of the spectrophotometric procedure is 0.0032 pg. per sq. em. for log 10/l = 0.001, and the fluorometric procedure with DAN is over 10 times as sensitive as the spectrophotometric procedure. Although DAN is over twice as sensitive as 3,3-diamin~benzidene it is more susceptible t o interferences. The direct determination of selenium in copper with 3,3-dianiinobenzidene has been reported ( 3 ) . l'his mas not possible with DAN. I n the determination of selenium in biological material bcth reagents were susceptible to some interference which was probably caused by traces of hypochlorite present aftcr perchloric acid
digestion. Experiments are under may t o develop a rapid procedure for the determination of selenium in biological samples. ACKNOWLEDGMENT
The authors express their thanks to K. L. Cheng for his suggestions on this work. LITERATURE CITED
(1) Broad, W. C., Barnard, A. J., Jr., Chemist-Analyst 50, 124 (1961). (2) Cheng, K. L., ANAL.CHEM.28, 1738 (1956). (3) Cheng, K. L., Chemist-Analyst 45, 67 (1956).
( 4 ) Kolthoff I. M., Elying, P. J., Treatise on Analytkal Chemistry, Part 11, Vol. 7, Interscience, Xew York, 1961. (5) Parker, C. A., Harvey, L. G., Analyst 87.558 (19621. (6) Schumann, H., Hoelling, JV.,2. Chenz. 1, 371 (1961). (7) Schwarz, K., Foltz, C. XI., J . Am. Chem. SOC.79, 3292 (1957). ( 8 ) Watkinson, J. H., AXAL.CHEM.32. 981 (1960). RECEIVED for review Sovember 29, 1962. Accepted June 5, 1963. Presented a t the Pittsburgh Conference on Analytical
Chemistry and Applied Spectroscopy, March 1963. This investigation was supported by research grant GM 09792 from the Sational Institutes of Health, U. S. Public Health Service. One of the authors, G. M., received support through a Sational Science Foundation Fellowship.
Zinc Complexing Properties with Dialkylphosphorodithioic Acids THOMAS H. HANDLEY Analytical Chemistry D,;vision, Oak Ridge National laboratory, Oak Ridge, Tenn.
RAQUEL H. ZUCAL' cind JOHN A. DEAN Department of Chemistry, University o f Tennessee, Knoxville, Tenn.
The composition of the zinc dialkylphosphorodithioate complex which partitions between aqueous acid solutions and CC14 has been established as [(RO)ZPSS]2Zn. Chara zteristic constants have been evaluated; these include the over-all extracticsn constant, the partition coefficient of the zinc complexes, and the stab lity constant of the zinc di-n-butyl and diisobutyl esters. The extractioq mechanism is discussed briefly and the characteristic constants obtained for zinc are compared with those reported for dialkylphosphate complex of zinc.
D
acids have been screened for their extraction of metal ions ( 3 ) from an acid media. and the behavior of several individual esters of dialkylpho*phorodithioic :tcids have been reported ( 8 ) . The manner in w h c h extractablc species form with this class of reagents, and a quantitative evaluation of the extraction characteristics of the zinc complexes were the obje1:t.s of this study. Few pure sulfur chelate complexes have been fully investigated. The dialkylphosphor odithioic acids are known to be quite strong acids (8 and Table I). Among the dialkyl IALKYLPHOSPHORODITHIOIC
1 Present addreos, Co nision Nacional Energia Atomica, Ruenos Aires, ;irgentina.
esters there is only a small change in the acid dissociation constant, but a very large change in the partition coefficients. There is no evidence for association of the reagent in the organic phase (8). EXPERIMENTAL
Dialkylphosphorodithioic Acids. Di-n-butyl phosphorodithioic acid was obtained from Victor Chemical Co., Chicago, Ill. As received, the reagent was shown t o be 99.6% pure by electrometric titration. T h e ammonium salt of the diethyl ester was obtained from t h e Monsanto Chemical Co., St. Louis, Mo. Other dialkyl esters were obtained from Lubrizol Corp., Cleveland, Ohio. Esters which required purification were converted t o the ammonium salts, followed by extraction of impurities with benzene or CC14 and reformation of the acid with addition of mineral acid to the
l a ble 1.
aqueous layer. The free acid may be extracted with CCla and the CCla removed by warming under reduced pressure. The ammonium salts may be recrystallized from ethyl acetate. Because the free phosphorodithioic acids are somewhat sensitive t o hydrolysis, solutions should be prepared fresh each day, or the free acids converted to alkali or ammonium salts. Determination of Metal Distribution Ratio. T h e distribution ratio, D = - -If lo was measured either by use
'
1 M 1'
of radiotracers with carriers (3') or by titrimetric EDTA methods. Except in strongly acidic solutions, the ionic strength was maintained at 1.0. Extractions were performed with a Burrell wrist-action shaker by shaking equal volumes (10 mi.) of the organic and aqueous phases for 10 minutes at 25' & 2' C. in separatory funnels of conventional design. Preliminitry ex-
Characteristic Constants for Zinc Dialkylphosphorodithioates
System: CCL-aqueous acid medium Ester log K* log KcPc lOgP, pKm 10gPc log Kc Diethyl ca. -10 ca. - 9 . 3 0.45 -0.10 ... Di-n-butyl 1.22 6.57 2.52 0.22 2:7t 3.81 Diisobutyl 1.29 5.77 2.63 0.10 2.70" 4.00 Di-sec-butyl 1.37 ... ... ... 2.77" ... Di-( 2-ethy1)hexyl 1.40 ... ... ... 2 86" ... a Estimated from extrapolation parallel to the complete distribution ciirve for di-nbutyl ester on logarithmic coordinates. ~~~
~~~~~~
VOL. 35, NO. 9, AUGUST 1963
~
1163
0
-
t
.!
a Z
-3m + z -
SolvP,ntSyrtem .CClq
I
-
H20
A Hexnne- H20
-I
0
where K , is the acid dissociation constant and P,, the partition coefficient of the reagent (Table I and 8). Since for zinc the coordination number is twice the charge, the appropriate number of bidentate ligands will both neutralize the metal ion charge and completely destroy the primary hydration shell, transforming the hydrated ion into an unhydrated organic molecule, the extractable species, ZnRw
--
“-Amyl Aceiale- HZO
-
Zn(O&),+’
0.10-
O.O( E-
- , 0.001
I I I , , , J ~
, ,18181i
,
0.01 0.4 AQUEOUS HYDROGEN-ION CONCENTRATION, M
io
Figure 1. Effect of carrier diluent upon the distribution ratio of zinc (0.0154M) from aqueous HCI medium with 0.207M di-n-butyl phosphorodithioic acid
+ 2HR(o) a ZnR2co) + 2H+ + 4&0
(2)
At the low concentrations of zinc employed, 0.008 to 0.015M1 it seems certain that the zinc ions exist initially as a mononuclear species in the aqueous phase. Hydrolysis species [such as
loading corrections have been mode
periments indicated that the distribution equilibrium was achieved within 10 minutes.
4
RESULTS
Effect of Carrier Diluent. The distribution of zinc was studied as a function of p H in three solvent systems: CC14-water, methyl isobutyl ketone-water, and n-amyl acetatewater. Di-n-butylphosphorodithioic acid concentration was identical in each system. As shown in Figure 1, the curves, log D us. log [H+], are essentially superimposed, although small differences in the distribution ratio do exist and were reported earlier (3) for a number of solvents. There is no indication that significant complexes form between species containing zinc and carrier diluent molecules unless a similar mixed complex is formed with each solvent, which is not likely in view of their marked structural differences. Identification of Phase Species. The composition of the zinc di-nbutylphosphorodithioate complex was established by studying the distribution of the metal a t different concentrations of the free ligand (Figure 2 ) and a t different concentrations of hydrogen ion (Figure 3), in each case with several different concentrations of reagent, [HR],, in the organic solvent layer. Data for the diethyl ester are included in Figure 2. The ascending portions of Figures 2 and 3 possess slopes of 2 and -2, respectively. I n Figure 2 the aqueous ligand concentration was computed from the relationship [R-] ([HRlo, Lnitial2 lZnItrlo)W[H+l Pr (1) 1164
ANALYTICAL CHEMISTRY
..
I
the curves in Figures 1-4, the distribution ratio is given by
D=- [Zn%lo
to a good approximation and, therefore,
+
log K * = log D 2 log [H+]
1
~~~~~~~~
l l l i l f l l
I
11111111
I
oooj
1.0
Figure 2. Distribution of zinc as a function of aqueous ligand concentration at several concentrations of hydrogen ion for di-n-butyl phosphorodithioic acid (upper curve) and for diethyl phosphorodithioic acid (lower curve)
Zn(OH)+] can be ignored in the pH interval 0 to 7.6. Absence of an intermediate ZnR+, except in the vicinity of the “knee” in Figures 2 and 3, or complexes such as Zn(OH)R(ot and ZnRs(HR),(o) in the organic phase, is proved by the manner in which D varies with [H+] and [R-] for the different initial reagent concentrations (6, 7 ) ; in particular, by the convergeiice of the distinct isograms of Figure 3 to a common plateau. Evaluation of Characteristic Constants. The extraction constant ( K * ) for the partition equilibrium (Equation 1) is given by
+
(5)
K* = K,P.(K./P,)’ (6) where K , is the formation constant and P, the partition coefficient of the metal complex. The value of the extraction constant can be obtained by substituting the appropriate experimental quantities into Equation 3. From the horizontal portions of Figures 2 and 3, the value of the partition coefficient] P, = [ZnR&/ [ZnRz],can be estimated. These values, plus values for the acid dissociation constant and the partition coefficient of the reagent which were determined earlier (8),enable the formation constant to be calculated (Equation 6). These characteristic constants are given in Table I for a number of zinc dialkylphosphorodithioate complexes. Values for the partition coefficients of the diisobutyl-, di-sec-butyl-, and di(2-ethy1)hexyl phosphorodithioate complexes were estimated from extrapolation parallel to the complete distribution curve for the di-n-butyl complex on logarithmic coordinates (Figure 4). DISCUSSION
t(tfaJ
001 01 AOUEOUS L I G A N D [R-] CONCENTRATION, M
log K* = log (ZnR& 2 log [H+l - 2 log [HRlc
- 2 log [HRlo
Also,
J
001 I 00004
(41
[Zn+I]
-
log [Zn+*I (3)
On the ascending rectilinear portions of
The mechanism of the zinc ion extraction into dilute solutions of the dialkylphosphorodithioic acids in carbon tetrachloride is a normal metal chelation without the added complication of solvating reagent molecules or the
0
t
0
2,
~
001 0
~ 04
AQUEObS HYDROGEN-ION CONCENTRATION, M
Figure 3. Distribution of zinc as a function of aqueous HCI concentration for three different concentrations of di-n-butylphosphorodithioic acid present in CCI,
existence of dimeric forms of these reagents in the nonpolar solvent. In progressing from the diethyl ester to esters with longer chain alkyl groups, there occurs a marked increase in the extractability. Inspeation of a S t u a r t Briegleb model of the zinc dialkylphosphorodithioate complex provides a possible explanation. When the linear portion of the alkyl chain exceeds two carbon atoms, normal rotation about the carbon-carbon bisnds provided a protective screen araund the central metal atom, restricting; access of solvent or solute molecules to the site. Branching in the alkyl group increases the extractability to a slight extent, paralleling the amall increases in acidit,y observed for the parent reagentci (8). The stability constant of the zinc complexes provides quantitative evidence for appreciable stability of the phosphorodithioate complex. Data are lacking for a direct comparison with dialkylphosphates. However, the extraction constant for the zinc di-(2-ethyl)hexylphosphorodithioate is 25 as compared with a value of 0.056 reported for
the formation of a 4membered ring with the central zinc ion with less strain and distortion of bond angles (1, 2 ) . LITERATURE CITED
v
Di-(Z-ethyl) hexyl
I
’k
d ,,L , COCil
L
001
AOUEOUS HYDR0GEN.iON
L
L
01
L
IO
CONCENTRATION, M
Figure 4. Distribution of zinc as a function of aqueous HCI concentration for four esters of phosphorodithioic acid, each 0.2M in CCld
zinc di-(2-ethyl)hexylphosphate, which is probably ZnR*(HR)Z in the organic phase (4,6), an increase in extractability by a factor of 450. Conceivably the larger radii of the sulfur atoms permits
(1) Chaney, E.G.,Fernando, Q., Freiser, H..J. Phus. Chcm. 63. 2055 (1959). (2) $ernando, Q., Freher, H., J.‘ Am. C h m . Sne. 80, 4928 (1958). T. H., Dean, J. A., ANAL. CHEM.34, 1312 (1962). ( 4 ) Kimura, K., Bull. Chem. Soc. Japan . 33, 1038 (1960). (5) Lamb, E., U. 8. Atomic Energy Comm. Rept. CF-60-6-132,p. 5 (1960). (6) Rydberg, J., Arkiv Kemi 8, 113 (1955). (7) Schweitzer, G K., Southeastern Regional Meeting, ACS, Gatliiburg, Tenn., 1962. (8) Zucal, R. H., Dean, J. A., Handley, T. H., ANAL.CHEM.35,988(1963).
RECEIVEDMarch 25, 1963. Accepted May 27,1963. Presented a t Southeastern Regional Meeting, ACS, Gatlinburg, Tenn. R. H. Zucal is indebted to the International Atomic Energy Agency for the award of a fellowship which made this work possib!e. ORNL is operated by Union Carbide Nuclear Co. for the U. S. Atomic Energy Commission.
2-Diethy ICI minoet ha nethio I Hy drochIoride as a S pect rophoto met ric Reagent f o r Rhodium SURESH C. SRIVASTAVA Deparfmenf of Chemislry, I ooisiana Sfate University in New Orleans, lakefront, New Orleans 22, la.
,Rhodium produces a yellow color on heating with an excess of 2-diethylaminoethanethiol hydrochloride in aqueous solution. Studies have been performed to investigate the optimum conditions for the spectrophotometric determination of trace amounts of rhodium using this reagent. The sensitivity of the color reaction is 0.0064 micron per square ‘centimeter. The region of maximum absorption of the complex lies at 330 mp when measured against a reagent b h k . About 30 minutes’ heating is required for the color to develop fully and the optimum pH range is 1.5 to 3.5. The system adheres to Beer’s law over a concentration range of 0.:3to 15.4 p.p.m. of rhodium, and the optimum range for the most accurate spectrophotometric measurements is 0.6 to 8.3 p.p.m. The average and maximum relative standard deviations, respectively, are 0.63 and 1.26%, as shown by the absorbance readings of samples containing 3.09 p.p.m. rhodium. The interferences due to other platinum metals have been studied.
T
a few analytical procedures described in the literature for the spectrophotometric determination of trace amounts of rhodium, but still there is a need for better and more sensitive reagents. Beamish and McBryde (2) reviewed the existing colorimetric methods for the determination of rhodium and concluded that the tin(I1) chloride method (1, 2, 6) is, by far, the most satisfactory one. The use of tin(I1) bromide reagent for the determination of rhodium has also been described (3, 4, 7). Jacobs (6) and Wilson and Jacobs (10) have studied N,N’-bis(3-dimethylaminopropyl) dithio oxamide and p-nitrosodimethylaniline, respectively, as reagents for the colorimetric determination of rhodium. Wagner and Yoe (9) have also described the spectrophotometric determination of rhodium with thiomalic acid and the simultaneous determination of rhodium and palladium. Most of the reagents that have been employed are not water soluble and also lack sensitivity. This paper describes the use of 2-diethylaminoethanethiol hydrochloride (DAT) HERE HAVE BEEN
aa a spectrophotometric reagent for the determination of trace amounts of rhodium. The reagent and the rhodium complex are soluble in aqueous media and the sensitivity of the reaction is fairly comparable with the procedures described in the literature. EXPERIMENTAL
Materials. STANDARD RHODIUM SOLUTION.A standard solution of rhodium nitrate was obtained from J. Bishop and Co. The stock solution was diluted suitably to give a final solution containing 1.00 mg. per ml. of rhodium. Rhodium content was determined gravimetrically by first precipitating rhodium as the sulfide then igniting to the oxide and finally reducing to the metal in presence of hydrogen. REAGENT SOLUTION. A stock solution containing 5 mg. per ml. was prepared by dissolving 2.6 grams of reagent grade Zdiethylaminoethanethiol hydrochloride (K and K Laboratories) in 500 ml. of distilled water. The solution was standardized by determining its sulfur content. It was stable for several days when stored in an inert atmosphere VOL. 35, NO. 9, AUGUST 1963
1165
