*
deviation was 7x. Reproducibility was similar for analysis of extracts containing 100 p g of BaP per gram. Results obtained by the spectrophotometric methods described in this paper were compared with those obtained by the column chromatographic spectrophotometric and the thin layer chromatographic spectrophotofluorimetric methods (Table 11). Based on seven or more determinations for each method, the percent relative standard deviations calculated were *lo, 110, *7, and =7tz for methods 1, 2 , 3, and 4, respectively. The investigation showed also that benzo[e]pyrene could be
determined in the same fractions eluted for the determination of benzo[a]pyrene. Further research is needed to determine how many polycyclic hydrocarbons can be analyzed by this procedure. ACKNOWLEDGMENT The technical assistance of Ethel Grisby and Margaret Smith is acknowledged.
RECEIVED for review April 6, 1967. Accepted June 29, 1967. Use of commercial products does not constitute endorsement by the Public Health Service.
Separation of Zirconium(1V) and Thorium(lV) from Each Other and from Numerous Metal Ions by Aqueous Paper Chromatography Mohsin Qureshi and Fahmida Khan Cheniical Laboratories, Aligarli Muslim Unicersity, Aligarli, U . P . , India VERYFEW systematic studies have been reported on the paper chromatography of zirconium and thorium systems, which pose many analytical problems. Zr(IV) has been separated from Th(1V) by Balkrishnamurty(/) and by Lebez and Pirs ( 2 ) . The separation of Balkrishnamurty requires 9 hours, and the work of Lebez and Pirs could not be reproduced by the junior author. The effect of varying Zr-Th ratios has not been studied in either case. Th(1V) was separated from some metal ions by Almassy and coworkers (3). Phosphate did not interfere. Zr(IV), was separated from numerous metal ions by Lederer ( 4 ) while he was studying the behavior of various cations in HCI-alcohol systems. The effect of foreign ions (1-4) has not been investigated. Almassy et a/. ( 5 ) use three different solvent systems to separate one of the cations Ti(IV), Zr(IV), and Th(IV) from the rest. This is obviously inconvenient and time consuming. The paper chromatography of Zr-Th systems, therefore, needs further study. Sandell ( 6 ) has reported that thorium can be separated from titanium, zirconium, iron, and most metals (except the rare earths) by precipitating with oxalic acid (not oxalate) in dilute mineral acids. He (7) also recommends the use of (1 : I O ) hydrochloric acid for the separation of zirconium phosphate from numerous metal ions by precipitation. If suitably, modified these aqueous precipitation systems may provide rapid and selective separations by paper chromatography. The present report summarizes the results of such an approach. EXPERIMENTAL
Apparatus. Whatman No. 1 strips, 14.5 X 3 cm, were developed in 20 X 5 cm glass jars using the ascending method. Reagents. Reagent grade chemicals were used. (1) V. V. ( 1961).
Balkrishnamurty, J . Sci. hid. Res. (India), 20B, 453-4
(2) D. Lebez and M. Pirs, “ J . Stefan” Ztrst. Repts. (Ljib/jutza), 3, 171-3 (1956). (3) G. Almassy, M. Ordogh, and B. Hadobas, Magy. Tud. Akad. Kozp. Fiz.Kilt. Itit. Kozlemen., 7, 67-75 (1959). (4) M. Lederer, A m d . Chim.Acta, 5, 185-90 (1951). (5) G. Almassy and Z . Nagy, Mugy. Kem. Folyoirut., 60, 206-9 (1954). (6) E. B. Sandell, “Colorimetric Metal Analysis,” 3rd ed., Interscience, New York, 1959, p. 838. (7) E. B. Sandell, lbid., p. 966.
Cation Solutions. All test solutions were 0.1M unless otherwise stated. Zirconium and thorium nitrates were dissolved in 0.1M nitric acid, thorium chloride in 0.1M HCI, and hafnyl sulfate in 2M perchloric acid. Arsenious oxide was dissolved in 0.5M nitric acid. Sodium molybdate and sodium tungstate were dissolved in distilled water. ZrO(II), Sn(II), Sn(IV), Sb(III), and Ti(IV) chlorides were dissolved in 4 M hydrochloric acid. The rest of the cations were dissolved as nitrates in O.1Mnitric acid. Indicators. Cations detected are listed for each indicator as follows. Yellow ammonium sulfide: Ag(I), Hg2(II), ammonium Hg(II), Bi(III), Pb(I1); 1 aluminon in 1 acetate: Cr(III), AI(III), Ca(II), Ba(II), Sr(II), Mg(I1);
Table I. Chromatography of Zr and Th in Some Oxalic Acid Systems R values Solvent system Th Zr Hf Conc. HC1 satd. soh. of oxalic acid water” 0.00 0.74-1.00 0.89-1.00 2:7:1 0.00-0.1 2 0.78- 1 .OO 0.87-1 . 00 2:6:2 0.00-0.04 0.74-1.00 0.91-1.00 2:5:3 0.00-0.06 0.75-1.00 0.89-1.00 2:4:4 0.00-0.18 0.75-1.00 0.83-1.00 2:3:5 2:2:6 0.00-0.20 0.76-1.00 0.87-1.00 0.00-0.04 0.83-1.00 0.90-1.00 2:8:10
+
+
Conc. HC1 4-10% oxalic acid 6.25z amrnoniurn oxalate* 1:l:S 1 :2:7 1 :3:6 1 :4:5 1:5:4 1 :6:3
+
0.00-0.12 0.80-1.00 0.87-1.00 0.00-0.12 0.75-1.00 0.87-1.00 0.00-0.06 0.75-1.00 0.90-1.00 0.00 0.79-1 .00 0.87-1 . 00 O.O0-0.18* 0.74-1.00’ 0.87-1.00 0.00-0.06 0.80-1.00 O.WO.03 0.75-1.00 0.89-1.00 0.0&0.17* 0.71-1.00* 1 :7:2 0.00-0.04 0.80-1.00 0.88-1.00 O.WO.19* 0.67-1 .oO* 1:S:l 0.00-0.04 0.75-1.00 0.90-1.00 4 Time of development 40 minutes, temp. 28-29’ C. Time of development 45 minutes, temp. 30-31” C. ’ * = Average values for two separations actually achieved. ~~
~
VOL. 39, NO. 11, SEPTEMBER 1967
1329
10% aqueous KI: Tl(I), Tl(II1); 0.1% alcoholic solution of alizarin: rare earths; 2% dithizone in chloroform: As(III), Sb(III), Zn(lI), Cd(I1); 1-2 % aqueous potassium ferrocyanide: Fe(III), Cu(I{), V(V); stannous chloride in 4 M HCl: Pt(IV), Pd(I1); alcoholic solution of 2% dimethylglyoxime: Ni(II), Co(11); 5 % aqueous chromotropic acid: Ti(1V). A conditioning time of 10 minutes was used throughout. All R, values are the average of two replicates and are given for the leading and the trailing edges of the spot. Sample size was approximately 0.0015 ml; the spot diameter was about 9 mm. Where an ion is chromatographed in two valence states, only the lower valence state has been indicated. The solvent ascent was 12.5 cm.
uo@I),
In order to be useful the separations of Zr and Th should include varying ratios of Zr and Th in the presence of foreign ions. This was verified experimentally and the results are given in Table 111. DISCUSSION
Thorium lacks outstanding separation reactions and the real problem in the determination of thorium by colorimetric methods lies in its separation from other elements. This satuhas been achieved very efficiently by using 4 0 x "Os rated with oxalic acid, which separates Th from numerous metal ions including Fe, AI, Ti, UOs, La, Ce(III), and Ce. These are the ions which are most troublesome in the separa-
RESULTS
R, values of individual ions and R/ values of the two ions when present in mixtures are given in Table I. Thirty-six cations were chromatographed in numerous systems containing oxalic acid or phosphoric acid. The results for the two best systems are summarized in Table 11.
Table 11. Separation of Thorium or Zirconium or Hafnium from Numerous Metal Ions Rf values 40% nittic acid saturated with HCI-HIPO4water (10 : 1 :9)b Cations oxalic acid0 0.00 0.00-1.00 Ag(I) 0.81-0.96 0.83-0.94 Pb(I1) 0.82-0.96 0.82-0.94 HdIV 0.00-1 .M) 0.82-1.00 Cu(I1) 0.854.96 0.88-1.00 Bi(II1) 0.90-1.00 0 . 81-0.98 Sb(II1) 0.79-1 .00 0.68-0.87 Sn(I1) 0.83-1.00 0.71-0.87 Sn(IV) 0.75-0.90 0.00-0.85 Pd(I1) 0.81-0.93 0.31-1.00 TU) 0.90-1.00 0.81-1.00 Fe(II1) 0.91-1 .OO 0.88-1.00 AI(I1I) 0.90-1 .00 0.93-1.00 Cr(II1) 0.88-1.00 0.81-1.00 Ni(I1) 0.80-1 .00 0.87-0.96 Co(I1) 0.90-1.00 0.81-0.96 Zn(I1) 0.90-1.00 0.86-1 .00 Ca(1I) 0.93-1.00 0.74-0.87 Ba(I1) 0.98-1.00 0.75-0.87 Sr(I1) 0.87-1.00 0.81-0.93 U(VU 0.87-1 .00 0.87-0.96 V(V) 0.85-1 .00 0.00 Zr(IV) 0.00 0.87- 1. 00 Th(1V) 0.89-1 .00 0.89-1.00 Ce(II1) 0.85-1 .00 0.87-1 .OO Ce(IV) 0.86-1 , 00 0.88-1.00 La(111) 0.89-1.00 0.87-1.00 Y(II1) 0 45-0.69 0.00-1.00 Ti(IV) 0.660.82 0.64-0.91 Mo(V1) 0 . 1 9 4 95 0.00-1.00 WW) 0.89-1 00 0.85-0.95 Selenite 0 83-0,99 0.68-0.87 Tellurite 0.90-1.00 0.00 HfO(I1) 0.79-1 .00 0.86-0.96 MdW 0.89-1 .oO 0.77-1 , 0 Mn(I1) 0.90-1.00 0.86- 1. 00 Cd(I1) I
I
I
I
a
Time of development 65 minutes, temp. 32-33" C . Time of development 55 minutes, temp. 31-32" C.
1330
ANALYTICAL CHEMISTRY
Table 111. Separation of Some Important Mixtures by Paper Chromatography R, values Mixtures Thorium with another Other cation ion (1 :1 ) ~ Th 0.00-0,06 0.87-1.00 Th-La 0 . 00-0. 09 0.86-1. 00 Th-Y 0.00-0.12 0.90-1.00 Th-Hf 0.00-0.25 0.75-1.00 Th-Ce Zirconium with another Other cation Zr ion (1 :l)b 0.88-1.00 0.00-0.31 Zr-La 0.89-1 .OO 0.00 Zr-Cd or Ce or Ni 0,78-0,90 0.00 Zr-Sb 0.81-1,OO 0.00 Zr-Pb or Fe or Cu 0.83-1 . 00 0.00-0.04 Zr-V 0.85-0.95 0.00-0,08 Zr-Th Zirconium and thorium Zr Th in various ratiosb Zr-Th 0.88-1 , 00 0.00-0.05 1:lO 0.87-1 . 00 0.00-0.06 1:100 0.87-1.00 0.00-0.08 10: 1 0.86- 1 . 00 0.00 100:1 Zirconium, thorium, and Th Zr a foreign ion( 1 : 1 : 3)n O.O@O. 08 0.86-0.99 Zr-Th-tellurite 0.00-0.19 0.81-1 .OO Zr-Th-La 0.00-0.18 0.88-1.00 Zr-Th-V Zr-Th-selenite or Y 0.00-0.12 0.88-1 . 00 or U 0.00-0.31 0.90-1.00 Zr-Th-Ce 0.00-0.10 0.88-1.00 Zr-Th-Ce( 111) 0.00-0.19 0.90-1.00 Zr-Th-Ti 0.00 0.88-1.00 Zr-Th-X 0.00 0.94-1 .00 Zr-Th-acetate Other ion Hf Hf with another ion (1 : 1) 0.80-0.90 Hf-UO2 0.00 0,79-0.93 0.00 Hf-V or Ce 0.76-0.94 Hf-Fe 0.00 0,68-0.86 Hf-Ce( 111) 0.00 Ti, Zr, Th Ti Zr Th Fe 0.43-0.70 0.90-1.0 0.00-0.22 Ti-Fe 0.43-0.79 0.90:i .oo a Solvent: 40% nitric acid saturated with oxalic acid, time 65 minutes, temp. 29-30" C * Solvent: HC1-HsPOa-water (10: 1 :9), temp. 34-35" C, time 55 minutes. X = or C103-l or Cr04-* or NO3-' or CI-1 or formate or arsenite.
tion and determination of Th. The same system also separates Zr and Th in the presence of various cations and anions in 65 minutes. Even phosphate and molybdate d o not interfere. This separation is therefore better than that of Balkrishnamurty which takes 9 hours by the ascending method. The oxalic acid system makes possible another difficult separation-Le., Zr-Th-Ti. This procedure is better than that of Almassy who uses three solvents to separate each one of the three cations from the other two. The system HC1-H3P04-water (1O:l :9) by volume is very successful for the separation of Zr from numerous metal ions including Th, V, Cu, Fe, and Ce. These are the ions which cause considerable difficulty in the separation and determination of Zr. The same system is also useful for the separation of Zr and Th in ratios of (1 :100)and (1OO:l). In both solvent systems mineral acid concentrations are high and the pH values are consequently very low. Under these circumstances the concentration of the precipitating ion-].e., oxalate or phosphate-is extremely small, and only Th or Zr is selectively precipitated. The high mineral acid concentrations also ensure that most of the anions of weak acids are present in acid form only, thus decreasing the interferences. Quantitative aspects of these separations were not studied but it is probable that coprecipitation will be small because the separations have been performed under conditions of high solubility. All these separations have been achieved in the presence of interferences which are likely to be encountered in practical analytical work. This study also shows that
selective precipitation reactions can be made almost specific when they are suitably modified and applied on paper. Thus in the usual oxalic acid precipitation of Th the interfering ions are La(III), Ce(III), SO^-*, and P04-3. Similarly, in the phosphoric acid precipitation of zirconium the interfering ions are Ti, Fe, and Th. Of these Fe and Th do not interfere in paper chromatography. The separations now depend not only on differences in solubility but also on differences in the adsorption of the precipitated species on paper. As a result the separations become possible in microgram quantities and are accomplished in a matter of minutes rather than hours. All the separations, which were obtained for zirconium nitrate and described in this report, are also obtained for zirconyl chloride and hafnyl sulfate, which have in all the systems studied the same R, values as those of zirconium nitrate. This is a great advantage because no particular method for the preparation of the original sample solution is necessary. ACKNOWLEDGMENT
The authors are grateful to A. R. Kidwai for research facilities and to A. K. Mukherjee (Drexel Institute of Technology) for the gift of hafnyl sulfate. RECEIVED for review October 24, 1966. Accepted May 18, 1967. Financial assistance to F. Khan was provided by the University Grants Commission.
Determination of Trace Metals in Terephthalic Acid by Ion Exchange Concentration and X-Ray Fluorescence J. G . Bergmann, C. H. Ehrhardt, L. Granatelli, and J. L. Janik Research and Decelopment Department, American Oil Co., Whiting,Ind. THEDETERMINATION of trace metal contaminants in purified organic chemicals has become more difficult as purity requirements have become more rigid. Faced with the need to determine 0 to 2 ppm of Fe, Ni, Cr, Mn, Ca, Co, Ti, and M o in purified terephthalic acid (TA), we decided to adapt the ion exchange concentration-x-ray fluorescence method that has been developed for determining traces of Ni and V in petroleum (1). The TA sample is ashed and the ash is dissolved by sequential acid treatments designed to eliminate traces of silica and to solubilize the metal oxides without losing Cr02C12. All of the Fe, Ni, Co, Mn, Ca, and most of the Cr and Ti are collected on a cation exchange disk. The eluant is evaporated; the remaining C r and Ti and the Mo are converted t o the ammonium salts, collected on an anion exchange disk, and counted by XRF. Total analysis time is about five hours. T o minimize the blank, the ion exchange disks are pretreated with 5 % sodium citrate to remove Ca, and then with 4%HC1 to remove the remaining metals, EXPERIMENTAL
Materials. Metals for calibration were spectroscopically pure iron wire, nickel sponge, cobalt oxide, manganese dioxide, and calcium carbonate. Water was ion exchanged. All (1) J. G. Bergmann, C . H. Ehrhardt, L. Granatelli, and J. L. Janik, ANAL.CHEM., 39, 1258 (1967).
other chemicals were reagent grade. Metal-collection disks were cut from Amberlite SA-2 cation exchange paper and SB-2 anion exchange paper. Standard solutions are prepared as follows: STANDARD IRONSOLUTION,100 ppm. Dissolve exactly 0.100 gram of pure iron wire in a minimum amount of "03 and evaporate to moist salts. Add 1 ml of 7 2 z HCIOl and dilute to 1000 ml. STANDARD NICKELSOLUTION, 100 ppm. Dissolve exactly 0.100 gram of pure nickel sponge in a minimum amount of nitric acid, evaporate to moist salts, and dilute to 1000 ml. STANDARDCHROMIUMSOLUTION,100 ppm. Dissolve exactly 0.769 gram of C T ( N O ~ ) ~ . ~ H inZwater O and dilute to 1000 ml. STANDARD COBALTSOLUTION, 100 ppm. Dissolve exactly 0.127 gram of pure COO in a minimum amount of HCI, evaporate to moist salts, and dilute to 1000 ml. STANDARDMANGANESE SOLUTION,100 ppm. Dissolve exactly 0.158 gram of pure M n 0 2 in a minimum amount of HCl, evaporate to moist salts, and dilute to 1000 ml. STANDARDCALCIUMSOLUTION,100 pprn. Accurately weigh 0.250 gram of pure C a C 0 3 and transfer to a 1-liter flask. Add H N 0 3 slowly and with shaking until the C a C 0 3 just dissolves, and dilute to volume. STANDARD 1.O-ppm SOLUTIONS.To simplify calibration, prepare only two 1.0-ppm solutions from these metals. Freshly prepare the working solutions by transferring 10.0 ml each of any three of the 100-ppm solutions to a I-liter volumetric flask and dilute to volume. Prepare a similar working solution containing the other three metals. VOL. 39, NO. 1 1 , SEPTEMBER 1967
1331
