Anal. Chem. 1983, 55, 910-913
910
obtained in less than 10 min. However, as the set of differential equations became stiffer, the computation time increased. The set of differential equations for the model whose results are shown in Figure 7 are considerably stiffer than those for the other simulations and, as a result, the solution required 1 h of computation time. While such long computational times currently limit the usefulness of on-line simulation when using 16-bit computers, such limitations will be drastically reduced in the next 2-5 years when the price of 32-bit computers drops to where they can cost effectively be utilized for on-line experimentation and control of complex systems. CONCLUSIONS
It is often difficult to assess the effects of changes in the various parameters of complex processes and complex chemical experimentation. A very useful aid is a simulation study particularly when it can be performed easily and quickly on the computers used concurrently on the experiment. One such system to aid in the design of enzyme kinetic experimentation was described, and the results of several simulations were discussed. It is predicted that when inexpensive 32-bit computers become available, on-line graphics-supported simulation studies will become a powerful tool to aid in experimental design and the automation of complex processes. LITERATURE C I T E D (1) Frazer, J. W.; Rigdon, L. P.; Brand, H. R.; Pomernacki, C. L. Anal. Chem. 1979, 51, 1739-1747. (2) Frazer, J. W.; Rigdon, L. P.; Brand, H. R.; Pomernacki, C. L.; Brubaker, T. A. Anal. Chem. 1979, 51, 1747-1754.
(3) Frazer, J. W. Anal. Chem. 1980, 52, 1205A-1212A. (4) Frazer, J. W.; Brand, H. R. Anal. Chem. 1980, 52, 1730-1732. Frazer, J. W. Am. Lab. (F8h'f/dd, Conn .) 1981, 13 (4), 60-78. , Hindmarsh, A. C.; Byrene, G. D. "EPISODE: An Effective Package for the Integration of Systems of Ordinary Differentlal Equations"; UCID 30112, Rev. 1, 1977. Readers who desire further information on LLNL internal documents should address their inquiries to the Technical Information Department, Lawrence Livermore National Laboratory, Livermore, CA. (7) Segal, I.H. "Enzyme Kinetics"; Wiiey: New York, 1975.
12 I
RECEIVED for review September 27, 1982. Accepted January 10, 1983. Work performed under the auspicies of the U S . Department of Energy by the Lawrence Livermore National Laboratory under Contract No. W-7405-ENG-48. This document was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor the University of California nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial products, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or the University of California. The view and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government thereof, and shall not be used for advertising or product endorsement purposes.
Determination of Extraction Constants of Thallium(I ) Dithiocarbamates Govindaraju Soundararajan
and Masilamanl Subbaiyan
Department of Analytical Chemistty, University of Madras, Guindy, Madras 600 025, Tamil Nadu, Indla
Extraction constants of dlthlocarbamates of thallium( I), having dlfferent N-alkyl substltuents, have been determined by studying thelr exchange reactions wlth dlthlzone In carbon tetrachlorlde. Extractlon constants, so determined are correlated wlth their chromatographic mlgratory behavlor (on thin layers of slllca gel), N-alkyl chaln length, molar solublllty In carbon tetrachiorlde and their characteristic Infrared frequencies. Application of these chelates in carbon tetrachloride as extractants for some elements wlth higher extraction constants (Fe, Co, Zn, and Cd) and their stablllty In carbon tetrachlorlde and toward acid aqueous solutions (HC104 and H,S04) are studled.
Colorless and neutral metal chelates of N,N-disubstituted dithiocarbamates in nonpolar solvents are widely employed as extractants for selective separation and estimation of sulfophilic elements from acid solutions with less danger of acid decomposition (I). Dithiocarbamates of zinc, cadmium, and lead are frequently employed as extractants in trace analyses (2). Though the univalent metal dithiocarbamates are more favoered as extractants than di- or trivalent metal
chelates ( 3 ) ,only a few works have so for been reported (1, 4, 5) employing thallium(1) diethyldithiocarbamate as extractant. Detailed knowledge about the extraction constants (Kex)is an essential prerequisite for the application of metal chelates as extractants in trace analyses. But the K,, values reported (2,5-8) for several metal dithiocarbamates are insufficient and irreproducible. Also the dependence of K,, values on N-alkyl substituents of dithiocarbamates have not been studied so for. This work reports the K,, values of thallium(1) dithiocarbamates (Tl(dtc))with different N-alkyl substituents by spectrophotometrically studying their exchange reaction with dithizone (H,Dz) in carbon tetrachloride (CC14) solution (3, 8). Since the extractive separation and chromatographic separation go hand in hand, K,, values of Tl(dtc) are correlated with their migratory behavior on thin layers of silica gel. Some metal ions (Fe(III), Co(III), Zn(II), and Cd(I1)) whose dithiocarbamates have greater K,, values than Tl(dtc) are extracted and determined with the latter as extractants. THEORY Ligand Exchange Equilibrium. Neutral metal dithiocarbamates in organic solvents undergo ligand exchange reactions when suitable ligands are added. Exchange constants
0003-2700/83/0355-0910$01.50/00 1983 American Chemlcai Society
ANALYTICAL CHEMISTRY, VOL. 55, NO. 6, MAY 1983
.-
.
Table I, log K,, Values and Some Other Properties of Thallium(1) Dithiocarbamatesa Me,dtc
*
log E log Kex. log S M Rmd A B C
Et,dtc
2.99 i: 0.02
JJ
slopee
Pr,dtc
-0.51
0.01
-2.05 0.34 0.47 0.51 1.40 1.54
-0.88 -0.28 -0.06 0.04 0.85
Bu,dtc f 0.03
4.09 0.59 2.18 -1.24 -0.95 -0.92 0.27 0.83
3.76 i: 0.03 0.26 -0.19 -0.77 -0.44 -0.32 0.54 1.06
3.51 i: 0.01
1.11
911
(PiP)(dtC) 3.11 i: 0.05 -0.39 -1.96 -0.08 0.18 0.28 1.28 1.33
a Key: Me,dtc = dimethyldithiocarbamate; Et,dtc = diethyldithiocarbamate; Pr,drc = di-n-propyldithiocarbamate;Bu,dtc = di-n-butyldithiocarbamate; (pip)(dtc) = 1-piperidinecarbodithioate. E ( T M D z a d k ) , the error presented is the average of six independent measurements with varying initial concentrations of Tl(dtc) and H,Dz in 20 mL of CC1, at 30 "C. Molar R , = thin-layer chromatographic migratory parameter with: A = benzene, B = toluene, C = solubility in CCl, at 30 "C. xylene, and D = CCl, as eluents. e Slopes of R, vs. log X, plots, where X, is fraction of active component-benzene in thLe
binary mobile phase with CCl, as diluent. of these reactions are determined from the equilibrium concentrations of the reactants and products or from the extraction constants of the corresponding metal chelates (9). When HzDz in CCl, is added to Tl(dtc) in CC14,the following equilibrium is established.
Tl(dtc) f HzDz + TlHDz
+ Hdtc
(1)
The equilibrium constant of this exchange reaction can be represented as E(TIHDz.Hdtc)
--
[Tl(dtc)l [HzDz] -=[Tl(HDz)l [Hdtcl
KTl(dtc) KT~(HD~)
(2)
where
LOO
I
I
500
603
I 700
Wave Length ( n m )
(3) and
JKTKHD~) =
[TI(HDz)Iorg[H+Iaq [T1+laq[H&Iorg
(4)
log K,, of Tl(HDz) is reported (10,11) to be -3.5. From this, the extraction constant of Tl(dtc) is calculated from the equation E(TIHDz-Hdtc)
-3*5 = log KTl(dtc)
(5)
The equilibrium concentrations of HzDz and TlHDz are determined spectrophotometrically, by measuring the absorbances at 620 nm and 505 nm, respectively, Figure 1. Equilibrium concentrations of Tl(dtc) and Hdtc are calculated from the following molar equalities: [Tl(dtc)],, = [Tl(dtc)l0,ig -- [TlHDz] [Hdtc],, = [T'l(dtc)lo,ig- [Tl(dtc)],, = [Tl(HDz)]
(6) (7)
or [Hdtcleq = [H~Dzlorig- [H~Dzleq= [Tl(HDz)l
(8)
ElXPERIMENTAL SECTION Apparatus. A Beckman Model 25 UV-visible spectrophotometer was usled. Reagents. All reagents employed were of analytical grade. HzDz was purified by solvent extraction with CCl, as reported earlier ( 1 0 , I I ) . HzDz (-lo4 M) was diluted appropriately before use with freshly distilled CC14 and purity was checked spectrophotometrically at 620 nm (log e 4.52). CCl, was purified as reported earlier (12) and twice distilled before use every time. Thallium(1) dithizonate (TlHDz) in CCl, was prepared in substoichiometlric condition, using an excess of TINO:, relative
Figure 1. Absorption spectra of H,Dz and TlHDz In CCI, at 30 'C: (a) H,Dz (2.6 X 1 M); (b) TlHDz (1.5 X 10-5 M).
to HzDz. The concentraction of TlHDz was verified from the amount of HzDzused and by measuring the absorbance of TLlHDz at 505 nm (log E 4.53). Sodium salts of dithiocarbamic acids were prepared hom methyl-, diethyl-, di-n-propyl-,and di-n-butylaminesand also from piperidine following the procedure reported (13)and recrystallized from isopropyl alcohol. The purity was checked after their conversion into copper chelate by spectrophotometry after appropriate dilution with CCl, at 436 nm (log e 4.11). Tl(dtc) were prepared by treating T1' ion with sodium salt of the ligand in 1:l molar proportions. The white precipitates collected were repeatedly recrystallized from chloroform. The purity was checked by elemental analyses, melting point determinations (13,14) and spectrophotometry as copper(I1) chelates (15). Elemental analyses confiied the monochelate compositions. Standard solutions of Tl(dtc) (10" M to M)were prepared in CCl, and further standardized as Cu(I1) chelates. Great care was taken in maintaining the purity of CC14,H2Dz,and Tl(dtc) chelates. Solubility Determinations. Saturated solutions of Tl(dtc) in CCl, were prepared at 30 "C. The concentrations in known volume of this solution after appropriate dilution were determined by spectrophotometry as Cu(I1) chelates at 436 nm. Investigation of Exchange Reaction between Tl(dtc) and H2Dz. To known amounts of Tl(dtc) in CC,, varying quantities of HzDz in CC14were added, and the total volume was maintained constant at 20 mL. Equilibrium was established within 1-3 min by mechanical shaking at 30 "C. The equilibrium concentrations of Tl(dtc), Hdtc, HzDz,and TlHDz were determined as previously described under "Ligand Exchange Equilibrium". Extraction and exchange constants were calculated from eq 5 and 2, respectively.
RESULTS AND DISCUSSION The exchange and extraction constants of different dithiocarbamates of thallium(1) are given in Table I. Figure 2 line "a" shows the linear variation of K , values with number of -CHz- groups in N-alkyl substituents. K,, values are also
912
ANALYTICAL CHEMISTRY, VOL. 55, NO. 6, MAY 1983
Table 11. Infrared Spectral Data of Thallium(1) Dithiocarbamates a
*
u(C--N),' cm-'
v(C-;;S),' cm-'
Tl(Me,dtc) Tl(Et,dtc)
TlC3H,NS, TlC5H,,NS,
Tl(Pr,dtc)
TlC,H,,NS,
1490 s 1475 s 1410m 1460 s
965 s 982 s 910m 962 s
1 4- -1. 0 .m ..
T1(Bu,dtc)
TlC,H,,NS,
8.9 ~ 0w . 955 s
Tl((pip)(dtc))
TlC,H,,NS,
chelate
O'?
1'
t-/-7-/ [
I
-06
-04
-02
0
02
04
1
06
L o g Kex (3)
Flgure 2. Llnear variation of K,, with N-alkyl chain length and TLC "R," values: (a) Variation of log K,, values with the number of CH,groups in N-alkyl group; (b) variation of log K, values with R , values of corresponding chelates on thin layers of silica gel. Eluent = carbon
tetrachloride. found to vary linearly with molar solubility in CC14. Extraction Constants and TLC (Thin-Layer Chromatography) Migratory Behavior. Silica gel adsorbs metal chelates via H-bond formation through its -OH adsorption centers. The stronger the metal chelate, i.e., lower the electron density in the outskirts of the metal chelate molecule, the higher is its mobility during chromatography (16). This is verified by the linear relationship between K,, values and R, values irrespective of the developing solvent employed (Figure 2 line "b" and Table I). This result is similar to that reported for dithizonates, benzoyl acetonates, and salicyl aldoximates (16). Table I also shows the decrease in R, vs. log X,slopes ( X , = Fraction of benzene in CC14-binary mobile phase) with increase in K,, values. This supports the decrease in adsorption constants due to the decreased availability of lone pair of electrons on N-atom for H-bond formation, through which a "dtc" moiety gets adsorbed with -OH groups on adsorbent surface. This correlation may help as a practical guide for a preliminary estimate of the probable chromatographic migratory order of different types of other metal dithiocarbamates, provided the influence of the metal ions on the physical properties of their dithiocarbamates is known in detail. Extraction Constants and IR v(C=N) Frequencies. IR spectral data of Tl(dtc) are produced in Table 11. Infrared absorptions due to stretching vibrations of the "thioureide" bond (C=N) are observed as strong bands in the region 1470 to 1490 cm-'. Absorptions due to v(C=S) are observed in the region 950 to lo00 cm-l. The v(C=N) band position is affected mainly by the nature of the central atom and the nature of the N-alkyl group. Also it is a measure of the availability of lone pair of electrons on N-atom. Since K,, is a measure of charge density in the outskirt of molecular frame, it may be correlated with IR v(C=N) frequencies. K,, is found to increase with decrease in (C=N) frequencies from dimethyl to dipropyl (also piperidyl) chelates. Though inductive effect increases with increase in alkyl chain length, the growing mass of the alkyl groups kinematically retards the hyperconjugation of a-CH2groups and thereby decreases C=N bond order (17). The greater v(C=N) frequency of T1(Buzdtc) may be due to the competing kinematic and electronic factors on vibrations. The lower v(C=N) of Tl((pip)(dtc))compared to that of T1(Et2dtc)is due to the controlled release of electron pair from N-atom due to former's rigid ring structure. Stability of Tl(dtc) Solutions in CC1& The M to M solutions of these chelates in CCll were found to be
mol compn
1470 s 1410m 1470 s 1422m
925m 1000 s 965s
a Spectra recorded with Perkin-Elmer 199 IR spectrophotometer (scanned region = from 3500 to 200 cm'' in chloroform). Molecular compositions obtained from the results of elemental analyses, Carbon and hydrogen estimated with a Perkin-Elmer 240 B analyzer. Nitrogen estimated by micro-Kjeldahl method. Sulfur estimated by microgravimetry as barium sulfate. Key: s = strong, m = medium, and w = weak absorption bands.
*
Table 111. Decomposition of Tl(dtc) in CC1, in Contact with Acid Aqueous Solutions chelate (molar concn)
(acid soln)
Tl(Me,dtc) (1.12 x 10-3 M ) Tl(Et,dtc) (1.78 x 10-3 M ) T1(Pr ,d t c ) (1.57 X M) Tl(Bu,dtc) (2.21 x 10-3 M ) TU (pip dtc 1) (9.43 x 10-4 M )
1 0.1 1 0.1 1 0.1 1 0.1 1 0.1
half-life, min H,SO, 6 100 8 121 9 139 19 288 7
115
HC10, 4
88 5 92 8 96 13 120 3 84
stable for over a period of 2 weeks, when stored in colored bottles. No deterioration of their titer exceeding the experimental accuracy of 2.5% was observed. Stability in Contact with Acid Aqueous Solutions. Table I11 shows the stabilities of various Tl(dtc) in half-life period (in minutes) toward acid aqueous solutions of HCIOl and HzS04. Stability is found to increase with increase in N-alkyl chain length, supporting earlier findings (18, 19). T1(Bu2dtc)is found to be the most stable toward acid solutions among the chelates employed. The lengthy n-butyl substituent resists the approach of HCtoward N-atom, which initiate the acid decompositions. The compact piperidyl group in Tl((pip)(dtc)), though more basic than diethylamine in T1(Etdtc), provides room for easy approach of H+ and as a result is less stable. All solutions in CC14showed a decrease of titer, which followed fist-order kinetics. In general stability is found to increase with K,, values. Thallium(1) Dithiocarbamates as Extractants. Table IV shows the results of application of Tl(dtc) in CC14 as extractants for Fe(III), Co(III), Zn(II), and Cd(I1) ions from 0.5 N HC104 solutions. For an equilibration period of 15 min the extraction efficiency is found to increase with increase in N-alkyl chain length. Stabilities of dithiocarbamates of these metal ions are also found to be in the order Fe < Zn < Cd < Co. Zn and Cd are quantitatively extracted within a minute of equilibration. But Co(II1) and Fe(II1) required 5-10 min. The rate of extraction observed was as reported earlier (20). The low values of Zn(II), Cd(II), and Fe(II1) recovered are due to the acid decomposition of the extracted chelates during the equilibration period of 15 min. Table V shows the results of analyses for Fe and Zn in some aluminum alloys with T1-
ANALYTICAL CHEMISTRY, VOL. 55, NO. 6, MAY 1983
-.Table IV. Extractions of Zn(II), Cd(II), Fe(III), and Co(II1) Ionsa with Thallium(1) Dithiocarbamates as Extractants
913
.-
quantity extracted (pg) with quantity metal ion taken, pg Me,dte Et,dtc Pr,dtc Bu,dtc (Pip)(dtc) Zn(I1) 70 48 58 63 69 46 Cd (11) 120 82 110 113 120 88 Fe(II1) 60 30 48 63 58 30 c o (111) 65 61 62 64 65 63 a Extractions from 0.5 N HC10, solutions with equal volumes of aqueous and organic (CCl,) phases (50 mL). Tl(Me,dtc) = 1.12 X M;Tl(Et,dtc) = 1.78 X lo-’ M; Tl(Pr,dtc) = 1.57 X M;Tl(Bu,dtc) = 2.21 x lo-“ M;Tl((pip)(dtc)) = 9.43 X lo-, M. Equilibration period is for 1 5 min. Fe estimated by spectrophotometry at 515 nm. Co estimated by spectrophotometry at 367 nm and 650 nm. Zn and Cd estimated after stripping into aqueous phase with 2 N HC1, by EDTA titration with xylenol orange as indicator (23 ).
-
-
~
-
I
Table V. Results of the Analyses of Some Aluminum Alloys for Traces of Zinc and Iron with Tl(dtc) in CCl, as Extractants
extractanita T1( Et,dtc) Tl((pip)(dtc))
element deter- specimined fied Fe Zn Fe Zn
1.00 0.25 1.00 0.25
%
Pound
error
0.9771 0.2452 0.9768 0.2473
2.29 1.92 2.32 1.08
a Extractions from 0.5 N NaC10, solutions. The Fe(dtc), extracted was estimated by spectrophotometry at 51 5 nm. The results confirmed with the estimations as “thiocyanati?” complex after stripping Fe(II1) into aqueous phase with 2 N HCl. Zn estimated after stripping into aqueous phase with 2 N HC1, by EDTA titration with xylenol orange as the indicator,
(Etzdtc) in CC14 as extractant. Kinetics of Extraction of Fe(II1) and Co(II1) (in substoichiometric condition (21)). Curve I in Figure 3 shows the extraction of Co(II1) with substoichiometric quantity of T1(Et,dtc) in the absence of Fe(II1). Extraction is completed in 12-15 min. Co(II1) -I- 3Tl(Etzdtc) aq org
+
+
C ~ ( E t z d t ~ ) g3T1’ 0% aq
But in the presence of Fe(III), the extraction behavior follows curve I1 requiring more than 30 min for completion of extraction. This rnay be due to the competing reaction of Fe(II1) with T1(Etzdtc).
-
Fe(II1) -I- 3T1(Etzdtc) aq Qrg
+ 3Tl’
Fe(Et,zdtc)3 0%
aq
The comparatively slow reaction rate of the exchange Fe(Etzdtc)3 + Co(II1)
Fe(II1)
+C~(Et~dtc)~
determines the final slow rate of extraction of Co(II1) under substoichiometric condition. According to Elek (22),when l / n log K,, values of two metals differ by less than 2 units, there is the poseibility of simultaneous extraction of both the metals under substoichiometric conditions. The plot of [(Fe)/(Fe(Etzdtc)3),,,]1~3vs. [(Go/(Co(Ehdt~)~),,~ ]For ] ~extractions /~ of a mixture of Fe and Co from 0.1 N KCIOl solution with different substoichiometric quantities of T1(Etzditc)in CC14resulted in a straight line with the slope 2.86 and it is found 1
/3
log Kex(co) - 73 log
I
K X ( F ~ )
= 0.46
This value of 0.46 favorably compares with the calculated value of 0.38from Kexvalues reported earlier (6). This straight
1
,
20
I
LO Time ( m i n )
I
63
Flgure 3. Extraction behavior of Fe(II1) and Co(II1) with TI(Et,dtc): (curve I ) extraction of 150 pg of Co(II1) from aqueous solution (50 mL contalnlng 0.5 N NaCIO,) at 30 OC with 50 mL 1.78 X M Ti(Et,dtc) in CCi, as extractant: (curve 11) extraction of Co(II1) ions in the presence of 60 pg of Fe(II1).
line plot does not pass through the origin. This may be (due to the formation of mixed ligand complexes with the anions present in the aqueous phase. Registry No. Fe, 7439-89-6; Co, 7440-48-4; Zn, 7440-66-6; Cd, 7440-43-9; T1, 7440-28-0;H,Dz, 60-10-6; T1(Mezdtc),14930-31-5; Tl(Et,dtc), 18756-72-4; T1(Przdtc), 18946-99-1; T1(Buzdtc), 18946-98-0;Tl((pip)(dtc)),84538-29-4. LITERATURE C I T E D Wyttenbach, A.; Bajo, S. Anal. Chem. 1975, 4 7 , 1813. Bajo, S.; Wyttenbach, A. Anal. Chem. 1979, 51, 378. Ruzicka, J., Stary, J. Talanfa 1987, 14, 909. Van Erkelens, P. C. Anal. Chim. Acta 1962, 2 6 , 32. Yeh, S. J.; Lo, J. M.; Shen, L. H. Anal. Chem. 1980, 52, 528. Stary, J.; Kratzer, K. Anal. Chlm. Acta 1988, 4 0 , 93. Shen, L. H.; Yeh, S. J.; Lo, J. M. Anal. Chem. 1980, 52, 1882. Stary, J.; Ruzicka, J. Talanta 1988, 15, 505. Stary, J.; Burcl, R. Radiochem. Radioanal. Len. 1971, 7 , 235. Iwantscheff, “Das Dithlzon und Seine Anwendung in der Mikro and Spurenanalyse”; Verlag Chemie: Wlnheim, 1958. Stary, J. “The Solvent Extraction of Metal Chelates”; Pergamon: Oxford, 1964. Hnadley, T. H. Anal. Chem. 1984, 36, 153. Coucouvanis, D. Prog. Inorg. Chem. 1971, 11, 271. Akerstrom, S. Ark. Kemi 1985, 2 4 , 495. Hulanlckl. A. Talanta 1987, 14. 1371. Gallk, A. Anal. Chlm. Acta 1973, 6 7 , 357. Oktavec, D.; Beinrohr, E.; S i b , B.; Graa], J. Collect. Czech. Chem. Commun. 1980, 4 5 , 1495. Miller, D. M.; Latimer, R. A. Can. J. Chem. 1982, 4 0 , 246. Aspila, K. I.; Sastri, V. S.; Chakrabarti, C. L. Talanta 1969, 16, 1099. Bajo, S . ; Wyttenbach, A. Anal. Chem. 1975, 4 7 , 2. Ruzicka, J.; Stary, J. “Substolchiometry in Radiochemical Analyses“; Pergamon: Elmsford, NY, 1988. Elek, A. J . Radloanal. Chem. 1970, 4 , 281. Vogei, A. I. “Quantitative Inorganlc Analyses”, 3rd ed.; Longmans: New York, 1984; p 443.
RECEIVED for review November 19,1982. Accepted Jannary 3, 1983.
