001M Co(II), 0.1697M nitrate solution pre16.3 the Co(I1)-EDTA complex diminpared from the pure metal ishes the sharpness of the end point. Use arsenazo indicator. Cu(I1) reacts 0.1579 f 0.0002M Cu(II), 0.1574M nitrate solution pre18.8 “irreversibly” with xylenol orange pared from the pure metal at the end point. For selectivity, Fe(II1) can be pre0.1796 f 0.0002M Fe(III), 0.1791M perchlorate solution 25.1 separated by extraction into 2-octaprepared from the pure metal none from 7M HCI medium ( 2 ) . 0.04277 =!= 0.00003M 20.3 Ga(III), 0.04283M nitrate solution prepared with Ga(NO,), .8H20 0.05009 i 0.00005M 21.8 Hg(II), 0.05010M nitrate solution prepared from the pure metal 0.05348 f O.OOOOOM 25.0 In(III), 0.05350M nitrate solution prepared from the pure metal Because Mn(I1)-EDTA, log K, = 0.1819 f 0.0002M ~13.9 Mn(III), 0.1820M nitrate solution pre13.9 zt 0.1, is much less stable than pared from the pure metal Ce(1II)-EDTA, log K , = 16.0, only a small quantity of cerium(II1) is released and the end point, though still discernible, is not very sharp. For selectivity, Ni(I1) can be pre18.6 0.04979 =k 0.00006M Ni(II), 0.04985M solution prepared from separated by precipitation with dithe pure metal methylglyoxime from tartrate medium. 0.05158 f O.0000lM 18.0 Pb(II), 0.05159M solution prepared from Pb(N03)2 Because of limited supplies, quantitaSatisfactory 23.1 Sc(III), -0.050M tive studies were not done. Because of the high stability of the Sc(II1)EDTA complex and the chemical similarity of Sc(II1) to the heavy lanthanides, the titration is expected to be quantitative and highly precise. The TI(II1) concentration was found to 0.04419 i O.OOOO5M TI(III), 0.05196M nitrate solution prebe 0.04406M by a reverse EDTA pared by HC104-HNOa dissolution of titration procedure. Attempts to TI203 oxidize all of the thallium to the (111) oxidation state were not successful. 18.8 0.05103 rt 0.00005M V(V) was reduced with NH20H.HCl. V(IV), 0.05127M solution of NH4VO3 The intense blue color of the VO2+EDTA complex alters the end point color transition. Zn(II), 0.04284M chloride solution pre16.5 0.04284 f 0.00002M Direct titration of the zinc to a xylenol orange end point gave 0.04284M. pared from the pure metal The method is applicable to nitrate, 20 to 25 0.05513 f 0.00005M Zr(IV), 0.05580M chloride solution p r e chloride, and perchlorate solutions of pared with ZrO2CI2~ 8H20 zirconium but not to sulfate solutions. Samples should be adjusted to 3M HN03 and boiled for 5 to 10 min initially to break up polymers (3).
Hf also is thought to be determinable. a Each result, the average of at least 3 titrations, is listed with the calculated standard deviation. Xylenol orange or arsenazo can be used interchangeably except in the case of copper(I1).
after the addition of 1 ml of pyridine, add concentrated ammonium hydroxide dropwise till it appears, then add 2 ml of additional pyridine. Titrate with standard 0.05M EDTA solution to the complete dissociation of the cerium(II1) indicator complex-Le., until a color change no longer is
observed on further addition of titrant. The end point transition is violet to red to yellow for xylenol orange and violet to red to reddish-orange for arsenazo. (3) B. C. Sinha and S. Das Gupta, Analyst, 92, 558 (1967). VOL 40, NO. 12, OCTOBER 1968
1899
Table 111. Titration Results of Metals Not Determinable by the Proposed Method Metal ion and solution used AI(III), 0.05240M chloride solution prepared from the pure metal
Ba(II), 0.050M nitrate solution Ca(II), 0.050M nitrate solution CrUIIl 0.1923M nitrate solution prepared from the pure metal Mg(II), 0.050M nitrate solution Sn(I1,IV), -0.OSMchloride solution Sr(II), 0.050M nitrate solution Th(IV), 0.04164M Solution of Th(NO&.4 HzO
Log 4 metalEDTA complex 16.1
7.8 10.7
-23.5 8.7
... 8.6 23.2
Results and remarks Results are low because of hydrolysis of Al(II1) at pH 5.6 =k 0.25. Also, AI(II1) reacts irreversibly with the indicator at the end point. By titrating rapidly and observing the end point carefully, a result of 0.0492M was obtained. Ba(I1) is not titrated at all at pH 5.60 =t0.25. Ca(I1) is titrated , Sr, Th, Tl(III), V(IV), Zn, and Zr. Excluded from the tests were the lanthanides and yttrium which, though accurately determinable by the proposed procedure, can be determined by direct EDTA titration under similar conditions. Tables I1 and I11 summarize the experimental results and pertinent accompanying remarks. Those metals that are determinable are listed in Table 11; those that are not, in Table 111. Of the 25 metallic elements studied, 17 metals including Am, Bi, Cd, Co, Cu (by using arsenazo indicator), Fe, Ga, Hg, In, Mn, Ni, Pb, Sc, T1, V, Zn, and Zr are determinable without bias at a precision of better than 0 . 2 0 z relative standard deviation. This list can be expanded to include the lanthanides and yttrium and probably hafnium which resembles zirconium and curium, berkelium, and californium which are chemically similar to americium(II1) for a total of 37 metallic elements, The nondeterminable elements fall into three groups: those (alkali metals, Be, Ba, Ca, Mg, and Sr) that do not form stable EDTA complexes at pH 5.6 i 0.25, those (Al, Cr, and Sn) that hydrolyze under the analysis conditions, and those (Al, Th) that react “irreversibly” with the indicator. Copper, when xylenol orange is used, falls into the third group. Desirable Features of the Method. Desirable features of the proposed titration procedure are wide applicability, extreme simplicity, and high reliability. The end point color transitions for both xylenol orange and arsenazo are sudden and easily recognizable and the titration is not affected adversely by high salt concentrations; hence, it is compatible with separation schemes that introduce large amounts of electrolyte. For example, no harmful effect has been observed even in the presence of 2 ml of concentrated hydrochloric acid, nitric acid, or perchloric acid. Another desirable characteristic i s that the pH can be adjusted easily with complete freedom from problems of hydrolysis. This is especially advantageous in the determination of metal ions, such as Bi(III), Fe(III), Hg(II), and Zr(IV), which readily hydrolyze during pH adjustment with base. The proposed procedure is recommended for the standardization of pure metal solutions that are encountered routinely in all branches of chemistry and in quality control programs. It is not designed for the analysis of complex multicomponent solutions ; however, some characterization of such systems is possible by the application of masking agents and selective separation procedures and by the use of EDTA titration techniques involving masking and demasking.
Possible Future Extensions. The applicability of the proposed idea can be broadened in several ways. First, NaCeEDTA 8Ht0 and other analogous lanthanide salts might be used with xylenol orange in the development of a general procedure for the determination of a wide variety of metals at microgram levels. Second, ytterbium-EDTA (log Kj ‘U 19.9) might be used to determine selectively those metals that form highly stable EDTA complexes. Finally, substitution of DTPA (diethylenetriamine pentaacetic acid) for EDTA
may permit the determination of those elements that currently are not determinable because of their tendency to hydrolyze and to react irreversibly with the indicator. YtterbiumDTPA, for example, has been used effectively for the titration of thorium in pyridine-acetate medium to a xylenol orange end point. RECEIVED for review February 5, 1968. Accepted June 25, 1968.
Determination of Inorganic Phosphate in the Presence of Adenosine Triphosphate by the Automatic Reaction Rate Method S. R. Crouch and H. V. Malmstadt Department of Chemistry and Chemical Engineering, University of Illinois, Urbana, Ill.
I N A RECENT article ( I ) , we presented a new analytical method for inorganic orthophosphate based on automated measurements of the initial rate of formation of molybdenum blue from phosphate, molybdate, and ascorbic acid. In comparison to the classical molybdenum blue procedure for phosphate (2), which requires 5-10 minutes for color development, the reaction rate procedure requires only 20-30 seconds of reaction time per sample. Thus, the rate procedure should be advantageous for the determination of inorganic phosphate in biological samples which contain readily hydrolyzable phosphate esters. Several authors have discussed the problems of phosphate determinations in such samples (3, 4). The hydrolysis of phosphate esters, such as adenosine triphosphate (ATP), is both acid and molybdate catalyzed ( 4 , 5). One common modification of the molybdenum blue procedure (6) allows the reaction to be carried out at pH 4 minimizing the acidcatalyzed hydrolysis. However, as has been pointed out (3, pH 4 is optimum for the molybdate-catalyzed hydrolysis of the terminal phosphate of ATP. Because of the slowness of the color development, almost all the common methods for the determination of phosphate utilize 5-10 minutes of reaction time during which extensive hydrolysis of phosphate esters can occur. EXPERIMENTAL
The spectrophotometric reaction rate measuring system has been previously described ( I ) . Samples were analyzed by both the reaction rate method ( I ) and the method of Fiske and Subbarow (2). A 10-minute color development period was used with the latter method. A Spectronic 20 colorimeter (Bausch & Lomb) was used for analyses by the method of Fiske and Subbarow. Solutions containing ATP were prepared from the disodium salt (Sigma Grade, Sigma Chemical Co., St. Louis, Mo.). (1) S. R. Crouch and H. V. Malmstadt, ANAL.CHEM.,39, 1090 (1967). (2) C . H. Fiske and Y . Subbarow, J . Biol. Chem., 66, 375 (1925). (3) M. Blecher, Anal. Biochem., 7, 383 (1964). (4) B. B. Marsh, Biochim. Biophys. Acta, 32,357 (1959). (5) H. Weil-Malherbe and R. H. Green, Biochem. J., 49, 286 (1951). (6) 0.H. Lowry and J. A. Lopez, J.Biol. Clzem., 162,421 (1946).
61801
I[d G/
15 sec
,
TIME
Figure 1. Recorded reaction rate curves for inorganic phosphate and for ATP (a).
(b).
4 ppm P 10-ZM ATP
Blood serum samples were deproteinized with trichloroacetic acid and centrifuged as previously described ( I ) . For the rate measurements, neutralization of the serum supernatant is necessary because of the strong dependence of the rate on sample acidity (7). Serum supernatant analyzed by the Fiske and Subbarow method was not neutralized. RESULTS AND DISCUSSION
In Figure 1, recorded reaction rate curves for the formation of phosphomolybdenum blue from inorganic phosphate and from ATP are shown. After an initial increase of absorbance upon starting the reaction with ascorbic acid, ATP shows only a very small rate even at the 10-2M level. After about
(7) S. R. Crouch and H. V. Malmstadt, ANAL.CHEM.,39, 1084 (1967). VOL 40, NO. 12, OCTOBER 1968
1901
