2646
HOWARD H . CADYAND ROBERT E. CONNICK
The point that a change in solvent from HzO to CHIOH causes a considerable change (four-fold decrease) in rate, but yet no effect on the rate is admitted arising out of altering the solvent water by compressing i t to a molar volume of 15 ml., requires some comment. I t can be argued that because the solvent in the neighborhood of an ion of high charge is already highly compressed, no large change in the solvent in this region is effected by increasing the pressure. Change from one solvent to another does, however, alter the immediate environment of the cation. Thus, the change in rate caused by changing solvent does not appear to be incompatible with the assumption that the pure solvent effect caused by increase in pressure does not give rise to a large change in rate. I n a number of substitution reactions, it has been observed that the rate of aquation or hydrolysis increases as the acidity decreases. The marked decline in rate of exchange for ROH2+++ as alkali is added, therefore, seems surprising a t first sight. Closer examination of the system shows that the result is not inconsistent with the observations on other complexes. A specific rate ratio of lo3 to lo5
[COSTRIBUTION FROM
Vol. 80
has been estimatedI6 for the replacement of X in (NH3)4CoNH2X+ as compared to (NH3)&oX++, and a similar ratio can be assumed for (NH3)dCoNH20Ha++ as compared to (NH,)6CoOHz+++. However, in a solution in which (NH3)6CoOH++ is the dominant form, only a small fraction is present as ( N H ~ ) ~ C O N H ~ O Hthis ~ + +fraction , being determined by the relative acidities of coordinated HzO as compared to coordinated NH3. The fraction in question may be as small as lo-’; therefore, the specific rate of exchange when (NH3)5CoOH++ is dominant is expected to be to 10+ times that for ( N H ~ ) G C O O H ~ + +Experimentally, +. i t is found that this ratio is no greater than 2 X Acknowledgments.-We wish to express our gratitude to Professor A. W. Lawson for placing the high pressure equipment a t our disposal; to the office of Naval Research for financial support (Contract N6ori-20) and t o Allied Chemical and Dye Corporation and E. I. du Pont de Nemours and Company for providing fellowship awards for H IR. H . (16) R. G. Pearson a n d F.Basolo, THISJOURNAL, 78, 4878 (1956).
CHICAGO 37, ILLISOIS
RADIATIOS LABORATORY AND DEPARTNEXT OF
CHEMISTRY, UNIVERSITY OF CALIFORNIA, BERKELEY]
The Determination of the Formulas of Aqueous Ruthenium(II1) Species by Means of Ion-exchange Resin : R u + ~R, u C I + ~and RuCI,+ BY HOWARD H. CADYAND ROBERT E. CONNICK RECEIVED OCTOBER7, 1957 The formulas of some aqueous ruthenium(II1) species have been determined by a method which utilizes the properties of ion-exchange resins. The method is applicable when equilibria between the species of interest and other species of significant concentration are established slowly, and when the species of interest can be isolated in solution, as for example certain complex ions. I n one experiment the charge per metal atom is found from the equivalents of charge exchanged with a n ion-exchange resin per gram atom of the metal. I n a second experiment the total charge per species is determined from the concentration dependence of the ion-exchange equilibrium of the unknown species with an ion of known charge. The total charge divided by the charge per atom gives the degree of polymerization of the species. The stoichiometric formula can be inferred when a single anion is involved in complexing. The method has been employed to establish the formulas A species shown to have a total charge of t l is probably RuCla+. The separation of the of the species Ru+3and RuCl’2. ruthenium species was accomplished by “pushing” with cerous ion a dilute mixture of species on an ion-exchange column, with the result that nearly pure bands of the individual species were obtained at high concentration.
Little is known regarding the formulas of aqueous ruthenium species in the +2, + 3 and + 4 oxidation states.lJ Preparations customarily yield mixtures of species corresponding to different complexes and varying degrees of polymerization, as well as mixtures of oxidation states. Because equilibria between the species are established slowly it is impractical in most cases to use equilibrium measurements to establish the formulas, Colligative dcterminations are not employed because of the difiiculty of preparing pure species. (1) (a) T h e older literature is summarized in Gmelin’s Handbuch der anorganischen Chemie, System No. 63, 8 t h Edition, Verlag Chemie, Berlin, 1938; (b) more recently D. D. Deford has critically reviewed t h e literature in his doctoral thesis a t t h e University of Kansas ( M a y , 1948), which has been reproduced by t h e United S t a t e s Atomic Energy Commission a s a n unclassified document, NP-1104, N o v . 30, 1919. (2) Recent work is t o be found i n : ( a ) P. Wehner a n d J C . Hindman, THISJOURNAL, 7 2 , 3911 (1950); (b) P. Wehner a n d J. C. Hindman, J. Phys. Chem., 66, 10 (1952); ( c ) L. W. Niedrach a n d A. D. Tevebaugh, THISJOURNAL, 73, 2835 (1951); (d) J. K. Backhouse and F. P. Dwyer, Proc. Roy. SOC.New Soulh Wales, 83, 138 (1949); 83, 146 (1949).
It was desirable to have as general a method as possible for the identification of ruthenium species. One has been developed which makes use of ionexchange measurements. Two experiments are performed. First, the charge on the species per ruthenium atom, a, is found. Second, the total charge per species, b, is measured. The ratio b/a equals the number of ruthenium atoms per species. If the oxidation number of the ruthenium is known, the number of negative charges contributed to the species by anions can be inferred. If only one anion is involved, the complete formula of the species is determined. The method has been used to identify R u + and ~ R U C I + ~ .A chloride complex of the + 3 oxidation state with b equal to +1 was detected and it is presumed to be RuC12+. Determination of Charge per Metal Atom.The method is based on the fixed exchange capacity of the resin, measured in equivalents of charge. The charge per metal atom, a, is equal (3) R. E. Connick and H . H. C a d y , THISJ O U R N A L , 79, 4242 (1957).
DETERMINATION OF FORMULAS OF AQUEOUS Ru(II1) SPECIES
June 5, 1958
to the equivalents of charge displaced from the resin divided by the gram atoms of metal atom taken up by the resin, when ion exchange occurs. For the ruthenium experiments it was convenient to use a column method4 rather than a batch exchange, because, in addition t o the determination of a, the ruthenium species could be separated and concentrated. The procedure was to place the ruthenium a t the top of a cation-exchange column and elute with a cerous perchlorate solution containing some perchloric acid, The cerous ion is held sufficiently strongly to the resin to displace essentially completely hydrogen ion and the ruthenium species of interest. As the ruthenium species are displaced they separate into bands with the least easily held species a t the head. These bands consist of essentially pure ruthenium species which occupy nearly all of the exchange sites within each band. As each band emerges from the column a number of samples are taken for analysis of ruthenium and free hydrogen ion. The value of a is given by the expression a =
(H+)i
+ 3 ( C e 9 i - (H+)t (R4f
where parentheses are used to indicate concentration in moles per liter of solution, i and f designate the elutriant and eluate, respectively, and the ruthenium concentration in the eluate is in gram atoms per liter. I n the above treatment i t is assumed that the unknown cation has the same charge when on the resin as in solution. If changes in complexing occur during the exchange process, the value of a cannot be calculated from equation 1. Changes in hydrolysis on exchange constitute a special case for which i t can be shown that equation 1 holds. By way of example, assume that the following exchange involving the metal cation M + 3 and hydrogen ion occurs Mt3
+ 2H+ + HsO = MOH+* + 3 H +
(2)
where the bar over a formula designates the resin phase. It is seen that the number of hydrogen ions appearing in the solution corresponds t o the charge on M in the solution. It is simple to show that this result is generally true, even when other reference cations in addition to hydrogen ion are involved in the reaction. Any cation which has changed its degree of hydrolysis on entering the resin will occupy one less exchange site per OH- group added. Consequently, i t displaces from the resin one less equivalent of charge per OH- group. Simultaneously i t has produced in the aqueous phase one more hydrogen ion per OH- group added, in the act of hydrolysis. Therefore the equivalents of positive charge released in solution correspond to the case of no change in hydrolysis and equation 1 applies. Equation 1 gives always the charge per atom on the species in the aqueous phase, regardless of hydrolysis. Should there be a change in the complexing (other than hydrolysis) of the unknown cations in the exchange process, equation 1 will not yield the (4) Analogous t o t h e method used b y D. A. Everest and J. E. Salmon for determining t h e charge per metal a t o m in pentagermanate ion. J . Chem. Soc., 1444 (195.5).
2647
correct value of a. I n principle i t would be possible to measure the aqueous concentration of the anion forming the complex, thereby giving the data necessary for a correct calculation of a, but in practice this would be difficult experimentally. I n the case of the ruthenium species in the + 3 oxidation state studied here, it is believed that there is little likelihood that such a change in complexing (other than hydrolysis)6 could occur during an experiment, because i t is known that these complexes form and break up only slowly. When changes in complexing of the unknown cation do occur, one obtains from equation 1 the charge per metal atom of the species in the resin, if the anion in the aqueous phase does not combine with hydrogen ion. This may be seen from the following type reaction where X- is a complexing anion MX+?
+ 3 H f = s3 + 3H+ + X-
(3)
If the acid H X is dissociated in the aqueous phase, as shown in the equation, the equivalents of positive charge displaced by M are equal to the charge on M in the resin phase. Equation 1 then yields a for the species in the resin. When H X is undissociated in the aqueous phase, equation 1 yields a for the species in the aqueous phase because only the free hydrogen ion concentration is to be used in (1). Hydrolysis is a n example of such a case. When the reference cation (e.g., Ce+S)undergoes hydrolysis or other forms of complexing, equation 1 will yield a for the aqueous species if the coefficient multiplying the concentration of the reference ion in equation 1 corresponds to the number of exchange sites occupied by the reference ion in the resin phase, regardless of the complexing in the aqueous phase. I n this case (Ht)i should be the stoichiometric hydrogen ion concentration calculated assuming the degree of hydrolysis found for the reference ion in the resin phase. Changes in degree of polymerization can affect the a values of equation 1 only when changes in complexing (other than hydrolysis) occur simultaneously. Determination of Charge per Species.-The charge per species is determined from the dependence of the equilibrium distribution of the unknown ion on the concentration of the known exchangeable ion in the resin and aqueous phases. Consider for example the reaction
where Q is the equilibrium quotient expressed in concentrations, K is the equilibrium constant, and signifies an activity coefficient. Itis seen that the distribution of M, given by (M*)/(M+b) varies as the b power of the hydrogen ion distribution, (H+)/(H+). Equilibration a t two different hydrogen ion concentrations and substitution ( 5 ) Hydrolysis can occur rapidly since only a proton has t o be moved off the oxygen of a water molecule attached t o t h e ruthenium. Changes i n complexing, o t h e r t h a n hydrolysis, require t h e replacem e n t of a t o m s in t h e first coordination sphere of t h e ruthenium, which is a relatively slow process.
2648
HOWARD H. CADYAND ROBERT E. CONNICK
Vol. 80
into equation 5 permits the simultaneous solution aqueous phase plus twice the anionic charge lost by for the two unknowns b and Q, if their values are the unknown ion in passing from the aqueous to the same for the two experiments. Actually Q the resin phase. Intermediate cases of complexing will vary because of activity coefficient changes, as will give intermediate values of b. shown on the right of equation 5. I n practice this Experimental variation was minimized in two ways. First, the mole fractions of ruthenium and hydrogen ion in Apparatus.-Ml absorption spectra were measured on the resin were kept a t very nearly the same absolute either a Cary Recording Spectrophotometer iModel 11, value in the two experiments so that the activity Serial 4, or a Beckman Model DU Spectrophotometer using quartz cells. A Beckman Model G pH meter was employed coefficients in the resin phase would not change ap- to determine the hydrogen ion concentration in solutions preciably. This was done by keeping the ruthe- where direct titration was not possible. nium mole fraction in the resin very small. There is Solutions.-The water used in all experiments was either an additional effect on activity coefficients in the normal distilled water or conductivity water prepared by redistillation of distilled water from alkaline permanganate resin caused by the change in activity of water in solution. In neither case could evidence of chloride ion be the aqueous phase, but a t the concentrations of obtained by addition of silver nitrate. The perchloric acid perchloric acid employed (
