427 1
REACTION BETWEEN PLATINUM CHLORIDE AND CARBON MONOXIDE
Sept. 5 , 19.54
Now taking the very high sulfate concentrations, where (Sn+4) 0 then C" =
(Cs,
- c")(HPSOI)KP
An alternative description which would also agree with the spectrophotometric data is suggested as
+
SnO+2 Sod' SnOSO4 Hi304
and
+
2SnOSOa
__
kl
Jr H P S ~ O ( S O ~ ) Pkz'
(5) (6)
which are combined to give
Csn = total tin concentration Then if D o b s d = ed'd ( i e . , if e3 >> ez) 1/Dobsd
Sn0+2
( ~ / P ~ C S ~ ~ ) ( ~ / K-t ~ (1 H ) ~(111) SO,)
Thus we plot 1 / D o b s d against 1/(H?S04) and obtain a good straight line, indicating that our reasoning was essentially correct. Then the data indicate that K 2 is quite small, since e3 is obviously large and hence that the complex anion is present to any extent only in concentrated sulfuric acid solutions. The electromigration experiments are in accord with the above conclusions. In 1.5 f sulfuric acid the tin migrates readily to the cathode, but in 6.0 f acid little migration is observed and that is toward the anode. At 3 f sulfuric acid some migration (as a streak rather than a spot) to the cathode is observed, which might indicate a very slow rate for reaction 3. It is believed that all these equilibria are slow, since the spectra of freshly prepared samples were about the same as in the perchlorate solutions and slowly changed to the recorded values. Equilibrium was reached in about 24 hours. Hence all solutions were aged for one to two days before spectrophotometric examinations were undertaken.
[CONTRIBUTION FROM THE
+ 2HSOa-
HSnO(SO&
K1
(7)
And in concentrated sulfuric acid we would have H&O(S04)2
+ HzSO4
H2Sn(SO4)9
+ HPO K P (8)
In view of the strong tendency toward hydrolysis exhibited by tin(IV), this explanation seems more reasonable than the previous interpretation. However, if we remember that the spectrum changes very little in perchloric acid of from 1.5 to 9 f, this second description loses some force. Unless k for the following SnO++
+ 2H30+ I-Sn+4 f 2H20
(9)
is small indeed, we should detect this reaction spectrophotometrically. Since we see no hydronium ion dependence in the spectrum, even in strongly acid solution, it seems more reasonable to assume the hydrated stannic ion is in fact the predominant species at hydronium ion concentrations above 1.5
f.
Acknowledgment.-This work is part of a program supported by the Atomic Energy Commission a t Michigan State College. EASTLANSING, MICHIGAN
DEPARTMENT OF CHEMISTRY
OF THE UNIVERSITY OF MICHIGAN]
The Reaction between Platinum Chlorides and Carbon Monoxide. The Preparation and Configuration of Dichlorodicarbonylplatinum(I1) BY J. M. LUTTON AND R. W. PARRY RECEIVED APRIL13, 1954 Pure dichlorodicarbonylplatinum( 11) can be synthesized in good yield by the reaction between PtCll and high pressure carbon monoxide a t 125". On the other hand, very poor yields of PtCIt(CO2) are obtained a t all temperatures and pressures if PtCL is the starting material. Pure PtC12(C0!2 loses CO irreversibly in vacuum or under nitrogen, but is stable under a CO pressure of one atmosphere. The relationship of these facts to the equilibria involved is considered. A value of 4.85 f 0.05 Debye for the dipole moment of dichlorodicarbonylplatinum(I1) in benzene solution confirms an earlier approximate value and eliminates the trans structure as one of the principal forms. Kurnakov's thiourea test also implies a cis structure for the compound, but results of the chemical test are not unequivocal. A preparation of PtClz from commercial PtO? is described.
The unusual chlorocarbonyls of platinum were first prepared by Schutzenberger2 using the reaction between platinum sponge a t 250" and a rapidly flowing gaseous mixture of chlorine and carbon monoxide. The reaction, which has been studied in this and other laboratories, proceeds rapidly to give a mixture of the three known chlorocarbonyls in good yield provided active platinum sponge, such as is obtained from the thermal decomposition of ammonium chloroplatinate, is used. A number of modifications of this procedure have been described. Phosgene gas has been used in (1) Abstracted in part from a dissertation submitted by J. Lutton to the Horace H. Rackham School of Graduate Studies of University of Michigan in partial fulfillment of the requirements the degree of Doctor of Philosophy. (2) P. Schutzenberger, A n n . Chcm. Phrs., 16, 100 (1868); ibid., 350 (1870).
M. the for 21,
place of chlorine and carbon monoxide with relatively poor results.3 I n a second modification platinous chloride has been treated with CO. Literature reports on this procedure are not consistent; an excellent modern reference book4 reports that PtC12 takes up CO more readily than does platinum sponge, yet preparative procedures involving the use of PtC12 have been relatively ineffe~tive.~ I n view of these observations the reactions between carbon monoxide and the chlorides of platinum have been reinvestigated in some detail with results as described below. (3) F . Mylius and F . Foerster, Ber.. 24, 2424 (1891). (4) N. V. Sidgwick, "Chemical Elements and Their Compounds," Vol. 11, Oxford University Press, New York, N. Y.,1950, p. 1627. ( 5 ) (a) W. 0. Walker, Thesis, University of Chicago (1930); (h) this investigation.
At low temperatures (100-125") and all pressures PtC12 is inert to CO. When the temperature is raised (150-225') and reaction begins, the principal product a t all pressures is metallic platinum containing adsorbed carbon monoxide. Only small amounts of the dichlorodicarbonylplatinum(I1) were ever obtained using olive-drab PtClz as a starting material. I n contrast, the black powder of empirical composition PtC13 gave 50 to 75% yields of crystalline [PtCl,(C0)2]when treated with CO a t 40 atmospheres pressure and 125". KO lower carbonyls were present. Variations in starting material, temperature and carbon monoxide pressure produced the effects summarized in Table I.
carbon monoxide. In the high pressure synthesis the dichlorodicarbonylplatinum( I 1) is the sole product; in the low pressure synthesis a niixturc of lower carbonyls is obtained. Besides the dimerization reaction, active PtClzCO can undergo irreversible internal reduction to give Pt and COCl, (equation 6)6or it can lose its last CO to give active PtCl2 (equation 3 ) . The active PtC12 may c o i i dense to give the non-reactive polymer (equation 5 I or it may be rapidly reduced by CO to give Pt and COC12 (equation 4). In the presence of CO above 130" the reduction is a rapid process. No significant amounts of PtClz have ever been isolated from thc decomposition products of PtCl2(C0)*,but the reverse process-the formation of coordination coniTABLE I pounds from PtCl,-proceeds rapidly above 200" Platinum Temp , C O pressure, to give essentially quantitative yields of PtC1,chloride 'C dt m Result (PF3)zwhen the non-reducing PFJ is used in place PtClg 100-125 1 S o reaction and 150-200 1 About 2 to 3% PtCl?(CO), of the reducing CO. The reactivity of PtC& a t high CO pressure and Largely Pt metal PtClJ the inert character of PtC12 must be due to lattice Above200 1 Pt metal factors. PtClz is a surprisingly inert material, prePtCI2 100 40 to 150 S o reaction sumably due to its chain-like structure. Reactiv123 40 t o 150 3-3% PtClI(CO)? ity of the system is enhanced by breaking up this Some P t metal structure as in PtC13. The formation of the mixed 150 40 to 150 P t metal carbonyls from active platinum, carbon monoxide and chlorine is also reasonable since in this case the PtCla 100 120 Yo reaction 125 40 to 120 tjO-75cT, yield PtCl>(CO)J non-reactive variety of PtClz does not form; instead the active monomer reacts rapidly with CO a t one 150 40 to 120 I't metal with trace atmosphere to give the observed mixture of carbonPtCl,( C 0 ) J yls. These are swept out of the heated region beI n view of the fact that both platinum(I1) chlo- fore decomposition. The knowns role of spongy ride and [PtClZC0l2are inert to carbon monoxide platinum as a catalyst for the high temperature a t one atmosphere pressure and room temperature, combination of CO and Clz to give COCl, is also it was rather surprising to find that an atmosphere suggested by the foregoing relationships (l't of CO represses the dissociation of pure [PtClZ- Clz + PtC12 and equation 5 ) . (CO)2]. The dissociation proceeds irreversibly in a The Contiguration of Dichlorodicarbonylplatinitrogen atmosphere or high vacuum. The above num(I1) and its Reaction with Thiourea.-Chatt facts suggested that a consideration of the reaction and Williamsg reported a value of 4.65 f 0.5 as the possibilities in the system carbon monoxide-plati- dipole moment of PtClZ(C0)2 in benzene solution. num carbonyl halides-platinum halides might be The result was interpreted in terms of a predomiinteresting. nantly cis structure. An independent and more preThe Dissociation Process.-The experimental ob- cise measurement, completed in this Laboratory servations can be interpreted in terms of the follow- concurrent with publication of the data of Chatt ing equations and Williams, gave 4.85 + 0.05 debye. X molecular weight measurement on the benzene solution COII[PtC1*C0]2 (less reactive form) firmed the monomeric formula for PtC12(CO)2.T h e cis configuration is indicated. Chemical tests for configuration are also of interroom temp. est. Kurnakov'O noted that cis-dichlorodianiminePtCL(C0)2 PtCI?CO co platinum(I1) reacts with aqueous thiourea to give form, the tetrathioureaplatinum( 11) chloride as yellow robably monomer) needles while the trans isomer reacts to give a colorless solution from which snow white needles of dithioureadiammineplatinum(I1) can be recovered on 5 4 [PtCM cooling. This test has been suggested as a means of Pt COCls + PtCL CO 4 (non-ryactive (reactive form) above lattice) solid differentiating cis and trans isomers of platinum(I1) 150' lattice complexes. As pointed out by Quagliano and Schu(6) T h e conversion of [PtChCO] to Pt and COClz is assumed t u be From equation 1 i t can be seen that CO will reor a n equilibrium which lies well toward the P t since earlier press the decomposition of the dicarbonyl a t room irreversible observationsaindicate t h a t Pt and COClr d o not react easily over a wide temperature ; but since the dimeric crystalline solid temperature range t o give platinum chlorocarbonyls. [PtC12C0]2 is not reactive a t room temperature (7) (a) J. C h a t t , .VValure. 166, 637 (1950); (13) confirmed in t h i s (equation 2 ) , the CO will not convert crystalline Laboratory. J. N. Friend, "Textbook of Inorganic Chemistry," Vol. 5 , 1921, [PtC12C0]2to PtC12(C0)2. At somewhat higher p . (8) 98. temperatures (125') the conversion proceeds rap(9) J. C h a t t a n d A. A. Williams, J. Chem. SOL.,3001 ( 1 9 5 3 ) . idly, the extent depending upon the fugacity of the (10) S. Kurnakov, J . prnbl. Chem., 121 50, 483 (1948).
+
,
+
+
+
-,
Sept. 5, 1954
REACTION BETWEEN PLATINUM CHLORIDE AND CARBON MONOXIDE
bertll no general check on the applicability of the test using substances of known configuration has been recorded. Rapid mixing of a dry benzene solution of the carbonyl and a saturated aqueous solution of thiourea results in rapid formation of the yellow needles of Pt(Th)rClz. No reduction or formation of other platinum compounds could be detected. The tempting conclusion t o be drawn from the result described is that the presence of the cis isomer is established by chemical as well as physical means, but closer analysis suggests that the chemical test may not be reliable under certain conditions. If one accepts the usual assumption that the CO group has a strong trans effect, comparable to or greater than C1-, either CO or Cl- could be eliminated as the first step in the reaction, the choice between the two depending upon the relative strength of the metal ligand bond after labilization. Available evidence seems to indicate that the CO is liberated first since the related [Pt CO Py Clz] is a known compound. The initial reaction for the cis isomer is then
c1
)Pt(zz
c1
+ Th + cc11
This would be followed rapidly by replacement of the CO and C1-. Since all attached groups are now trans labilizing, Pt(Th)4Clz would be expected as a product just as in the case of the cis Pt(NH&C12. For the trans isomer, CO might again be liberated as in the initial reaction, since the two CO groups are assumed to be trans labilizing. The reaction then becomes
co
)Pt
