plier, this fact should become apparent in these stringent circumstances and reBy sult in deviation from linearity. this criterion no measurable amount of stray ultraviolet light \vas found in the system. The somewhat greater standard deviation found for A than for B is attributable largely to the fact that the extremely small photocurrents measured required maximal amplification, resulting in larger values for the dark current. Xerertheless, absorbances approaching 4.0 (referred back to water) have been determined with a fair degree of accuracy. The last point shown on 8.with a n expected absorbance of 1.2, is in serious error. and illustrates the limitation of the instrument, as used. It has not been
included in the calculations. This limitation may be minimized by reduction of the dark current ( 7 ) . SUMMARY
LITERATURE CITED
(1) Baum, IT. A, Dunkelman, L., J . Opt. SOC.Am. 11, 782 (1950). ( 2 ) Buell, 11.V., Hansen, R. E., J . Biol.
Chem., in press.
(31 Buell. M. V.. Hansen. R. E.. Science 126, 842 (1957). (4) Dunkelman, L., J . Opt. SOC.A m . 45, 134 (1955). \
The spectrophotometer described is capable of making accurate measurements of the absorption of ultraviolet light of wave length as short as 194 mp. This technique may be used to characterize compounds by their absorption spectra in the far-ultraviolet, and follow the course of reactions in this region by difference spectra. Absorption spectra in the far-ultraviolet of certain biologically important compounds are presented elsewhere ( 2 ) .
,
(5) Hanovia Chemical and hlfg. Co., 100 Chestnut St., Newark, S . J., Brochure 127. (6) Xlehler, A. H., Bloom, B., Ahrendt, &I. E.. Stetten. De W.. Jr.. Science
RECEIVED for review ;iugust 11, 1958. Accepted December 15, 1958. Investigation supported by a grant h-646(C) from the Sational Institutes of Health.
Spectrophotometric Determination of Palladium with Phenyl-alpha-pyridyl Ketoxime BUDDHADEV SEN Coates Chemical Laboratories, Louisiana State University, Baton Rouge, La.
b Phenyl-a-pyridyl ketoxime was found to b e a highly selective and sensitive reagent for palladium. The palladium complex of the ketoxime has two characteristic absorption maxima, one a t 410 mp and the other a t 340 mp. At these wave lengths the photometric error is a minimum in the concentration ranges 2 to 10 and 1.5 to 8 p.p.m., respectively. The molar extinction coefficient of the palladium chelate is 3 X a t 410 mp and 5 X l o 4 a t 340 mp. The reagent also forms complexes with iron, cobalt, nickel, and copper, which are all soluble in organic solvents. The interference due to these ions can be eliminated b y using (ethylenedinitri1o)tetraacetic acid (EDTA). Gold and cyanide are the only interfering substances.
lo4
P
HEhTL-a-PYRIDTL
lietoxime(1) was
reported (6)as a
fi
A 1
1
n c ~chelating agent for a number of
cations. It was found to be highly selective for palladium(II), with which it reacted to form a yellow, n-ater-insoluble chelate of definite composition. An analysis of the palladium complex
showed that palladium and the reagent combine in the ratio of 1 mole to 2 moles. The various reagents and procedures for the spectrophotometric determination of palladium were discussed by Beamish and RlcBryde (1, 2 ) in t n o excellent review articles. West ( 7 ) and Rlellon and Boltz (4) summarized the reagents used for the spectrophotometric determination of platinum metals during the last few years. The principal difficulty encountered in many of the proposed methods for palladium is the interference of the bivalent ions of the transition metals, particularly the colored ones, and of other metals of the platinum group. LTsually the interference of the metals other than the platinum metals can be eliminated by using EDTA.. Phenyl-a-pyridyl ketoxime formed stable colored complexes n-ith nickel(I1) , cobalt(II), iron(II), copper(II), gold(111), and palladium(I1). All were precipitated a t a definite pH, from the aqueous phase and extracted with organic solvents such as chloroform and carbon tetrachloride. The efficiency of precipitation and extraction for most of the chelates from the aqueous phase was best between pH 6 and 11. Honever, the incipient precipitation started a t a much lower pH. The palladium complex began to precipitate a t pH 2. K i t h the exception of copper(II), the spectral characteristics of the complexes were the same in both aqueous and nonaqueous phases. Figures 1 and 2 shou-
the absorption spectra of a number of complexes in chloroform, and also that of copper in the aqueous phase. The present paper describes the procedure for the spectrophotometric determination of palladium, using phenyl-a-pyridj-1 ketoxime. EXPERIMENTAL
Instruments. Spectral studies and light adsorbance measurements \yere made with a Beckman Model DK I recording spectrophotometer. Matched silica cells of 1-em. light path were used. p H measurements were made with a Beckman RIodel G p H meter. Preparation of Phenyl- a-pyridyl Ketoxime. Hydroxylamine hydrochloride (50 grams) dissolved in a minimum volume of n ater was added t o 200 ml. of a 107, sodium hydroxide solution. T o this were added 20 grams of 2-benzoylpyridine dissolved in 300 nil. of 95% ethyl alcohol. The mixture rras refluxed for 8 hours and evaporated on a water bath until all the solid separated as a crust on the top and the liquid was brownish. Sometimes the oxime separated as an oil which solidified on cooling. The magenta solid was filtered and weighed (yield 30 grams). The crude product was crystallized several times from 95% ethyl alcohol until the oxime was obtained as a white crystalline substance melting at 161' C. Two or three more crops of crystals were obtained from the mother liquor. The combined yield was 70% of the theoretical. The VOL. 31, NO. 5, MAY 1959
881
IO
IO
20
20
30
30
40
40
W V
W V
z a 50 IL v)
5
l-
60
60
2
U
E
50
I-
Z
2
70
I-
80
70
80
90
90
100
100
700
500
400
350
320
300
700
500
400
WAVE LENGTH rnp
Figure 1. Spectral absorption curves of metal phenyl-crpyridyl ketoximates in chloroform Fe++(12 p.p.m.1 Au + + (20 p.p.m.) P d + ( l O p.p.rn.) Reagent in CHC13 5. Reagent in ethyl alcohol
882
ANALYTICAL CHEMISTRY
320
300
Figure 2. Spectral absorption curves of metal phenyl-apyridyl ketoximates in chloroform
1. 2. 3. 4.
analysis of phenyl-a-pyridyl ketoxime was as follows: Found: C. 72.95%: H. 5.10%; h', 14.20%. Calculated:' C; 72.71%; H, 5.08%; N, 14.14%. Standard Solutions. All chemicals were of analytical reagent grade. T h e palladium solution n a s prepared from palladium(I1) chloride and was 4144 n-ith respect to hydrochloric acid. It was standardized gravimetrically by precipitating palladium with dimethylglyoxime. The solutions of other cations were prepared from their sulfates, nitrates, and chlorides and estimated by standard procedures. The buffers were prepared according to Vogel (6). Reagent Solution. A 1% solution of phenyl-a-pyridyl lietoxime in 95% ethyl alcohol was used a s the reagent. Adsorption Spectra of Metal Phenyl-a-pyridyl Ketoximates. Preliminary studies indicated t h a t for most of t h e cations, chelate formation and extraction of the chelates by organic solvents were most favorable between pH 6 and 11. Solutions for the study of the spectral adsorbance were prepared as follows. Depending upon the intensity of adsorption by the complex, 100 to 600 y of the cation (4to 24 p.p.m. in the final solution measured) were treated Tyith 2 ml. of a 1% solution of the reagent in a 50-ml. beaker and diluted with water to about 15 nil. The p H of the solution was adjusted between' 8.5 and 10 with sodium carbonate solution and recorded with a p H meter. The solution was then transferred to a 50-ml. separatory funnel using a mini-
350
WAVE LENGTH rnp
mum volume of water to wash the beaker. The volume of the aqueous phase was maintained at about 20 ml., and 10 ml. of chloroform was added to it. The two phases were mixed by vigorous shaking and allowed to separate. Two extractions were performed. The volume of the chloroform extract was made up to 25 ml. with chloroform. The aqueous phase was evaporated almost to dryness and tested for the particular cation by applying the appropriate spot test. The aqueous phase was free of the cation being extracted after the second extraction. The solution for the study of adsorption by copper(I1) complex in water was made as above; the volume, however, was made to 25 ml. with n-ater. Figures 1 and 2 show the spectral curves and Table I summarizes wave lengths a t the peaks. A chloroform blank was prepared by diluting 2 ml. of the reagent solution to 20 ml. with water, adjusting the pH to 9, and extracting twice with chloroform. The final volume was 25 ml. with the chloroform. An aqueous blank was prepared by diluting 2 ml. of the reagent to 20 ml. with water, adjusting the p H to 9, and diluting to 25 ml. with water. Absorption studies were made against proper blank solutions. Figure 1 shon s the spectral characteristic of the reagent in chloroform and in ethyl alcohol os. chloroform and ethyl alcohol, respectively. Influence of pH on Chelate Formation and Extraction. The influence of
1.
Cu++(24 p.p.rn,)
2. 3. 4.
Cu++ in HsO ( 2 4 p.p.m.1 co"(4.8 p.p.m.) Ni + + ( 4 p.p.m.)
p H on the formation and extraction of palladium phenyl-a-pyridyl ketoximate was studied. i Concentration of palladium (10 p.p.m.) and the reagent and t h e procedure adopted for preparing solutions were the same as in the spectral absorption study. I n one series of experiments, the pH n-as adjusted by the addition of different volumes of sodium carbonate solution (Figure 3). I n another series, after p H had been adjusted to the desired value, it was stabilized by the addition of 2 ml. of a buffer in order to determine the most desirable buffer system. Acetate, phosphate, borate, ammonia-ammonium chloride, and alkali carbonate buffers were used. Only the phosphate interfered seriously a ith the absorption peak in the ultraviolet region; in the visible region it was unaffected (Figure 3). Efficiency of Extraction. A number of solutions of each cation (500 y) in 20 ml. were prepared and adjusted t o a different p H b e t m e n 2 and 11. Each solution n.as treated with 2 ml. of t h e reagent solution and extracted one t o three times n i t h chloroform. The aqueous phases were evaporated almost to dryness and tested for the cations. Aqueous phases were free of palladium(II), iron(II), cobalt(II), and nickel(II), after the first extraction. Gold(II1) and copper(I1) were completely removed by two extractions in the above pH range. This was further
10
Table 1. W a v e Lengths of Light Absorption Peaks of a Number of M e t a l Phenyl-a-pyridyl Ketoximates in Chloroform
20
Absorp-
30
tion
Rapid Increase
M y
Absorp-
Peaks,
40 W
u
50 I-b
60
z
U
E
70 80 90
Cation Fe++
Wave Length 530
Au++ CU'+
450 475,350
in
tion after 460 mp
Figure
and Curve
KO.
Fig. 1, curve 1 n
L
Fig. 2, curve 1
Cu (in H?O) 415 co++ h'i+' Pd++
390mp 600mp 550mp 410, 317
Reagent (in CHC13) Reagent
360mp
4
340mp
5
(in
2 3 4
Fig. 1, curve 3
CzH60H) 100
600
500
410
380
340
310 300
WAVE LENGTH m y Figure 3. Spectral curves of palladium phenyl-a-pyridyl ketoximate in chloroform extracted from aqueous phase of different p H 1. 2. 3. 4.
pH 1.7 pH5.0 pH 11.5 pH6.4
verified by measuring the absorbance of the chloroform phase after each extraction. Beer's Law. The spectral curves show t h a t palladium phenyl-a-pyridyl ketoximate exhibits two distinct and characteristic absorption peaks a t 410 and 327 mp. Beer's law n-as studied a t both these wave lengths in the concentration range of 0.5 to 1.I p.p.m. using a cell of 1-em. light path. One milliliter of the reagent solution was added to the palladium solution contained in a 25-ml. beaker. The volume of the mixture was made up with water to nearly 10 ml. and the pI-1 of the solution was adjusted between S and 9 with a sodium carbonate solution. The solution was then transferred to a 50-ml. separatory funnel using a minimum volume of water to wash the beaker. The solution was extracted three times, using 5 ml. of chloroform each time and the final volume of the extract was made up with chloroform, to 25 ml. Absorbance of the chloroform extract was measured a t 410 and 340 mp. The concentration range for minimum photometric error was between 1.5 and 8 p.p.m. a t 340 mp and between 2 and 10 p.p.m. a t 410 mp. However, this range can be varied by changing the length of the light path through the solution. The average molar extinction coefficient of the chelate was found to
5. 6.
pH 9.3 pH 10.8 7. pH 7.5 Phosphate buffer
be 3 X lo4 a t 410 mp and 5 X lo4 a t 340 my. Effect of Other Ions. Binary mixtures were prepared containing 100 y of palladium and 5 mg. of a second cation. The mixture was treated with 2 ml. of the reagent solution and was diluted with water to approximately 10 ml. The extraction was performed three times using 5 ml. of chloroform each time. The combined extracts were diluted to 25 ml. with chloroform. Absorbance was measured a t both 410 and 340 mp. Beryllium, magnesium, strontium, barium, zinc, cadmium, mercury(II), aluminum, lead, bismuth(111), vanadyl(II), manganese(II), iron (111), ruthenium (VI) rhodium (III), osmium(VI), iridium (N) , platinum(V) , and silver ions were not extracted and hence did not interfere. Interference due to copper(II), iron(II), cobalt(II), and nickel(I1) was eliminated by using 2 ml. of 0.1M EDTA solution. EDTA did not interfere when absorbance was measured a t 410 mp. Interference by EDTA when absorbance was measured a t 340 mp was eliminated by treating the blank with a n identical amount of EDTA. Bismuth(II1) and iron(II1) tended to hydrolyze and the hydroxides usually separated at the interface of the two phases. The suspensions were eliminated by filtration. Gold(II1) was the only seriously interfering cation. Among the common complexing agents such as citrate, oxalate, tartrate, and ~
cyanide, only cyanide interfered seriously. Effect of Reagent, Temperature, and Time. At the optimum p H even a 1000 gold excess of reagent was without any adverse effect a t 410 m l . At 340 mp, the absorbance was higher because of the absorption of the reagent. This was completely eliminated by treating the blarik similarly. Variation in temperature of the chloroform extract between 0" and 45" C. did not affect the color intensity Reaction and maximum color development were instantaneous. There was no fading of color in 2 weeks.
DISCUSSION
Phenyl-a-pyridyl ketosime n-as found to be a sensitive and selective reagent for palladium. The procedure for the spectrophotometric determination of palladium is not rigorous. It is possible to separate palladium, using phenyla-pyridyl ketosime, from a number of elements with which palladium is usually associated. The curves in Figure 3 indicate that betn een pH 6.4 and 10.8 the structure of the chelate formed is the same. Preliminary spectrophotometric studies (3) and analysis of the precipitate indicated that palladium and the reagent combined in the ratio of 1 to 2 moles. However, these studies did not reveal whether the palladium complex exists as structure I1 or 111. The results of the investigation on the likely structure of the palladium chelate and other analytical applications of phenyl-a-pyridyl ketoxime will be published later. VOL. 31, NO. 5 , M A Y 1959
883
Pd/2
N
O=N-Pd/P
0 ‘’
I1
Philip K.West for allo\Ting the author to use the facilities of his laboratories. Partial financial support by the Division of Research Grants, Public Health Service, is thankfully acknowledged. Reilly Tar and Chemical Corp. made a gift of 1 pound of 2-benzoylpyridine.
11928’1.
(4)’ hlellon, AI. G., Boltz, 11. F., ANAL CHEW30, 554 (1958). (5) Sen, B., Chern. & Znd. (London) 1958, cco
dUY.
(6) Vogel, A. I., “Text-Book of Quantita-
LITERATURE CITED
tive Inorganic Analysis,” Longmans, Green & Co., London, 1944. ( 7 ) West, P. W.,ANAL.CHEM.30, 748 (1058).
(1) Beamish, F. E., McBrS.de, IT. A. E., Anal. Chim. Acta 9 , 349 (1953).
RECEIVEDfor review July 30, 1968. .4ccepted December 11, 1958.
I11
ACKNOWLEDGMENT
Grateful thanks are expressed to
(2) Zbid., 18, 55 (1958). (3) Job, P., -4nn. chim. (Paris) 9, 113
Tracer Investigation of Iridium- Plati num Se pa ration by Anion Exchange D. D. BUSCH,’
J. M. PROSPERO, and R. A. NAUMANN
Frick Chemical and Palmer Physical laboratories, Princeton University, Princeton, N. J.
b The separation of 1 0-mg. quantities of iridium and platinum by the use of Dowex A-1 anion resin has been investigated using radiotracer techniques. Elution curves are exhibited for iridium(lll), platinum(ll), and platinum(lV) recovery from resin COIumns, and a procedure is reported for the rapid separation of iridium and platinum.
100 80
-
100
80
-
30
-
20
-
> k
>
c
l0-
2
T
separation of platinum and iridium by ion exchange has been the subject of a number of investigations (2-7). Procedures involving anion exchange have been particularly appealing because of the extremely stable halide complexes formed by these elements. Hon ever, these methods nere limited in that large volumes of eluent (and consequently long elution times) were required: in particular, anionic platinum complexes w r e bonded by the resin to such an extent that quantitative elutions n ere difficult to obtain. I n a recent article, Berman and RIcBryde ( 1 ) have reported the successful recovery of fractional milligram quantities of platinum as chloroplatinate ion from Amberlite IRA-400 anion exchange resin, using perchloric acid as eluent. Independent work in this laboratory using radiotracer techniques to investigate the separation of 10-mg. quantities of iridium and platinum with Dowex -4-1 anion exchange resin has yielded similar results. 4 procedure has been developed permitting the rapid HE
Present address, Department of Chemical Engineering, University of California, Berkeley, Calif. 884
ANALYTICAL CHEMISTRY
-
60 -
60 50 40
-
i0 8-
8 :
u
5-
>
I
4-
3-
g-
4 -
f .
1, 0
-
6 -
6-
I
I
/
175%
6 9 4 % ELUTION f
I
I
l
5
l
1
IO
MILLILITERS OF ELUENT (503 HCIO4)
MILLILITERS OF ELUENT
Figure 1. Elution curve of chloroplatinate with 50% perchloric acid
Figure 2. Elution curve of mixed chloroplatinite and chloroplatinate
Platinum, 10 mg.
Platinum, 10 mg.
separation of radiochemically pure platinum from iridium.
thermostatted reservoir through jacketed columns. The cyclotron-produced activities 2.7-day platinum-191 and 4.3-day platinum-193m served as the platinum tracer and pile-producpd 75day iridium-192 as the iridium tracer. As 63 to 66 k.e.v. x-ray and harder gamma radiations are associated with these tracers, the activities of all solutions and of the resin itself were directly counted with a sodium iodide scintillation spectrometer set to record quanta with an energy greater than 45 k.e.v. The fractions to be counted were of standard volume and were contained in standard size test tubes mounted in standard geometry with respect to the detector.
APPARATUS A N D REAGENTS
Dowex 1 resin, 200 to 400 mesh size, 10% cross-linked (Dow Chemical Co.,
Midland, Mich.), was employed in columns 5 mm. in diameter by 5 cm. in length, with column free volume of 0.75 ml. Before use, the columns mere mashed with 5 column volumes of 9.1hydrochloric acid, followed by 5 column volumes of water. All elutions were carried out a t 80’ C., a t a rate of 1 ml. per minute. Temperature was controlled by circulating water from a
