to Sc(II1) is used are presented in Table 111. The separation of Sc(II1) from large amounts of foreign metals is quantitative. La(”, Sm(III), and Lu(1II) are selected as typical rare earth metals. I3ecause Y(III), rare earths(III), Zr(IV), and Th(1V) do not shoiv noticeable adsorption on the anion exchanger from HC1 media, the procedure described seems to be useful for their separation. In addition, rare earths(III), T h ( I V ) ,and U(T’1) fractions are recovered sharlily from the column without any troublesome tailing. However, when large amounts of Th(IV), Zr(IV), or TJ(V1) (or all of them) are present, the Sc(II1) breakthrough tends to appear in a slightly earlier fraction of effluent, but the separation is still good.
The flow rate of eluent does not affect the chromatographic separation of Sc (111); a range of the flow rate from 0.5 to 1.5 ml. per minute can be used without disturbing the elution patterns of Sc(111) or the other metals. .imong the metals tested qualitatively, Ta(V) and W(V1) showed strong adsorption on the resin from the sulfate media, 0.1M (SHa)zS04-0.025Jl HzS04, while Fe(1lI) a,nd In(II1) were weakly adsorbed. h t the lower acid concentration employed here [0.025.11 HzS04-0.1V(XH&SOa], Ti( IV) is seriously hydrolyzed, partly passing through the column and partly being adsorbed strongly on the resin. Ti(IV) does not show any adsorption from the 0.5X HzSO~-O.lM (NH&SO4 medium.
LITERATURE CITED
(1) Bunney, I,. R., Ballou, S . E., Pascual, J., Foti, 8., ASAI,. CHEM.31, 324 (1959). ( 2 ) Hamaguchi, H., Kishi, RI., Onuma, X., Kuroda, R., Talanta 11, 495 (1964). (3) Harnaguchi, H., Kurodn, R., Aoki, K., Sugisita, It., Onurna, N., Ibid., IO, 153 (1963). (4) Kagle, R. A , , Murthy, T. K. S., Analyst 84, 37 (1959). ( 5 ) Sekine, T., Saito, S . ,A-ature 181, 1464 (1958). (6) Stevenson, P. C., Nervik, W. E., C. S. At,. Ehergy Cornm. K e p t . NAS-NS 3020, 1961. ( 7 ) Vickery, R. C., “The Chemistry of Yttrium and Scandium,” Perganion Press, Sew York, 1960. RECEIVEDfor review April 3, 1964. Accepted June 16, 1964.
Cation Exchange Separation of Titanium and Zirconium Using Perchloric Acid Application to the Analysis of PZT Ceramics R. G. DOSCH and F. J. CONRAD Organization 1 1
74,Sandia laboratory, Albuquerque, N. M .
b A
description is given of a cation exchange procedure for a quantitative separation of zirconium and titanium from lead and bismuth, and from each other, using HCIOd-NHJ mixtures and HClOd as eluents. This procedure has been applied to the analysis of highfired lead oxide-zirconium oxidetitanium oxide ceramic material (PZT).
H E USEFUL PIEZOCLIXTRIC propT e r r i e s of ceramic 13aTi03 has stimulated a search for other ferroelectrics suitable for fabricating piezoelectric ceramics. Shirane, Suzuki, and Takeda ( 1 4 ) and others (13) described the properties of solid solutions of PbTiOs and I’b%r03 and Jaffe, Koth, and Narzullo (6! 2 studied piezoelectric propertirs of lead zirconatelead titanate (WT) solid-solution ceramics. Small com1)ositional changes can produce marked changes in final piezoelectric properties. These variations can be minimized by close control of the constituents a t all stages of processing. To date, the composition of 1’ZT ceramics usually are determined empirically by memuring their physical and electrical 1)roperties. Knowledge of the composition during processing, and particularly in the finished ccramic, could be more easily obtained if the dis-
2306
ANALYTICAL CHEMISTRY
solution procedure and the method for separating Ti and Zr from the resulting solution, and from each other, were simplified. Brown and Rieman (3) used ion exchange techniques employing a n HCIcit,ric acid eluent and I3elya Alimarin, and Kolosova ( 2 ) used 1X HCl in an ion exchange procedure for the separation of Ti from Zr. However, because either relatively large effluent volumes are required or pH control is critical, use of these methods has been limited. Other investigators ( I f , 12) have ut,ilized anion exchange techniques and HzS04-Sa2S04and HCI media, respectively, to separate Zr from a number of elements, including Ti. Use of the chloride ion i F undesirable in cases where cations form either insoluble chlorides or unknown complexes. Thus, many current ion eschange procedures suffer from one of three disadvantages: use of relatively large effluent volumes, critical control of eluent pH, or the presence of an undesirable anion. I n a n investigation of the behavior of titanium( IV) in HClO,, 13eukenkamp and Herrington ( 2 ) reported the possibility of titanium(1V’i-perchlorate complexes forming at HC104 concentrations above 2.11. From their work in separating vanadium from titanium and other cations, Fritz and Abbink
(6) proposed that Ti(1V) could be separated from other tri- or quadrivalent metal ions using HC1O4-H2O2 eluent systems. This paper describes a cation exchange procedure for a quantitative separation of zirconium and titanium from each other, using an HCIOd eluent system. The method as applied to the analysis of PZT ceramics is briefly described. Solutions containing titanium and zirconium in amounts between 10 to 100 mg. of each element have been separated from each other. EXPERIMENTAL
Column Preparation. The borosilicate glass ion exchange columns were 30 X 1.0 cm. (i.d.) a n d 10 x 1.0 cm. (i.d.). The resin used was I>owex AG-5OWX8, 100- to 200-mesh, hydrogen form. Twenty-seven grams of resin, d r y weight, were used in t,he larger columns, and 9 grams, d r y weight, were used in the smaller columns. The resin was washed several times with distilled water, and the finest particles were decanted after each washing. The resin then was washed with 1.1- IICI, followed by distilled water until a negative chloride test was obtained. Reagents. TITANIUMELUENT. ; isolution of 257, HCIO, was prepared by diluting 178 ml. of 707, HC104 to 500 ml. with distilled water.
MLS. E F F L U E X T
Figure 1 . Elution curves of titanium and zirconium
P Z T . Samples of standard PZT material were dissolved by fuming with 727, HCIOa and taken to a volume of 3 to 4 ml. The solution was diluted to 100 ml. with distilled water, several drops of 30% H202were added, and the resulting solution was placed on the 10 x 1.0-cm. (id.) columns, allowing a flow rate of 2 ml. per minute. The zirconium and titanium was eluted using the HC104-NH,F eluent at a flow rate of 2 ml. per minute. The effluent solution of zirconium and titanium was fumed with H2S04 to remove the fluoride ion, and then treated in the same manner as described for titanium and zirconium in the preceding paragraph. TREATMENT OF SAMPLES. The preparation of P Z T samples was identical to the method described for preparation of the standard PZT material in the previous section. RECOVERY OF
TIT.kKIU>f
AND
ZIR-
Sulfuric acid solutions of Figure 2. Elution curve for titanium 1 % HClO4-O.lM "IF:, flow r a t e 1 ml. p e r zirconium and titanium were put on the minute 25% HCIO4, flow r a t e 1 ml. p e r minute 30 x 1.0-cm. (i.d.) columns a t a flow rate of 1 ml. per minute. The titanium was eluted using 257, HC1O4 as the T I T A K I U h I AXD 2;IRCONIChI ELUENT. eluent, and the zirconium was subseeffluent collected in the separation of quently eluted, using the HClO,-r\",F 14 ml. of 7Oy, HCliO, and 3.7 grams of lead and bismuth from titanium and A.R. ammonium fluoride were diluted eluent. At a flow rate of 1 ml. per zirconium. Removal of the fluoride reminute, 125 ml. of 25y0 HC10, was to 1000 ml. with distilled water. sults if the effluent is fumed with H2S04 CUFFERROX SOLUTION.Six grams of required to elute the titanium and 60 until the H2S04wet,s the entire inside of ml. of HC104-SH,F eluent were used cupferron were dissolved in 100 ml. the flask (5). The presence of the of distilled water. The solution was to elute the zirconium. The titanium peroxide complex of titanium is not filtered through Wh.atman KO.42 filter and zirconium were analyzed gravimetrically using cupferron as precipipaper before use. essential to the separation of zirconium STANDARD TITAN [UM AND ZIRCOKIUM tant (9, 10). and titanium. However, because of t'he SOLUTIOKS. Spectropure titanium ELUTIONCCRVES. Elution curves of color of the complex, it was possible to metal was dissolved in concentrated titanium and zirconium were determined observe the elution of titanium from the hydrochloric acid under reflux condiwith separate columns for the percolumn. I n several cases, a yellow tions. Zirconium oxychloride was dischloric acid-ammonium fluoride eluent band was observed a t the top of the solved in 5% HCIOa solution, and the system, using standard techniques. column after the titanium had been The columns used were 30 X 1.0 em. solution was standardized gravimeteluted from the t'op of the column. ( i d . ) , and the flow rate was 1 ml. per rically using cupferron (9,IO^: This band formed when the sulfuric acid STAND.4RD P Z T AfATERIAL. PZT minute. Each titanium fraction and material ureDared from A.R. grade lead each zirconium fraction was analyzed concentration vias lower than 1% while oxide, biim;th oxide, titanium dioxide, by a gravimetric procedure using cupputting the zirconium-t,itan,ium solution a n d zirconium dioside. The mixture ferron as the precipitant ( 9 , 10). The on the column. Failure to elute the was ball-milled, dried, and calcined at elution curve of titanium, using 25% yellow band caused a low recovery of 900' C. This material was passed HC104 as the eluent, was obtained by tit'anium. A quantitative recovery of through a S o . 325 I3ureau of Standards the same procedure as described above. titanium was achieved by eluting Table I consieve before being used. alternately with 5 ml. of 2 5 7 , pertains the per cent compositions of the RESULTS AND DISCUSSION chloric acid and 5 nil. of distilled water PZT material used. The titanium until the band was no longer visible. The elution curves for titanium and and zirconium contents of the standards were determined by neutron activation The analytical results from a series of zirconium, using the 1% HClO4-0.1-1f analysis in order to check their validity determinations made on mixtures of Y H 4 F eluent, are shown in Figure 1. as standards. zirconium and titanium are tabulated Figure 2 illustrates the elution curve of Procedure. TRE:ATMEKT O F STAXDin Tables I1 and 111. Table I1 contains titanium using 25% HC10, as the ARDS. Ti-Zr. Known mixtures of data obtained from mixtures prepared eluent. Since the titanium-fluoride titanium a n d zirconium prepared from from standard zirconium and t,itaniuni complex is not retained by the resin, t h e s t a n d a r d solutions were evapsolutions. while Table I11 contains d a t a fluoride must be removed from the o r a t e d to near dryness in t h e presence of hydrogen persoside. T e n milliliters of concentrated sulfuric acid were added, a n d the resulting mixture was evaporated to a, volume of 1 to 2 ml. Table 1. Weight Per Cent Composition of PZT Standards After dilution to 1 to 27, &So4 concenPZT tration, several drops of 30y0 hydrogen Standard PbO BilOp ZrG Ti02 KzCC)Sa peroxide were added, and the solution was boiled to reniove excess hydrogen 1 67 56 1.43 19.56 11.24 0.21 peroxide to prevent possible resin 2 66.65 1.41 23.67 8.27 ... 3 67.70 1.43 19.60 11 27 ... damage and format,ion of anomalous 4 68.79 1.45 15.03 14.62 0.11 complexes. The titanium and zirconium were separated by a n ion The potassium present in PZT Standards 1 and 4 accompanies the Ti and Zr, and exchange procedure described in the does not interfere with subsequent analysis. The analysis of potassium was done by emission spectrography. section on "Recovery of Titanium and Zirconium.'' CONIUM.
0
VOL 36, NO. 12, NOVEMBER 1964
2307
Table II.
Analysis of Ti-Zr Mixtures
TiO2, nig.
-
‘raker1
Found
43 . 9 69.0
45.7 69 2 95.5 8% 9 60 1 97 1
96.G 82.8 6!$.0 06. G ...
Average ri Liverage c;
I? e c o very 99.6 100.3 98.9 100 1 100 1 100 5
... ... Recovery of Ti02 = 99.9% I{ec,overy of Z r O p = 99.9”,
Table 111.
I’ZT
ZrOz, my.
~-
7;
Taken
Found
.___
53
Recovery
...
...
...
40.1 40 1
40.0 40 1 40 60 1 59 9 40 1
99.8
40.1 60 60
40.1
100
99.8
100.2 99.8 100
Analyses of Standard PZT Material ~~~~~
etantlard
Ti09
content
1Iean error
content
1
1 1 12c,
-0 12
10 ,i7ic
ZrOz
re>ulting from the analysis of several PZT niaterialh. Table 111 also includes data on the I’%T material obtained by ncutron acti\xtion anal! elcnicnth present in I’ZT, except potayiiiun, arc left on thc cation colunin after the first ion exchange separation. ‘rhwe can 1,e v p a r a t e d or striliprd from the column n-ith HXOa (4)and determined using qtandard methods:. The m d i o d described has been sueces>fiilly applied to the analysis of a srrie. of PZT materials which \yere calcined a t 900”C. and hot-pressed at teniiieratures ranging froin 1050’ to 1300” C. The ap1,lication of this method to our specific problem war-
1Iean error
By activation analysis.. Ti02
content
ZrO,
content
ranted no concern with iinliurities However! additional ~ ~ o r isk being carried out a t the present time to determine the ion cwhangr. behavior of a large number of cations using H U 0 4 solutions as eluents. l’rocedures arc being developed for the prel)aration of si)ectrol)ure coniiiounds of several ?lements, including He and %r, using ion exchange techniques with HC104eluent? to separate the major conbtituent from its iiiiliurities. ACKNOWLEDGMENT
The author< exi)r+s their appreciation to 13. 1’.Kenna, Sandla Corpora-
tion, for performing the neutron activation analysis work and allowing us to use his results; and to G. H. Haertling, Sandia Corlioration. for preparing l’ZT material for use as standards. LITERATURE CITED
(1) Belyavskaya, T. .%., Aliniarin, I. P., Kolosova, I. F., Zhur. -Anal. Khint. 13, 668-70 (1958). ( 2 ) Beukenkamp, J., Herrington, K. I)., J . .lm. C‘heni. Soc. 82, 3025 (1960). (3) Brown, tj‘. E.>Rienian, IT.,I b i d . , 74, 1279 (195%). (4) Buchanan, R . F., Faris, J. P., Orlandine, K. A , > Hughes, J., C‘SAEC, TID7560, l7Y-SS (1058j. ( , 5 ) Fritz, J. S., -ibbink, J. E.$ ANAL. CHEM.34, 1OSO (1962). (6) Jaffe, H., Koth, R., Marzullo, S., J . 9 p p Z . I’hys. 25, 809 (1954). ( i )Jaffe, B., Roth, R., Marzullo, S., J . R e s . .Yut. Bur. Std. 55, 239 (1955). (8) Kenna, B. T., Conrad, F. J., ANAL. CHEM. 35, 1255 (1063). ( 9 ) Kolthoff, I. AI., Elving, P. J., ed., “Treatise on Analytical Chemistry,” Part 11, Vol. 5, Section .I by E, It. Schaffer, pp, 29-30, Interscience, New York, 1962. (10) Kolthoff, I. l l . j Elving, P. J., ed., “Treatise on Analytical Chemistry,” Part If, Vol. 5, Section . Iby R. B. Hahn, p. 95, Interscience, Yew Tork,
1962.
(11) Korkisch, J., Farag, A., 2. iinal. Cheni. 166 (a),81-8 (1959). (12) Ilichaelis, C., Eveslage, S., Coulter, P.,Fortman, J., AKAI,.CHEM.34, 1764 (1962). (13) Shirane, G., Suzuki, K., J . Phys. Soc. Jupczn 7, 333 (1952).
(14) Shirane, G.! Suzuki, K . , Takeda, Ibid., 7, 12 (1952). RECEIVEDfor review June 8, 1964. Accepted September 1, 1964. Supported by the ynited States Atoniic Energy Commission. Reproduction in whole or in part is permitted for any purpose of the I‘.S. Government..
Anion Exchange Separation of Gallium, Indium, and AI uminum JOHANN KORKISCH and
ISIDOR HAZAN
Analytical Institute, Universify o f Vienna, / X , Wahringerstrasse 38, Austria Two methods are described for the chromatographic separation of aluminurn, gallium, and indium on the strongly basic anion exchange resin Dowex 1, X8. Media consisting of 90 and 7Oy0 2-methoxyethanol-l10% 6N and 30% 2 N hydrochloric acid, respectively, followed b y 1 N hydrochloric acid are used in one case to effect the sequential elution of aluminum, gallium, and indium in this order. By means of the other method, gallium i s first eluted with 80% acetone-20% 3 N hydrochloric acid whereafter 90% acetone-] 0% 6 N hydrochloric acid i s used to remove
2308 *
ANALYTICAL CHEMISTRY
indium from the resin. Subsequently aluminum i s desorbed b y washing the resin with 7070 acetone-30% 2N hydrochloric acid. By means of these methods the quantitative separation of milligram quantities of these elements can be effected even when aluminum is present in great excess. Based on these facts successful analyses of samples of aluminum metal for gallium and indium have been performed.
o
ISTL~TIGATIOSS concerning the anion exchange separation of aluniinutn. galliuni. and indiiini in niixcd aqiieoui and nonaqueous solventa
containing hydrochloric acid have been carried out in the past. In purc aqueous hydrochloric acid solution.;, however: a sel)aration method ha,s been iised by Kraiis, Selson, and Smith (8) for the sequentiaJ flution of aluminuni, indium. and galliimi in this order using hydrochloric acid of varying cwncentration. This method has heen modified and Jefffry (a) for the se1)aration of gallium in the prcwnct of a large excess of aluminum and was applied t o the analysis of 13ayr liquors. .I railid method was de.-cribcd 11y S a d e zhinn (S) Tvhich i; euital~lefor isolating siiinll nnioiints of galliiuii from con-