- Email: [email protected]
Extraction of palladium with tri-isobutylphosphine sulphide (cyanex 471) in toluene from chloride solutions containing thiocyanate
Talaata. Vol. 38, No. 5, pp. 483-488, 1991
0039-9140/91$3.00+ 0.00 Copyright 0 1991 Pergamon Press plc
Printed in Great Britain. All rights reserved
EXTRACTION OF PALLADIUM WITH TRI-ISOBUTYLPHOSPHINE SULPHIDE (CYANEX 471) IN TOLUENE FROM CHLORIDE SOLUTIONS CONTAINING THIOCYANATE M. HIDALGO, A. MASANA and V. SALVADO Unitat de Quimica, Collegi Universitari de Girona, U.A.B., 17071 Girona, Spain M. Mufioz
and M. VALIENTE*
Quimica Analitica, U.A.B., 08193, Bellaterra, Barcelona, Spain (Received 21 April 1989. Revised 24 October 1990. Accepted 16 November 1990)
Summary-The extraction of Pd(II) by hi-isobutylphosphine sulphide, TIBPS (Cyanex 471x), in toluene from aqueous chloride solutions (containing small amounts of thiocyanate) has been investigated. The extraction is enhanced by the presence of thiocyanate, owing to formation of mixed-ligand Pd(II)-Cl--SCN--TIBPS complexes. Analysis of the metal distribution suggests the formation of PdCl(SCN) .TIBPS, PdCl(SCN).2TIBPS, Pd(SCN),.TIBPS and Pd(SCN),.2TIBPS in the organic phase. The equilibrium constants are log&,, = 9.56, log&,, = 12.70, log&, = 14.73 and log& = 17.17, respectively. The ultraviolet absorption spectra of the organic phase support the hypothesis of formation of mixed-ligand complexes.
Solvent extraction has been widely used for the separation of palladium,’ especially with sulphur-containing extractants, which are known to be highly selective for extraction of metals classified as “soft acids”. Cyanex 471x is a phosphine sulphide reagent, the active component of which is triisobutylphosphine sulphide (TIBPS). It has been reported as effective for extraction of Pd(II),Z-5 Ag(I)6 and Hg(I1)’ and the separation of Pd(II)-Pt(IV) mixtures. However, the slow rate of the extraction of Pd(II) is one of the major problems in the separation of the platinum group metals.8 Al-Bazi and Freiser’ reported a large increase in the rate of extraction of Pd(I1) by Kelex 100 [7(1 -vinyl-3,3,5,5-tetramethylhexyl)-8-quinolinol] when sodium thiocyanate was present in the aqueous phase, and regarded this as due to the truns-effect in Pd(SCN),C12- produced in the aqueous phase. The aim of the present work was to investigate the effect of thiocyanate on the Pd(II)-ClTIBPS system and to explain the role of thiocyanate in enhancement of the extraction rate.
*Author for correspondence. TAL 38,sc
EXPERIMENTAL
Reagents
Tri-isobutylphosphine sulphide, TIBPS, was obtained from Cyanex 471x (kindly supplied by the American Cyanamid Co.) by recrystallization from an ethanol-water mixture,3 and used to prepare a stock 1 x 10e3M solution in toluene (Merck p.a.) which had been successively washed with OSM sodium hydroxide, distilled water, OSM hydrochloric acid and doubly distilled water before use. Sodium chloride (Merck p.a.) was purified as previously described.‘O Sodium thiocyanate solutions were standardized by titration with silver nitrate, iron(II1) ammonium sulphate being used as indicator. A stock solution of Pd(I1) (3 x 10e3M, pH = 1) was prepared by dissolving the required amount of the chloride (Spanish Society of Precious Metals) in 0.1 M hydrochloric acid, and was standardized gravimetrically with dimethylglyoxime.” In the aqueous phases the total concentrations of Pd(I1) used were 3 x 10V5,1.2 x lOA and 2.3 x 10-4M. The total thiocyanate concentration was varied between 2.5 x 10e5 and 1 x 10e3M. The pH was kept constant at 2.5 to avoid hydrolysis of palladium. The total chloride concentration was 1M.
M. HIDALCO~~ al.
484
‘.5-(tit)
The total TIBPS concentration was varied in the range from 2 x 10m4to 1 x 10e3M.
1.0 I
Apparatus
I I
A Bausch and Lomb Spectronic 2000 spectrophotometer was used.
p
0.5
I.
-
:r
0” J
00 0.0 0 .OO l 00 l * 00
O-
Procedure
Aqueous solutions with a total chloride concentration of l.OM and various concentrations of palladium and thiocyanate were prepared. Equal volumes (10 ml) of the organic and aqueous phases were mixed in stoppered glass tubes and shaken for 90 min on a rack rotating at 40 rpm and kept at 22 & 1”. Preliminary work had shown that shaking for 90 min was adequate for equilibrium to be reached (Fig. 1). The mixture was then centrifuged and the phases were separated. Excess of thiocyanate was added to the aqueous phase and the absorbance at 310 nm was measured to determine the palladium. I2 For several sets of conditions the Pd(I1) mass balance was checked by stripping the palladium from the organic phase with O.lM sodium thiocyanate and measuring the absorbance of the resulting aqueous phase. RESULTS
-
-1.0 1 -4.0
o [SCN-Jtot-2
E-5M
+ [SCN-ltot-5
E-5M
[SCN-Jtot-1
E-4M E-4M E-4M E-4M
n
0 [SCN-Itot-
03
. [SCN-Itot0 [SCN-Itot-
”
I
I
I
-3.5
-3.0
-2.5
Log [TIBPS]tot [SCN-]tot-1
q
E-4M E-4M E-4M
+ [SCN-]totr3
“5-(b)
9 [SCN-]tob2 1.0 -
D
0.5 -
5 P
v4’
8 A
+ .
o-
I18
0 -0.5
0
-
Ll
-l.pq.o_
-2.5
Log [TIBPS]tot
AND DISCUSSION
The palladium distribution given by D =
-0.5
.I-’ .‘ooa
1.5
(c)
coefficient, D, is 1.0
0
.
0
[PW)I,, /PW)I,,
0
(1)
Q
where [Pd(II),, and [Pd(II)],, are the total palladium concentrations in the organic and aqueous phase, respectively. Figure 2 gives the palladium distribution coefficient as a function of the total TIPBS concentration for total Pd(I1) concentrations of 3.2,
g -I
0.5 q q
0
0 .
: I
Irn
*
4 o [SCN-Jtot-3 * [SCN-Itotn [SCN-Itot-
-0.5
E-4M E-4M E-4M
I
-1:04.0
-2.5
Log (TIBPS]tot
5 q
4
(SCN-Itot-
.O E-4M
* [SCN-]tot=S.O
E-4M
[SCN-]tot-5.0
E-4M
n
0 [SCN‘]tot-0.0 0
0
0
E-4M 0
Fig. log
2. Experimental distribution data, log D us. ~IBPS],,, at different SCN- levels. (pd(II)],,,: (a) 3.0 x 10-5M; (b) 1.2 x 10-4M; (c) 2.3 x lo-“M.
0
Time min Fig. 1. Palladium concentration in the aqueous phase, plotted against time, for different SCN- concentrations. [pd(II)],,, = 3.0 x IO-‘M; [TIBPS],,, = 7.0 x 10-4A4.
12.8 and 24.8 ppm and different thiocyanate concentrations and Fig. 3 shows the data in the form log D US. log [SCN-],O,. Taking into account that in our case ([Cl-] = l.OM), in the absence of SCN- the Pd(I1) in the aqueous phase will be present as the PdCl:- complex, equation (1) can be expressed as: D=
PWIIo, [PdCl;- ] (1 + X,[SCN-
]‘/[Cl- 1’)
(2)
Extraction of palladium 2
o [TIBPB]tot-3.5 + [TIBPS]tot-4.2 . [TIBPS]tot-5.1 o [TlBPS]tot-6.0 l [TIBPS]tot=7.0
1
E-4M E-4M E-4M E-4M E-4M
-la
-3
-5
Log [SCN’Jtot Fig. 3. Experimental distribution data, log D, plotted against log [SCN-I,,,. [Pd(II&, = 3.0 x 10-SM. Solid lines are the calculated curves based on the proposed model.
where the Ki values are the equilibrium stants of the reactions
con-
PdCl:- + iSCN- s PdCl,_ ,SCN:- + iCl- (3) with i = 1, 2, 3 and 4. The values of Ki obtained from the literatureI were log K1 = 6.03, log K2 = 10.93, log K3 = 14.52 and log & = 17.55. Likewise, the general extraction reaction when SCN- is added could be expressed as: r PdCl:- + q SCN- +
pTIBPS,,
N
PdCl:- + SCN- + TIBPS,, + 3Cl-
(6)
+ 3Cl-
(7)
+ 4Cl-
(8)
= Pd(SCN)*. 2TIBPS,,, + 4Cl-
(9)
log K,,, = 9.56 (4)
where q = O-2. The equilibrium constant can be termed Krgp. Since TIBPS is a neutral extractan&’ it follows that its role in the extraction is solely solvation of the palladium complex extracted, which itself must therefore be neutral. The solubility of TIBPS in aqueous solution was determined by equilibrating hydrochloric acid with a solution of TIBPS in toluene, and analysing the aqueous phase for TIBPS by ICP spectrometry and HPLC. The solubility of TIBPS in aqueous solution was found to be < 1.10 x lo-6M. To evaluate the composition of the extracted species as well as the corresponding formation constants, the experimental data were analysed numerically with the LETAGROP-DISTR program, I4 which is based on minimization of the error-square sum, U: u = c (log &I - log &p)*
mass-balance equations for a proposed model: N is the total number of experimental points. Our proposed models include the mixedligand complexes present in the aqueous phase [equation (311and protonation of thiocyanate as a constant set in all calculations. The results of the LETAGROP-DISTR calculations for different models tested, based on equation (4), are summarized in Table 1. Polynuclear species of Pd(I1) were also tested but were always rejected. It is interesting that the results obtained for extraction of PdC12 solvated with TIBPS, model I, are far from agreement with the experimental data, and when mixed species are included in the model (e.g., VII), the PdCl, species is rejected. A binuclear species is also rejected in the calculations (models VI and IX). As seen from Table 1, model X, consisting of the species PdCl(SCN) - TIBPS, PdCl(SCN) - 2TIBPS, Pd(SCN)* *TIBPS and Pd(SCN), *2TIBPS formed in the organic phase gives the best fit to the experimental data. Thus, the chemical reactions regarded as responsible for the extraction seem to be
z$ PdCl(SCN)*TIBPS,,
= PdCl,, - 4)(SCN&.P (TIBPS),, + r(q + 2)Cl-
485
(5)
where D, are the experimental values of the distribution coefficient and D,, are the corresponding values calculated from the relevant
PdCl:- + SCN- + ZTIBPS,, $ PdCl(SCN)*ZTIBPS,, log Kn2 = 12.70 PdCl:- + 2SCN- + TIBPS,, z$ Pd(SCN)2.TIBPS,, log K,,, = 14.73 PdCl:- + 2SCN- + ZTIBPS,,,
log K,,, = 17.17 Figures 4 and 5 show the distributions of the species proposed in this model, as a function of the total TIBPS and thiocyanate concentrations, respectively. As seen, the species containing Pd(SCN), predominate under most conditions, and although Pd(SCN),*TIBPS is predominant in the range of extractant concentrations studied, Pd(SCN)* - ZTIBPS becomes more important as the concentration of TIBPS increases (as might be expected). In additional experiments the ultraviolet absorption spectra of the organic phases obtained
M.
486
HIDALGOet al.
Table 1. Results of numerical calculations for the species Pd, Cl,,_,,(SCN);pTIBPS [equation (4)] by the program LETAGROPDISTR; n (number of experimental points) = 89 Model
Species ry q9 p
logK/,,
a(logD)
V
1,092 1,192 191,l 1,191 1,192 1,191 1,&l l,l,l 1,291 2,2,2 1,191 1,291 1,0,2 l,l,l 1,192 1,2,1 1,191 1,192 1,291 2,2,2 l,l,l 1,1,2 1,291 1,292
10.27; 10.51 14.21 $0.19 10.83 f 0.20 9.90; 10.55 14.15; 14.38 9.74 f 0.17 14.79 + 0.09 9.00; 9.68 14.82 f 0.08 9.70 f 0.20 14.80 f 0.09 9.51; 9.80 12.76; 13.20 14.79 f 0.10 9.60; 9.90 12.68; 13.21 14.81 f 0.09 9.56; 9.86 12.70; 13.17 14.73 It 0.23 17.17; 17.83
0.76 0.42 0.42 0.42
0.45E2 0.16E2 0.17E2 0.15E2
0.20
0.35El
0.19
0.33El
0.20
0.34El
0.17
0.27El
0.18
0.28El
0.16
0.20El
I II III IV V VI VII VIII
IX
X
Remarks
rejected
rejected
rejected
The standard deviation o(logD) is defined as a(logD) = (V/(N - A$))“* where A$ is the number of constants to be adjusted. The error in the constants vs is given as + 3a (1ogK) but for a(K) z 0.2K, the “best” value of 1ogK and the maximum value [log(K + 3u(K)] are given.
by extraction of aqueous Pd(I1) solutions (in the presence and absence of thiocyanate) with a toluene solution of TIBPS were recorded. For the extraction without thiocyanate present the shaking time was extended to a week to ensure that equilibrium was reached. The spectra are shown in Fig. 6(a). As seen, there is a blue-shift of the absorbance maximum when the extraction is done with thiocyanate present in the
1.
Pd(SCN)
aqueous phase, whereas there is a red-shift in the spectrum of the initial aqueous phases when thiocyanate is added, Fig. 6(b). The blue-shift seems to be a consequence of the change in medium as well as the ligand-exchange (and is analogous to the spectral differences between palladium thiocyanate species in aqueous medium and in molten salts12) and the red-shift
1 .o
2 .TIBPS
1.
Pd(SCN)2.TIBPS
2.
PdCI(SCN)
3.
PdCI(SCN)
4. 1.0 0.9
4.
PdCI,
0.6 F
(SCN)2-
0.6 0.7
6.
PdCI(SCN)
0.6
7.
PdClSCN
.TIBPS
-3.4
-3.0
.PTIBPS
.TIBPS Pd(SCN)2.2TIBPS
5.
PdC12(SCN);-
6.
PdCI(SCN)
7.
Pd(SCN);-
g‘
.STIBPS
0.5 0.4 0.3 0.2 0.1 %O
-3.6
-3.6
-3.2
Log [TIBPS]
-2.6
-2.6
-2.4
tot
Fig. 4. Distribution diagram of metal species us. [TIBPS],,, . Ipd(II)],,, = 3.0 x IO-‘M; [SCN-],O, = 1.0 x lo-‘M.
Log [SCN-]tot Fig. 5. Distribution diagram of palladium-containing species 0s. aqueous [SCN-I,,. [pd(II)],,, = 3.0 x 10WsM; [TIBPS],,, = 1.O x lo-‘M.
Extraction of palladium
l-
(a)
2
g z
0200
250
300
350
400
X nm Fig. 6. (a) Ultraviolet spectra of organic species: [Pd(II)],, = 3.0 x 10-‘&f. 1, No SCN- present in the aqueous phase; 2, [SCN-I,,, = 5.0 x 10esM. (b) Ultraviolet spectra of aqueous complexes: [pd(II)],,, = 3.0 x 10esM. 1, No SCNpresent: 2, [SCN-],O, = 5.0 x 10e5M; 3, [SCN-I,,, = 5.0 x 10-4M: 4, [SCN-I,,, = 1.0 x 10-3M.
a consequence of ligand-exchange in aqueous medium.‘s*‘6 A comparison of these results with those reported by Walker and BautistaZ and by Inoue and co-workers3” leads to conclusions different from theirs. Walker and Bautista? explain the extraction of Pd(I1) by TIBPS in xylene, in the absence of thiocyanate, as due to formation of Pd(TIBPS)2 species in the organic phase, and Inoue and co-workers consider that a binuclear 2: 2 complex also contributes to the extraction into toluene. In both systems metal extraction is assumed to occur by a solvating mechanism. Neither Walker and Bautista nor Inoue and co-workers give values for the reaction constants. The main difference from our results is related to the formation of the binuclear species reported by Inoue and co-workers, and this may be due to the different ranges of Pd(I1) concentration used, which in our case were limited by the solubility of Pd(I1) in aqueous solutions containing thiocyanate. Furthermore, thio-
487
cyanate provides a different chemical environment, which may result in a different speciation. Species with 1: 2 metal : extractant stoichiometry are also found when triphenylphosphine in 1,Zdichloroethane is used as the extractant (in the absence of thiocyanate).” Our results also partially agree with those reported in the literature for the extraction of different metals with Cyanex 471x, e.g., Ag(1) is extracted from nitrate solutions as AgNO, *2TIBPS,6 and Hg(I1) from chloride media as HgCl,.2TIBPS,’ and a recent study concerning Au(II1) extraction from hydrochloric acid media with Cyanex 47 lx showed that the species AuC13*TIBPS and AuCl, *2TIBPS are responsible for the gold extraction.” In our study the effect of thiocyanate on the extraction of Pd(I1) from chloride solutions has been shown to be related to the formation of mixed-ligand complexes PdCl, _ , SCN:- in the aqueous phase, which react more rapidly than the corresponding tetrachloro-complex PdCl:- . As a result an increase in the rate of Pd(I1) extraction is obtained (see Fig. 1). The trends observed at the different thiocyanate concentrations may be related to the similar trend in the relative concentrations of the species PdCl,(SCN):- and PdCl(SCN):- under the corresponding experimental conditions. Such behaviour has previously been explained as a catalytic process based on a trans-ligand effect.’ However, taking into account our findings that thiocyanate is present in the extracted metal species, the effect of the thiocyanate should now be considered as a particular case of synergism rather than catalysis, although a trans-effect of the thiocyanate in the mixedligand aqueous complexes may well be important in formation of the metal species finally extracted. Acknowledgements-The
present work is part of the CAICYT project MAT88-752. M. H. also thanks the Catalan Institution CIRIT for its financial contribution towards the equipment needed to carry out this work.
REFERENCES 1. F. E. Beamish, The Analytical Chemistry of the Noble Metals, Pergamon Press, London, 1966. 2. R. D. Walker and R. Bautista. Proc. Int. Solo. Ext. Con& LSEC-86, Munich, 1986, Paper 11-107. 3. K. Inoue and Y. Baba, ibid., Paper 11-263. 4. Y. Baba, M. Ohshima and K. Inoue, Bull. Chem. Sot. Japan, 1986, 59, 3829. 5. Y. Baba and K. Inoue, Ind. Eng. Chem. Res., 1988.27,
1613.
488
M. HIDALGCI er al.
6. K. Inoue, Y. Baba and M. Tagaki, Proc. Inr. Solu. Ext. Cod, ISEC ‘86. Munich, 1986. Paper B-65. 7. Y. Baba, Y. Umezaki and K. Inoue, Solu. Ext. fan Exch., 1986, 4, IS. 8. S. J. Al-Bazi and A. Chow, Tufanra, 1986, 31, 815. 9. S. J. Al-Bazi and H. Freiser, Solo. Ext. ion Exch., 1986, 4 1121. 10. Some Laborarory Methodr, Inorganic Chemistry, The Royal Institute of Technology (KTH), Stockholm, 1959. 11. A. I. Vogel, Textbook of Quantitative Inorganic Analysis, 4th Ed., Longmans, London, 1978.
12. K. S. De Haas, J. Inorg. Nucl. Chem., 1973, 35, 3231. 13. A. A. Biryukou and V. I. Shlenskaya, Russ. J. Inorg. Chem., 1967, 12, 1362. 14. D. H. Liem, Acta Chem. Scud, 1971, 25, 1521. 15. L. I. Elding, Inorg. Chim. Acta, 1972, 6, 647. 16. A. K. Sundaram and E. B. Sandell, J. Am. Chem. Sot., 1955, 77, 855. 17. M. Mojski, Talanta, 1980, 27, 7. 18. V. Salvado, M. Hidalgo, A. Masana, M. Mutioz, M. Valiente and M. Muhammed, Solo. Ext. Ion Exch., 1990, 8, 491.
0039-9140/91$3.00+ 0.00 Copyright 0 1991 Pergamon Press plc
Printed in Great Britain. All rights reserved
EXTRACTION OF PALLADIUM WITH TRI-ISOBUTYLPHOSPHINE SULPHIDE (CYANEX 471) IN TOLUENE FROM CHLORIDE SOLUTIONS CONTAINING THIOCYANATE M. HIDALGO, A. MASANA and V. SALVADO Unitat de Quimica, Collegi Universitari de Girona, U.A.B., 17071 Girona, Spain M. Mufioz
and M. VALIENTE*
Quimica Analitica, U.A.B., 08193, Bellaterra, Barcelona, Spain (Received 21 April 1989. Revised 24 October 1990. Accepted 16 November 1990)
Summary-The extraction of Pd(II) by hi-isobutylphosphine sulphide, TIBPS (Cyanex 471x), in toluene from aqueous chloride solutions (containing small amounts of thiocyanate) has been investigated. The extraction is enhanced by the presence of thiocyanate, owing to formation of mixed-ligand Pd(II)-Cl--SCN--TIBPS complexes. Analysis of the metal distribution suggests the formation of PdCl(SCN) .TIBPS, PdCl(SCN).2TIBPS, Pd(SCN),.TIBPS and Pd(SCN),.2TIBPS in the organic phase. The equilibrium constants are log&,, = 9.56, log&,, = 12.70, log&, = 14.73 and log& = 17.17, respectively. The ultraviolet absorption spectra of the organic phase support the hypothesis of formation of mixed-ligand complexes.
Solvent extraction has been widely used for the separation of palladium,’ especially with sulphur-containing extractants, which are known to be highly selective for extraction of metals classified as “soft acids”. Cyanex 471x is a phosphine sulphide reagent, the active component of which is triisobutylphosphine sulphide (TIBPS). It has been reported as effective for extraction of Pd(II),Z-5 Ag(I)6 and Hg(I1)’ and the separation of Pd(II)-Pt(IV) mixtures. However, the slow rate of the extraction of Pd(II) is one of the major problems in the separation of the platinum group metals.8 Al-Bazi and Freiser’ reported a large increase in the rate of extraction of Pd(I1) by Kelex 100 [7(1 -vinyl-3,3,5,5-tetramethylhexyl)-8-quinolinol] when sodium thiocyanate was present in the aqueous phase, and regarded this as due to the truns-effect in Pd(SCN),C12- produced in the aqueous phase. The aim of the present work was to investigate the effect of thiocyanate on the Pd(II)-ClTIBPS system and to explain the role of thiocyanate in enhancement of the extraction rate.
*Author for correspondence. TAL 38,sc
EXPERIMENTAL
Reagents
Tri-isobutylphosphine sulphide, TIBPS, was obtained from Cyanex 471x (kindly supplied by the American Cyanamid Co.) by recrystallization from an ethanol-water mixture,3 and used to prepare a stock 1 x 10e3M solution in toluene (Merck p.a.) which had been successively washed with OSM sodium hydroxide, distilled water, OSM hydrochloric acid and doubly distilled water before use. Sodium chloride (Merck p.a.) was purified as previously described.‘O Sodium thiocyanate solutions were standardized by titration with silver nitrate, iron(II1) ammonium sulphate being used as indicator. A stock solution of Pd(I1) (3 x 10e3M, pH = 1) was prepared by dissolving the required amount of the chloride (Spanish Society of Precious Metals) in 0.1 M hydrochloric acid, and was standardized gravimetrically with dimethylglyoxime.” In the aqueous phases the total concentrations of Pd(I1) used were 3 x 10V5,1.2 x lOA and 2.3 x 10-4M. The total thiocyanate concentration was varied between 2.5 x 10e5 and 1 x 10e3M. The pH was kept constant at 2.5 to avoid hydrolysis of palladium. The total chloride concentration was 1M.
M. HIDALCO~~ al.
484
‘.5-(tit)
The total TIBPS concentration was varied in the range from 2 x 10m4to 1 x 10e3M.
1.0 I
Apparatus
I I
A Bausch and Lomb Spectronic 2000 spectrophotometer was used.
p
0.5
I.
-
:r
0” J
00 0.0 0 .OO l 00 l * 00
O-
Procedure
Aqueous solutions with a total chloride concentration of l.OM and various concentrations of palladium and thiocyanate were prepared. Equal volumes (10 ml) of the organic and aqueous phases were mixed in stoppered glass tubes and shaken for 90 min on a rack rotating at 40 rpm and kept at 22 & 1”. Preliminary work had shown that shaking for 90 min was adequate for equilibrium to be reached (Fig. 1). The mixture was then centrifuged and the phases were separated. Excess of thiocyanate was added to the aqueous phase and the absorbance at 310 nm was measured to determine the palladium. I2 For several sets of conditions the Pd(I1) mass balance was checked by stripping the palladium from the organic phase with O.lM sodium thiocyanate and measuring the absorbance of the resulting aqueous phase. RESULTS
-
-1.0 1 -4.0
o [SCN-Jtot-2
E-5M
+ [SCN-ltot-5
E-5M
[SCN-Jtot-1
E-4M E-4M E-4M E-4M
n
0 [SCN-Itot-
03
. [SCN-Itot0 [SCN-Itot-
”
I
I
I
-3.5
-3.0
-2.5
Log [TIBPS]tot [SCN-]tot-1
q
E-4M E-4M E-4M
+ [SCN-]totr3
“5-(b)
9 [SCN-]tob2 1.0 -
D
0.5 -
5 P
v4’
8 A
+ .
o-
I18
0 -0.5
0
-
Ll
-l.pq.o_
-2.5
Log [TIBPS]tot
AND DISCUSSION
The palladium distribution given by D =
-0.5
.I-’ .‘ooa
1.5
(c)
coefficient, D, is 1.0
0
.
0
[PW)I,, /PW)I,,
0
(1)
Q
where [Pd(II),, and [Pd(II)],, are the total palladium concentrations in the organic and aqueous phase, respectively. Figure 2 gives the palladium distribution coefficient as a function of the total TIPBS concentration for total Pd(I1) concentrations of 3.2,
g -I
0.5 q q
0
0 .
: I
Irn
*
4 o [SCN-Jtot-3 * [SCN-Itotn [SCN-Itot-
-0.5
E-4M E-4M E-4M
I
-1:04.0
-2.5
Log (TIBPS]tot
5 q
4
(SCN-Itot-
.O E-4M
* [SCN-]tot=S.O
E-4M
[SCN-]tot-5.0
E-4M
n
0 [SCN‘]tot-0.0 0
0
0
E-4M 0
Fig. log
2. Experimental distribution data, log D us. ~IBPS],,, at different SCN- levels. (pd(II)],,,: (a) 3.0 x 10-5M; (b) 1.2 x 10-4M; (c) 2.3 x lo-“M.
0
Time min Fig. 1. Palladium concentration in the aqueous phase, plotted against time, for different SCN- concentrations. [pd(II)],,, = 3.0 x IO-‘M; [TIBPS],,, = 7.0 x 10-4A4.
12.8 and 24.8 ppm and different thiocyanate concentrations and Fig. 3 shows the data in the form log D US. log [SCN-],O,. Taking into account that in our case ([Cl-] = l.OM), in the absence of SCN- the Pd(I1) in the aqueous phase will be present as the PdCl:- complex, equation (1) can be expressed as: D=
PWIIo, [PdCl;- ] (1 + X,[SCN-
]‘/[Cl- 1’)
(2)
Extraction of palladium 2
o [TIBPB]tot-3.5 + [TIBPS]tot-4.2 . [TIBPS]tot-5.1 o [TlBPS]tot-6.0 l [TIBPS]tot=7.0
1
E-4M E-4M E-4M E-4M E-4M
-la
-3
-5
Log [SCN’Jtot Fig. 3. Experimental distribution data, log D, plotted against log [SCN-I,,,. [Pd(II&, = 3.0 x 10-SM. Solid lines are the calculated curves based on the proposed model.
where the Ki values are the equilibrium stants of the reactions
con-
PdCl:- + iSCN- s PdCl,_ ,SCN:- + iCl- (3) with i = 1, 2, 3 and 4. The values of Ki obtained from the literatureI were log K1 = 6.03, log K2 = 10.93, log K3 = 14.52 and log & = 17.55. Likewise, the general extraction reaction when SCN- is added could be expressed as: r PdCl:- + q SCN- +
pTIBPS,,
N
PdCl:- + SCN- + TIBPS,, + 3Cl-
(6)
+ 3Cl-
(7)
+ 4Cl-
(8)
= Pd(SCN)*. 2TIBPS,,, + 4Cl-
(9)
log K,,, = 9.56 (4)
where q = O-2. The equilibrium constant can be termed Krgp. Since TIBPS is a neutral extractan&’ it follows that its role in the extraction is solely solvation of the palladium complex extracted, which itself must therefore be neutral. The solubility of TIBPS in aqueous solution was determined by equilibrating hydrochloric acid with a solution of TIBPS in toluene, and analysing the aqueous phase for TIBPS by ICP spectrometry and HPLC. The solubility of TIBPS in aqueous solution was found to be < 1.10 x lo-6M. To evaluate the composition of the extracted species as well as the corresponding formation constants, the experimental data were analysed numerically with the LETAGROP-DISTR program, I4 which is based on minimization of the error-square sum, U: u = c (log &I - log &p)*
mass-balance equations for a proposed model: N is the total number of experimental points. Our proposed models include the mixedligand complexes present in the aqueous phase [equation (311and protonation of thiocyanate as a constant set in all calculations. The results of the LETAGROP-DISTR calculations for different models tested, based on equation (4), are summarized in Table 1. Polynuclear species of Pd(I1) were also tested but were always rejected. It is interesting that the results obtained for extraction of PdC12 solvated with TIBPS, model I, are far from agreement with the experimental data, and when mixed species are included in the model (e.g., VII), the PdCl, species is rejected. A binuclear species is also rejected in the calculations (models VI and IX). As seen from Table 1, model X, consisting of the species PdCl(SCN) - TIBPS, PdCl(SCN) - 2TIBPS, Pd(SCN)* *TIBPS and Pd(SCN), *2TIBPS formed in the organic phase gives the best fit to the experimental data. Thus, the chemical reactions regarded as responsible for the extraction seem to be
z$ PdCl(SCN)*TIBPS,,
= PdCl,, - 4)(SCN&.P (TIBPS),, + r(q + 2)Cl-
485
(5)
where D, are the experimental values of the distribution coefficient and D,, are the corresponding values calculated from the relevant
PdCl:- + SCN- + ZTIBPS,, $ PdCl(SCN)*ZTIBPS,, log Kn2 = 12.70 PdCl:- + 2SCN- + TIBPS,, z$ Pd(SCN)2.TIBPS,, log K,,, = 14.73 PdCl:- + 2SCN- + ZTIBPS,,,
log K,,, = 17.17 Figures 4 and 5 show the distributions of the species proposed in this model, as a function of the total TIBPS and thiocyanate concentrations, respectively. As seen, the species containing Pd(SCN), predominate under most conditions, and although Pd(SCN),*TIBPS is predominant in the range of extractant concentrations studied, Pd(SCN)* - ZTIBPS becomes more important as the concentration of TIBPS increases (as might be expected). In additional experiments the ultraviolet absorption spectra of the organic phases obtained
M.
486
HIDALGOet al.
Table 1. Results of numerical calculations for the species Pd, Cl,,_,,(SCN);pTIBPS [equation (4)] by the program LETAGROPDISTR; n (number of experimental points) = 89 Model
Species ry q9 p
logK/,,
a(logD)
V
1,092 1,192 191,l 1,191 1,192 1,191 1,&l l,l,l 1,291 2,2,2 1,191 1,291 1,0,2 l,l,l 1,192 1,2,1 1,191 1,192 1,291 2,2,2 l,l,l 1,1,2 1,291 1,292
10.27; 10.51 14.21 $0.19 10.83 f 0.20 9.90; 10.55 14.15; 14.38 9.74 f 0.17 14.79 + 0.09 9.00; 9.68 14.82 f 0.08 9.70 f 0.20 14.80 f 0.09 9.51; 9.80 12.76; 13.20 14.79 f 0.10 9.60; 9.90 12.68; 13.21 14.81 f 0.09 9.56; 9.86 12.70; 13.17 14.73 It 0.23 17.17; 17.83
0.76 0.42 0.42 0.42
0.45E2 0.16E2 0.17E2 0.15E2
0.20
0.35El
0.19
0.33El
0.20
0.34El
0.17
0.27El
0.18
0.28El
0.16
0.20El
I II III IV V VI VII VIII
IX
X
Remarks
rejected
rejected
rejected
The standard deviation o(logD) is defined as a(logD) = (V/(N - A$))“* where A$ is the number of constants to be adjusted. The error in the constants vs is given as + 3a (1ogK) but for a(K) z 0.2K, the “best” value of 1ogK and the maximum value [log(K + 3u(K)] are given.
by extraction of aqueous Pd(I1) solutions (in the presence and absence of thiocyanate) with a toluene solution of TIBPS were recorded. For the extraction without thiocyanate present the shaking time was extended to a week to ensure that equilibrium was reached. The spectra are shown in Fig. 6(a). As seen, there is a blue-shift of the absorbance maximum when the extraction is done with thiocyanate present in the
1.
Pd(SCN)
aqueous phase, whereas there is a red-shift in the spectrum of the initial aqueous phases when thiocyanate is added, Fig. 6(b). The blue-shift seems to be a consequence of the change in medium as well as the ligand-exchange (and is analogous to the spectral differences between palladium thiocyanate species in aqueous medium and in molten salts12) and the red-shift
1 .o
2 .TIBPS
1.
Pd(SCN)2.TIBPS
2.
PdCI(SCN)
3.
PdCI(SCN)
4. 1.0 0.9
4.
PdCI,
0.6 F
(SCN)2-
0.6 0.7
6.
PdCI(SCN)
0.6
7.
PdClSCN
.TIBPS
-3.4
-3.0
.PTIBPS
.TIBPS Pd(SCN)2.2TIBPS
5.
PdC12(SCN);-
6.
PdCI(SCN)
7.
Pd(SCN);-
g‘
.STIBPS
0.5 0.4 0.3 0.2 0.1 %O
-3.6
-3.6
-3.2
Log [TIBPS]
-2.6
-2.6
-2.4
tot
Fig. 4. Distribution diagram of metal species us. [TIBPS],,, . Ipd(II)],,, = 3.0 x IO-‘M; [SCN-],O, = 1.0 x lo-‘M.
Log [SCN-]tot Fig. 5. Distribution diagram of palladium-containing species 0s. aqueous [SCN-I,,. [pd(II)],,, = 3.0 x 10WsM; [TIBPS],,, = 1.O x lo-‘M.
Extraction of palladium
l-
(a)
2
g z
0200
250
300
350
400
X nm Fig. 6. (a) Ultraviolet spectra of organic species: [Pd(II)],, = 3.0 x 10-‘&f. 1, No SCN- present in the aqueous phase; 2, [SCN-I,,, = 5.0 x 10esM. (b) Ultraviolet spectra of aqueous complexes: [pd(II)],,, = 3.0 x 10esM. 1, No SCNpresent: 2, [SCN-],O, = 5.0 x 10e5M; 3, [SCN-I,,, = 5.0 x 10-4M: 4, [SCN-I,,, = 1.0 x 10-3M.
a consequence of ligand-exchange in aqueous medium.‘s*‘6 A comparison of these results with those reported by Walker and BautistaZ and by Inoue and co-workers3” leads to conclusions different from theirs. Walker and Bautista? explain the extraction of Pd(I1) by TIBPS in xylene, in the absence of thiocyanate, as due to formation of Pd(TIBPS)2 species in the organic phase, and Inoue and co-workers consider that a binuclear 2: 2 complex also contributes to the extraction into toluene. In both systems metal extraction is assumed to occur by a solvating mechanism. Neither Walker and Bautista nor Inoue and co-workers give values for the reaction constants. The main difference from our results is related to the formation of the binuclear species reported by Inoue and co-workers, and this may be due to the different ranges of Pd(I1) concentration used, which in our case were limited by the solubility of Pd(I1) in aqueous solutions containing thiocyanate. Furthermore, thio-
487
cyanate provides a different chemical environment, which may result in a different speciation. Species with 1: 2 metal : extractant stoichiometry are also found when triphenylphosphine in 1,Zdichloroethane is used as the extractant (in the absence of thiocyanate).” Our results also partially agree with those reported in the literature for the extraction of different metals with Cyanex 471x, e.g., Ag(1) is extracted from nitrate solutions as AgNO, *2TIBPS,6 and Hg(I1) from chloride media as HgCl,.2TIBPS,’ and a recent study concerning Au(II1) extraction from hydrochloric acid media with Cyanex 47 lx showed that the species AuC13*TIBPS and AuCl, *2TIBPS are responsible for the gold extraction.” In our study the effect of thiocyanate on the extraction of Pd(I1) from chloride solutions has been shown to be related to the formation of mixed-ligand complexes PdCl, _ , SCN:- in the aqueous phase, which react more rapidly than the corresponding tetrachloro-complex PdCl:- . As a result an increase in the rate of Pd(I1) extraction is obtained (see Fig. 1). The trends observed at the different thiocyanate concentrations may be related to the similar trend in the relative concentrations of the species PdCl,(SCN):- and PdCl(SCN):- under the corresponding experimental conditions. Such behaviour has previously been explained as a catalytic process based on a trans-ligand effect.’ However, taking into account our findings that thiocyanate is present in the extracted metal species, the effect of the thiocyanate should now be considered as a particular case of synergism rather than catalysis, although a trans-effect of the thiocyanate in the mixedligand aqueous complexes may well be important in formation of the metal species finally extracted. Acknowledgements-The
present work is part of the CAICYT project MAT88-752. M. H. also thanks the Catalan Institution CIRIT for its financial contribution towards the equipment needed to carry out this work.
REFERENCES 1. F. E. Beamish, The Analytical Chemistry of the Noble Metals, Pergamon Press, London, 1966. 2. R. D. Walker and R. Bautista. Proc. Int. Solo. Ext. Con& LSEC-86, Munich, 1986, Paper 11-107. 3. K. Inoue and Y. Baba, ibid., Paper 11-263. 4. Y. Baba, M. Ohshima and K. Inoue, Bull. Chem. Sot. Japan, 1986, 59, 3829. 5. Y. Baba and K. Inoue, Ind. Eng. Chem. Res., 1988.27,
1613.
488
M. HIDALGCI er al.
6. K. Inoue, Y. Baba and M. Tagaki, Proc. Inr. Solu. Ext. Cod, ISEC ‘86. Munich, 1986. Paper B-65. 7. Y. Baba, Y. Umezaki and K. Inoue, Solu. Ext. fan Exch., 1986, 4, IS. 8. S. J. Al-Bazi and A. Chow, Tufanra, 1986, 31, 815. 9. S. J. Al-Bazi and H. Freiser, Solo. Ext. ion Exch., 1986, 4 1121. 10. Some Laborarory Methodr, Inorganic Chemistry, The Royal Institute of Technology (KTH), Stockholm, 1959. 11. A. I. Vogel, Textbook of Quantitative Inorganic Analysis, 4th Ed., Longmans, London, 1978.
12. K. S. De Haas, J. Inorg. Nucl. Chem., 1973, 35, 3231. 13. A. A. Biryukou and V. I. Shlenskaya, Russ. J. Inorg. Chem., 1967, 12, 1362. 14. D. H. Liem, Acta Chem. Scud, 1971, 25, 1521. 15. L. I. Elding, Inorg. Chim. Acta, 1972, 6, 647. 16. A. K. Sundaram and E. B. Sandell, J. Am. Chem. Sot., 1955, 77, 855. 17. M. Mojski, Talanta, 1980, 27, 7. 18. V. Salvado, M. Hidalgo, A. Masana, M. Mutioz, M. Valiente and M. Muhammed, Solo. Ext. Ion Exch., 1990, 8, 491.