2965 June 5, 1955 EXCHANGE OF CHLORINE BETWEEN CHLORIDE IONAND TETRACHLOROPLATINATE(~~) partially disssolved in a depth of several molecular layers a t the surface of the crystallites. The important facts which have been established are (a) the formation of pyrophosphate upon the ignition of precipitated BCP, (b) the slow rate of pyro-
[CONTRIBUTION NO. 363, INSTITUTE FOR ATOMIC RESEARCH
phosphate formation relative to a macrophase of acid phosphate, and (c) the correlation of the amount of pyrophosphate formed with the observed ratio of Ca to Pod, as shown by Table I. WASHINGTON, D. C.
AND
DEPARTMENT O F CHEMISTRY, IOWA STATE COLLEGE]
Exchange of Chlorine in Aqueous Systems Containing Chloride and Tetrachloroplatinate (11) ‘9’
s. MARTIN,JR.
BY LEROYF. GRANTHAM, THOMAS s. ELLEMAN AND DON RECEIVED DECEMBER 23, 1954
The kinetics of exchange of chlorine has been measured in aqueous solutions containing C1-, [PtClr]’ and [PtC13(H20)]by the use of elJE. Addition of [Pt(NHs)r](NOa):! to the solutions yielded precipitates of [Pt(NH3),] [PtCl,] (Magnus’ green or pink salts) and [Pt(NH3)rl [PtCL(H20)]2without fractionation or induced exchange. Exchange occurred by means
of the reversible aquation process: [PtClr]’ e.u., AH,* = 15 kcal. and AS2* = C1- and [PtC13(H*O)]- with a rate,
- 18 e.u. R
+ H20
ki
[PtC13(HzO)]-
k?
+ Cl-, for which AH,*
= 21 kcal., AS,* = -8
In addition to this process one other detectable exchange occurred between = K‘[PtCla(H20)]-, for which AH* = 25 kcal. and A S * = 2 e a .
Introduction The exchange of chlorine between chloride ion and the [PtC14]- ion in aqueous solution has been found to proceed with a conveniently measurable rate a t room temperatures. The tetrachloroplatinate(I1) ion, representative of the square, planar complexes of platinum(II), is inert toward substitution in the sense defined by T a ~ b e . Although ~ it has been recognized that water or hydroxide can be substituted for chloride in this it is surprising that for so familiar a system a quantitative treatment of such aquation (or hydrolysis) has not appeared in the literature. The rate of exchange of a ligand in a complex with its corresponding entity in solution gives very fundamental information about the nature of the species in the solution. The work of Adamson, Welker and Volpe,6 who observed diverse rate of exchange between cyanide ion and the cyanide ligand in a number of coordination complexes, illustrates the use which can be made of these techniques. Grinberg and Nikol’shaga6 have studied the exchange in systems with a number of the square platinum(I1) complexes. They reported that the exchange rate decreases in order with the ligands: CN- > I- > Br- > C1-. For the tetrachloroplatinate(I1) ion the exchange periods were reported to be long compared to the 38-minute half-life of the C13*isotope available. The abstract of their work does not mention the possibility of a photo-effect
on the rates of exchange, although this effect has frequently been noted in exchange work with related compounds. For example, Adamson and Grunland’ indicated that the exchange of Br between [PtB ~ G ]and - Br-, undetected in the dark, was exceedingly rapid in light. Rich and Taubes recently have reported studies of the exchange of C1 between [PtClaJ- and C1and of the exchange of Pt between [PtCldj- and [PtCla]’ with a discussion of the role of PtrrX. They had previously described the exchange system [AuC14]-: Cl-.9 The complex [AuC14]- is closely related to [PtC&]= in that i t also exists as the square, coplanar complex. A number of observations by Flikkema and Hammond1° have been of great assistance in the interpretation of the results of the present exchange experiments. They found t h a t a chloride ligand of [PtC14]- in its aqueous solution was replaced reversibly by HzO with a reaction period of a few hours according to the reaction li?
[PtClrl’ 4- HzO
[PtCla(HzO)]-
+ C1-
(1)
12,
The resulting ion, [PtCI,(H,O)]-, is a weak acid which can be titrated by hydroxide with a glass electrode. The extent of aquation can be followed by measuring the acid titer of the solution. Flikkema and Hammond found that the replacement of chloride by water was first order in [PtC14]=,and the replacement of water by chloride in the reverse reaction was first order in [PtC13(H20)]-and first order in C1-. In addition, a second aquation occurred with a much longer period. The formation of [PtClz(Hz0)2] was catalyzed by hydroxide, and the formation of appreciable concentrations of this diaquo-species was accompanied by the formation
(1) Presented before the Division of Physical and Inorganic Chemistry a t t h e 126th National Meeting of the American Chemical Society, New York, N. Y., September 17, 1954. (2) Work performed in t h e Ames Laboratory of t h e Atomic Energy Commission. (3) H. Taube, Chem. Reus.. 60, 60 (1952). (4) J. N. Friend, “Textbook of Inorganic Chemistry,” Vol. IX, P a r t I, 2nd Edition, Charles Griffin and Company, Ltd., London, 1922, pp. 286-288. (5) A. W. Adamson, J . P. Welker and M. Volpe, THIS JOURNAL, (7) A. W. Adamson and J. M. Grunland, TEIS JOURNAL, 75, 5508 72, 4030 (1950). (1951). (6) A. A. Grinherg and I,. E. hTikol’shaga, Zhur. Prik!ad. Khim., (8) R. L. Rich and H. Tauhe, ibid., 76, 2608 (1954). 2 2 , 542 (1949); 24, 893 (1051), C. A . , 46, 236011 (1951); 47, 4709a (9) R. L. Rich and H. Taabe. J. Phys. Chem., 68, 1 (1954). (1963). (10) D. S. Flikkema and G. S. Hammond, private communication.
in a volumetric flask with a larger amount of HzO a t the bath temperature. The desired quantity of solution was pipetted quickly into the reaction vessel. The time a t which solutions were mixed was the zero time for the cschange. With aged solutions of L P t C l 4 , the desired amounts of KCI, LiCIOa, HCl and K;.PtClr were dissolved in water i l l ;I volumetric flask and dlowed to stand for more that1 20 Experimental Materials.-Platinum was obtained ab the foil from J . hours. The chloride tracer, a l ~ a y scontaining less thall 2y0 of the total chloride iii less than l ) . O l ~ oof the total Bishop and Co. and as chloroplatinic(1V) acid from the volume, w a s added from it micropipet. The time a t which Mallinckrodt Chemical Works. The metal was converted the tracer was ~rdcled!vas the zero time for exchange in this to chloroplatinic(1V) acid by treatment with aqua regia followed by repeated evaporations with hydrochloric acid. case. A separate reaction mixture was prepared for each deterKQPtClawas precipitated by the addition of potassium carmination of exchange. I-nder each set of conditions, ten bonate and potassium chloride. to twenty reaction mixtures were prepared. At various Potassium tetrachloroplatinate( J I ) was prepared by the times following the mixing, either a portion of solid [Pt oxalate reduction of IGPtC16, originally suggested by Vezes.” : : its aqueous solution in excess was added It was found t h a t precipitated platinum served as a very ( N H I ) ~ ] ( S O J )or to the mixture; antl a precipitate began to form within 3 0 effective catalyst when a KzPtC16 slurry was refluxed with seconds. -4fter five minutes the precipitate was collected stoichiometric amounts of K2C204. The effectiveness of the platinum catalyst for this reduction has subsequently on an “S&S Red Ribbon” filter paper supported on :t been reported by Mikhelis.Iz The product was recrystal- sintered glass disc. The precipitate was washed three time.\ with H?O, three times with alcohol, and three times with lized three times and finally precipitated from an acetoneether. It formed :I very fine, evenly distributed deposit 011 alcohol mixture. A spectroscopic analysis of the salt indithe paptr. The samples were dried for at least one hour a t cated a trace of Ag, Cu, P d and a very faint trace of Ir. 70’ and then allowed t o cool in the air for at least another [Pt(NHl)r]Cln was prepared from KzPtCla by the method hour. To correct for vctriations in room humidit>-, which of Keller.I3 The chloride was converted t o the nitrate by passing its solution through a column of Dowex 1 anion- were found to give a tletectahle effect. tared filter papers exchange resin in the nitrate phase. KO chloride was de- were simultaneously carried through t h e same washing, drytected by the addition of silver nitrate to the [ P t ( S H 3 ) r ] ing and cooling operations. A statistical study showed that filter paper :tiid \wights could be reprotlured to about ~ k 0 1 ( S O 3 ) * solutions. Platinum was normally recovered by a magnesium or mg. After being weighed, samples were inouiitetl on cardboard formic acid reduction in HC1 solution and re-used in the squares and covered with standard Cellophane sheets. The experiments. Chlorine-36, half-life 4.4 X 105 J - , , decaying by a single samples were handled gently and mounted so the deposits 0.72 MeV.-&disintegration, was the tracer isotope. The remained uniform. Counting times were chosen so t h a t the counting rates ahove background would have a statistical radioactive chlorine was obtained from the Isotopes Divistandard deviation of 1 to .?yo. Counting rates were always sion, U.S.A.E.C., Oak Ridge, Tennessee, and was carried less than 4000 cts./min. antl were corrected for the counter in dilute HC1. As received, the specific activity was redead-time of 300 p sec., reported by the manufacturer. ported t o be 0.23 millicuries/g. The specific activity of a precipitate, S . was taken as the Lithium perchlorate which was used to adjust the ionic actual counting rate (corrected for self-scattering to a staiitlstrength was obtained from G . Frederick Smith Chemical ard sample thickness) divided h y the weight of the chlorinc Company. All the water used for the experiments was re(mg.) present in the sample. S , !vas the limiting specific distilled from alkaline potassium permanganate solutions. activity of the samples which n.as determined empirically Other chemicals and reagents were of analytical reagent for each set of reaction condition.: The “aDparent” fractioii .. grade quality and met A . C . S.specifications. of exchange, j,was defined as .Y 5 , . Equipment.-The temperature of the exchange reaction Nature of “Magnus’ Salt” Precipitates.-‘lh additioii was adjusted by placing the reaction vessels in a water-bath of the cation, [Pt(NHl)rl +-, to it solutioii with I P t C l r P controlled by a mercury thermoregulator t o =t0.Io. E x normally produces‘ v e r i quickly a i l exceedingly i;isolub’lc ploratory experiments indicated that exchange rates were h f a g ~ i u s , ‘for ~ increased somewhat by light. Therefore, all exchange green precipitate, originalij, I)repared which the formula [Pt(KHs)4]jPtCI+Jhas becn establishetl solutions were contained in glass tubes, 17-18 m m . in diameter, covered by a wrapping of black. opaquc tape. The Addition of [Pt(SH:,)r](N0.,)2to the cxchange solutioiii solutions were introduced by means of a pipet through a would therefore he expected to remove effecti\-ely the chlorine in [PtCI,]= from the solutioii, atid exploratory experitwo-hole stopper. After the last solutiori was added and ments shoil-ed that exchange of chloride with green precipishaken gently to mix, the holes were covered with tape, tates was negligible over a period of Id minutcs. The beand the tubes were placed in the thermostated water-bath. havior of the [PtCla)H20]- ion in these precipitations has The radioactive samples were counted with an end-window not been previously evaluated. Green precipitates n-erc Geiger-Muller counter, Model TGC-1 of Tracerlab, Inc. prepared from aged solution containing [PtC1.1(H20)] - and The counter was operated iii conjunction with a commer(PtCl,]= in a ratio of 3 to 2. The weights of the precipitate\ cial scaler, Berkeley Instrument Co., Model 100. Samples frequently were 807, of the weight which would have becn were mounted on cardboard and placed in a reproducible manner on a shelf at a distance of 7 mm. from the counter observed if all the platinum in anions had been present a s [PtCl,]=and collected :LS [ P t ( d l [PtCI,I. Since the h \ , window. A self-scattering function of the C P activitl- in drogen ion concentrat ioii inr ~tccom~ian!-ingt h e iirrjPt( NH:j)4][PtCL] precipitates was determined cmpiricallj. cipitatioii in ;I solution coi llg ~ t b o u t 0 . 0 I 1nolc I . over the raiigc of samplc thicktie.;s employed. The weight - W:LS IC’;. than 1 0 - I riiolc ’ I , , it \vas cvitlciil of precipitate was chosen in the raiigc of 6.0 to 8.5 ~ng.!cm.~, /PtC1:,(H20)] for which region the variatioii of the self-scattering function ‘rzti3t,x I \vas relatively small. pH measurements were obtaincd by means of a Beckmaii AN.4l.YSIS 01’ I’XECIPII ATES TKOhI F R E S H L Y P R E P A K t i U KzModel “G” pH meter with an electrode model #290 and PtCI; SOLUl IOS5 adjusted with a standard buffer solution of pH 4 . Procedure .-In exchange experiments with freshly prepared [PtCla]’ solution the HC1 tracer solution was added to a standard KC1 solution. This tagged chloride solution, after mixing, was added to the reaction tubes in the thermostat. KlPtClr and LiCIOa crystals were dissolved in a small amount of H20 and the solution was diluted and mixed
of a visible precipitate, probably by disproportionation, to give platinum metal. Since the reversible aquation of [PtC14]* offers a path for the exchange of chloride, it is of interest to firid whether an aclditional mechanism for the exchange exists.
June 5 , 19%
EXCHANGE OF CHLORINE BETWEEN CHLORIDE I O N AND TETRACHLOROPL.4TINATE(II)
that the amount of the hydrolysis product, [Pt(NH3)4] [PtC130H], was less than 1% of the total. It appeared, therefore, that the compound [Pt(NHa)4][PtC13(HzO)]z precipitated as a component of the green precipitate and this conclusion was confirmed by analyses. Analyses of a precipitate from a freshly prepared K2PtC14 solution are given in Table I together with the calculated compositions of possible compounds precipitated from aged solutions. It is to be noted that the analysis for ammonia will give the critical differentiation between the most likely components. A number of precipitates for aged, equilibrium solutions were also analyzed; and the results are given in Table 11. The equilibrium values of [PtC14]- and [PtCl3(HzO)]- were calculated from the values of the equilibrium constant given below. The evidence indicated that the two compounds, [Pt!NH3)4]1PtCI4] and [Pt(NH3)4][PtC13(Hz0)]2, were precipitated without detectable fractionation. Infrared spectra of the precipitates from the aged solutions showed strong 0-H bands which were not apparent in precipitates from freshly prepared solutions. Additional lines were also present in powder X-ray diffraction photographs of precipitates from aged solutions.
2967
= concn. of [PtCla(HzO)]- a t steady state of chemical
x,
equilibrium
K = x w ( b -k ( a - x,)
=
equilibrium coilstant of reaction 1
Figure 1 shows the titration of a 0.0166 molar KzPtC14 solution which had aged 24 hours and which contained enough LiC104to give an ionic strength of 0.318 before aging. The titration indicates the presence of an acid with an ionization constant K , = lo-' and with an equivalence equal to about 60% of the moles of total platinum. The titration curve, indicated by the open circles, was obtained in 30-40 min. during which time additional aquation could occur as the acid was neutralized. -4 one point titration was made in which the number of equivalents required for the end-point in the complete titration curve were added to a second sample. The pH given by the solid point was observed for this case. This result showed that the T A B L E 11 acid in the one-point titrations was overtitrated. MICRO-KJELDAHL ANALYSISOF PRECIPITATES FROM ACED From the number of equivalents which gave the K2PtClrKCl SOLUTIOXS same pH in the titration curve, the overtitration is Ammonia, g., Concn. Concn. Calcd. [PtClr][PtClr (HPO)1 seen to amount to 4.2y0in this experiment. With (rnmole/rnl.) Found for mixture (mmole/ml.) this correction the concentration of acid was calcu11.3 0.0000 11.6 0.0083 lated from the equivalents of base. The concentra9.5 ,0081 10.2 ,0085 tion of acid was taken as x,, the concentration of 9.2 9.5 ,0085 ,0081 [PtC13(H20)]-. From the observed values of x, 9.6 9.5 ,0085 ,0081 the equilibrium constant was calculated to be 0.018 9.9 9.4 ,0040 ,0043 molejliter at 2 5 " . ,0073
,0010
10.8
10.6
12.01
On occasions, the precipitates were brown or pink in color. Under a low-power binocular microscope they were observed to consist of the green needles of the typical Magnus' precipitates and very fine pink plates. The pink crystals were apparently similar t o those originally described by Jorgensen and Sorensen'" and studied by Cox, et al.,lSband by Drew and Tress.le The exact color of the precipitates depended on the relative amounts of the red and green crystals. KO crystals of the pink form were ever prepared large enough for single crystal X-ray techniques. However, almost pure pink forms have been prepared from both freshly prepared and aged solutions and analyses and powder diffraction photographs have indicated that pink crystals can be formed from the tetrachloroplatinate(I1) ion and the trichloroaquoplatinate(11) ion. It was noted that strikingly different amounts of the two forms would appear in a series of precipitations in which there was no apparent difference in the conditions.
Results Aquation Reaction of [PtC14]-.-The reversible aquation of [PtC14]= according to reaction 1 affords a mechanism for the exchange between chloride ion and the chloride ligands in the complexes. Since the equilibrium constant and rate constants for this reaction were not available, some experimental determinations of these quantities at 25' were made in order that the role of this process in the total exchange could be evaluated. The following symbols were used for the treatment. Chlorine was only introduced into the system in the form of K2PtC14or KC1. a,
b x
= original concn. of [PtClr]' (before aging) = o'riginal concn. of chloride ion (before agingj = COIICII. of [PtC13(HzO)]-a1 time t
(15) (a) S. M . Jorgensen and S. P . L. Sorensen, Z. anorg. Chem., 48, 441 (1906); (b) E. G. Cox, F. W. Pinkard, W. Wardlaw and G. H. Preston, J . Chem. Soc., 2527 (1932). (10) H. D. K. Drew and H. J. Tress, ibid., 1586 (1935).
I
1
I
I
2 2 IOD 0
0
3* 1
!
!3 -
J
0
MOLES
OF PT.
~
8.0
LL
!
I
I
!
I
0.4
0.6
0.0
'
0
6.0
I/ 0
0.2
VOLUME OF BASE ADDED (M1). Fig. 1.-Potentiometric titration curve of an aged solution of [PtCla]" with XaOH a t 25"; ionic strength equals 0 318; concentration of [PtClr]- before aging = 0.0166; aging time = 48 hr.
Similar titrations at other temperatures yielded the values of K equal to 0.013 a t 15' and 0.021 a t 30". From the temperature variation of K , AHo for the reaction was calculated to be $5100 cal. From the results of Flikkema and Hammond the rates of the two reactions for equation 1 are given b y the expressions RI k i ( a - 3 ~ ) ; Rz = kpx(b X) (2) where the k's are the specific rate constants. The differential equation for x is
+
dx/dl = RI
- Ri =
If xo is the value of w a t f tion 3 is
- X ) - kzx(b + X )
k i ( ~
=
(3)
0, the solution of equa-
III { ( x
- wm.)[xo - x , - x,)~x - X ,
(XO
+ Z / ( b-__ + K ) 2+ 4 K a ] / + d ( b+K)2+4K~]) - k d - \ / ( b f K ) 2 + 4Ka =
(4)
The aquation of the [PtCldl- ion was followed in a solution prepared by dissolving a weighed quantity of KzPtC14 in H20 and diluting i t in a volumetric flask. At various times after its preparation aliquots were withdrawn from the solution and an amount of NaOH solution sufficient to neutralize 30430% of the [PtClS(H20)]- was added, and the fiH of the resulting solution was recorded immediately. The actual per cent. of neutralization was taken from the fiH plot in a complete corrected titration curve. From the equivalents of base added and from the per cent. neutralization was calculated the concentration of [PtCIa(HzO) 1- at the time of analysis. The data from one such experiment is illustrated in a plot in Fig. 2. The half-tiinc o f the approach toward equilibrium in this experiment was 2.37 5 0.05 hr. From this value and the equilibrium constant given above, the rate constants for 25" were computed t o be: k1 = 0.1 39 hr. -* and R z = 7.TO liter moles-' hr.-'.
established. The courses of two of the actual exchange reactions are shown in Fig. 3 . The plots of this figure show that any separation-induced exchange was less than 2 to 4%, a feature which was also apparent in the other experiments. The figure also shows t h a t a single exponential function describes the exchange variable (1 - f ) over the range of the observations. A summary of the results under various initial concentrations is given in Table 111. I t is to be noted that the halftime of the exchange in these experiments was nearly independent of the concentration variables. TABLE I11 EXCHANGE EXPERIMENTS WITH FRESHLY PREPARED SOLUT I O N S O F &PtC14 A T 25', I O N I C STRENGTH = 0.318 Comma CI b
Ccmrn. [PtCI,]
Concn [PtCls(HzO)]
n
'l,'~,
0.0010 ,0083 ,0010 .134 .OlfB .0026 ,0938 ,0042 ,0016 ,0638 Concentratioiis expressctl in molcs/l. 0 0100
0.268
hr.
Obsd.
Cnlcd.
Xrn
16.8 16.9 13.1 15.0
15.3 15.2 13.2
13.8
For the treatment of the exchange expcrirncnts the following additional symbols were defined. atoms of Cl36 in [PtClr]- per ml. of s o h . atoms of C136in [PtCla(H20)]- per ml. of s o h zu = atoms of Cl36 in C1- per ml. of soin. I = II z' M = total radioactive atoms per ml. = ~ c / l ( n- 2); specific activity of [PtCI
