POTENTIAL O F 'THE
Ru(II)-Ru(III) COUPLE
3103
The Potential of the Ruthenium(I1)-Kuthenium(II1)
Couple]
by R. R. Buckley2 and E. E. Mercer Department of Chemistry, University of South Carolina, Columbia, South Carolina 29.908 (Received March 4, 1966)
The potential of the Ru(I1)-Ru(II1) couple has been determined by direct measurement using a glass electrode as a reference. The potential of the cell was measured at several temperatures and a t several ionic strengths at 25". From these data E" = -0.2487 v, AGO = 5.74 kcal/mole, AH" = --10.1 kcal/mole, and AS" = 53 eu a t 25" for the cell reaction, Ru2+ H+ = Rua+ 1/2H2. Approximate values for the Ru(I1)-Ru(II1) couple involving the chloride complexes are also presented.
+
The chemistry of ruthenium in aqueous solution is very complicated because of the large number of oxidation states, complexes, and polymeric species. A lack of knowledge of the thermodynamics of reactions has resulted from this complexity. With the isolation of the monomeric aquo complex of r~t,henium(II),~ the possibility of measuring the Ru(I1)-Ru(II1) potential was available in a solution where the species of both oxidation states of the metal were well characterized. A number of values for this couple can be found or calculated from data presented in the literature. However, in all of these cases the actual species present in soIution were not known with certainty or the measurements were of very low precision, and involved a number of approximations. Backhouse and Dwyer4 measured the potential of the Ru(I1)-Ru(II1) couple in 1.5 to 6.8 M hydrochloric acid. Yo attempt to maintain constant ionic strength was made, nor were the complexes present known with any certainty. The E" which they calculated was 0.084 v . ~From polarographic studies of the reduction of Ru(1V) made by Atwood6 an approximate value for the Ru(I1)-Ru(II1) couple can be calculated to be 0.06 v in perchloric acid medium. Here also the species present in solution were unknown, but there was a hydrogen ion dependence of the potential which indicated that hydrolytic ions were involved in the electrode reaction. -Endicott and Taube' have recently reported values of the potential for the couple Ru(NHJe2+-F1u(NH3)e3+ as - 0.24 v. This measurement did involve the estimation of rather large junction potentials. Since there is a great deal of uncertaintY in the potential of this COUple, we have under-
+
taken its measurement in p-toluenesulfonic acid solutions. Most other noncomplexing anions have been found to react with the uncomplexed ruthenium(I1). All measurements were made using a glass electrode as a reference to eliminate the need of correcting for liquid junction potentials.
Experimental Section Solutions. The solutions used in the measurements were prepared from p-toluenesulfonic acid which had been recrystallized from anhydrous ether. The ruthenium(I1) solutions were prepared as reported previously. The ruthenium(II1) solutions were made either by air oxidation of the ruthenium(I1) solutions or by electrolytic oxidation of the same solutions a t a controlled pot,ential of 0.10 v us. a saturated calomel electrode. All solutions were made using normal distilled water except in the dilution studies, where it was necessary to use deionized wat.er. All solutions (1) Presented in preliminary form at the SoutheastSouthwest Regional Meeting of the American Chemical Society, Memphis, Tenn., Dec 1965. (2) Presented in partial fulfillment of the requirements of the Ph.9. degree, University of South Carolina, 1965. Recipient of the Dupont Teaching Fellowship for the year 1963-1964. (3) E. E. Mercer and R. R. Buckley, Inorg. Chem., 4, 1692 (1965). (4) J. R.Backhouse and F. P. Dwyer, Proc. Roy. SOC.,N . S . W., 83, 138, 146 (1949). ( 5 ) All potentials reported here follow the conventions given in "Oxidation Potentials," by W. M. Latimer, Prentice-Hall, Inc., Englewood Cliffs, N. J., 1952. All half-reactions are written as oxidations, and couples with reducing agents stronger than hydrogen have a positive sign. ( 6 ) D. K. Atwood, Ph.D. Thesis, Purdue Cniversity, Lafayatte, In& 1960, (7) J. F. Endicott and H. Taube, Inorg. Chem., 4, 437 (1965).
Volume 70,Number 10 October 1966
R. R. BUCKLEY AND E. E. MERCER
3104
were freed from dissolved oxygen by purging them with purified nitrogen which had been saturated with water vapor by bubbling it through a solution of p-toluenesulfonic acid. Ruthenium analyses were performed as reported previously. Apparatus. The controlled-potential coulometer used in all experiments was designed by The operation of the coulometer required the use of three electrodes, a working electrode, a reference electrode, and an isolated electrode. An electrode made from a piece of 0.05-cm gold foil 1 cm square was used as the working electrode. Both the isolated and reference electrodes were connected to the solution through silicic acid bridges, which were made by addition of concentrated sulfuric acid to a saturated solution of sodium silicate, followed by washing with distilled water to remove free electrolyte. A saturated calomel electrode served as the reference, and the isolated electrode was a spiral of platinum wire immersed in 0.01 F p-toluenesulfonic acid. During an electrolysis, current flowed only between the isolated and working electrodes. For the measurement of the cell potential, the gold foil electrode was used as the indicating electrode, and a Beckman GP Type 41252 glass electrode served as a reference. The silver-silver chloride electrodes used to determine the standard potential of the glass electrode were prepared by a modification of the method of Shedlovsky and XacInnes. Spirals of 22-gauge platinum wire sealed into the end of 6-mm glass tubing were silver-plated from a cyanide bath at a current of 2 ma for 15 hr. The electrodes were rinsed with distilled water, then anodiaed in 0.1 F hydrochloric acid for 1 to 2 hr a t a current of 1 ma. Several such electrodes were connected together after preparation and allowed to stand in 0.1 F hydrochloric acid for 1 week. Electrodes prepared in this manner were reproducible to better than 0.2 mv. Cell potentials were measured to 0.1 mv using a Beckman Research Model p H Meter. Solutions were contained in a jacketed 2.5 X 8-cm cylindrical Pyrex container with a Teflon cap. Five openings in the cap supported the gold, isolated, calomel, and glass electrodes, and a gas inlet tube. The temperature of this vessel was controlled by circulation of water thermostated to =tO.Ol" through the outer jacket. Nitrogen was bubbled through the solution continuously during all experiments. It was verified that there was no detectable difference in potential recorded in the presence or absence of the gas purge. All polarograms were recorded on a Sargent Model XV Polarograph. Procedure. The standard potential of the glass The Journal of Physical Chemistry
electrode was determined in either 0.1 or 0.01 m hydrochloric acid by measuring the potential of the cell Ag, AgCl/HCl(m)Iglass
(1)
From the Nernst equation, then EOgl*ss
=
-Eeell
+ E"
2RT AgiAgC1
-
F
1n UHCI (2)
Values for E O A ~ ~ Aand ~ C ~aHcl were taken from the report by Zielen.lo The standard potentials of the silver-silver chloride electrode at other temperatures were calculated from the data presented by Harned and Owen." In each experiment the standard potential of the glass electrode was determined before and after the measurements of the cell potentials. In all cases these two readings agreed within 0.1 mv. The experiments in which the ratio of Ru(II1) to Ru(I1) was varied were conducted in the following manner. Ten milliliters of 0.1 F p-toluenesulfonic acid was placed in the cell, and all of the electrodes and the gas bubbler were inserted. The solution was purged for 30 min with inert gas and was prereduced a t -0.5 v us. sce. Five hundred p1 of the stock ruthenium(I1) solution in p-toluenesulfonic acid was added with a pipet. The solution was again reduced electrolytically at -0.4 v us. sce to the limiting current. The polarity of the electrolysis was then reversed, and measured fractions of the ruthenium were oxidized to the +3 oxidation state at a potential of 0.1 v. The electrolysis was interrupted at intervals of 10% of the total ruthenium present to obtain potential measurements. Stable readings of the cell potential were obtained within 3 to 5 min after discontinuation of the electrolysis. In the dilution experiments, 200 pl of a stock ruthenium(I1) solution was added to 2 ml of deaerated, prereduced 0.1 F p-toluenesulfonic acid. The Ru(II1) to Ru(I1) ratio was adjusted to unity by electrolytic oxidation. The sample was diluted in increments by addition of 2 ml of deaerated, deionized distilled water, or 0.1 F sodium p-toluenesulfonate. The potential of the cell was determined after each dilution.
Results ZielenlO has shown that a glass electrode may be used as a reference electrode in most solutions if a correction (8) R. E. Propst, U. S. AEC Report, DP-798 (1963). (9) T. Shedlovsky and D.A. MacInnes, J . Am. Chem. Soc., 58, 1970 (1936). (10) A. J. Zielen, J. Phys. Chem., 6 7 , 1474 (1963). (11) H. 5. Harned and B. B. Owen, "The Physical Chemistry of Electrolyte Solutions," 2nd ed, Reinhold Publishing Corp., New York, N. Y.,1950.
POTENTIAL OF THE Ru(I1)-Ru(II1) COUPLE
3105
is made for a slow drift of the potential with time. In most cases this drift is very small, and was below the limit of detection in our experiments in p-toluenesulfonic acid. The duration of most of these experiments did not exceed 3 hr. The glass electrode offers a convenient means of eliminating liquid junction potentials in the determination of the emf of cells. Since the glass electrode responds in a manner identical with a hydrogen electrode, the potential of the cell
I
A u I R u ~ + ( ~Ru3 ~ ) +(mz), , H +(m3)glass
(3)
is given by the equation Eeeii
=
--
E O R ~
Eogiass
-
RT
- In [URu ( I I I ) / a R u (1I)ai-I ]
(4) The concentration dependence of the cell potential was tested at constant ionic strength for changes in [Ru(111)], [Ru(II)], and [H+]. Some typical data are presented in Table I. The slopes of the plots of E o e l l Eoglass us. log [Ru3+]/[Ru2+] [H+] a t all temperatures used were in agreement with that given by the Kernst equation within experimental error. The formal potentials obtained at an ionic strength of approximately 0.1 are summarized as a function of temperature in Table I [.
+
Table I : Some Cell Potentials Measured a t fl = 0.103 zk 0.002,
T
= 25.40'
Large activity coefficient effects were expected to occur in this cell since the reaction involved highly charged ions. In an attempt to obtain a value of E" at infinite dilution, the cell potential was determined a t several different ionic strengths, at 25", and a fixed ratio of ruthenium(II1) to ruthenium(I1) of unity. In all of these measurements, however, the concentration of the acid was greater than lod2M to avoid the necessity of corrections for the hydrolysis of the metal cations. The formal potentials at each ionic strength were calculated from the experimental data, and are related to the standard oxidation potential by the expression
E0t = E ' R ~- 0.05916 log { Y R ~ ( I I I ) / Y R ~ ( I I ) Y H } (5) The values of the y's were calculated from the equation log y
0.981 0.736 Ci. 490 0,245
9.56 9.53 9.51 9.48 4.81 3.26 1.98 1.25
3.07 2.08 1.26 0.80
0.2584 0.2818 0.3047 0.3308 0.3129 0.3217 0.3349 0.3489
0.2337 0.2318 0.2338 0,2346 0.2349 0.2337 0,2341 0.2360
Table I1 : The Formal Potential for the Half-Reaction,
+
R u * + = Ru3+ e-, a t Constant Ionic Strength, 0.103 f 0.002
p =
OC
No. of measurements
25.40 25.00 19.40 15.00
15 7 7 7
T,
The limits of the uncertainty are 95y0 confidence level.
1
+ 0.3281a0.\/~
+ H + = Ru3+ +
have been calculated at 25" :
AGO
(7 1 = 5.74 f 0.03 kcal/ '/2H2
-GuE== .249
I
o
h
0
0.2342 i.0.0008n 0.2341i0.0005 0.2464i.0.0005 0.2578 i.0.0007
-0 . 5 0 8 5di ~~
where z is the charge of the ion, u o is the effective diameter of the hydrated ion in angstroms, and p is the ionic strength. The values of uo used for each ion were estimated from similar ions given by Kielland.12 For Ru3+and H+, uo was assumed to be 9, while a value of 5 was used for Ru2+. The standard oxidation potential of the electrode a t each ionic strength is shown in Figure 1, and is independent of ionic strength well within experimental error. The extrapolated value of E O R was ~ -0.2487 v. From the potential and the thermal coefficient of the cell, the following values for the reaction Ru'+
0.245 0,490 0.736 0.981 3.07 2.08 1.26 0.80
=
-02
004
.06
a08
.IO
Figure 1. The standard oxidation potential as a function of ionic strength at 25". E'R"was calculated using eq 5 and 6 in the text.
(12) J. Kielland, J. Am. Chem. Soc., 59, 1675 (1937).
Volume YO, Number 10
October 1966
R. R. BUCKLEY AND E. E. MERCER
3106
mole, A S o -= 53 f 2 eu, A H o = -10.1 f 0.6 kcal/ mole. In addition to the above precise potential determination, we have obtained approximate potentials for the Ru(I1)-Ru(II1) couple for a number of the stable chloride complexes by analysis of polarographic results. The mononieric ruthenium(II1) chloride complexes were prepared and isolated in the manner described by Connick, et ul.13-15 The polarograms were observed as reductions, using a saturated calomel electrode as a reference. The half-wave potentials of the reduction of the di- and trichloro complexes were shown to be independent of the free chloride ion present and, therefore, the electrode reactions were the reverse of those given in Table 111. Because there were no corrections made for junction potentials, ionic strength, or diffusion coefficients, we have chosen rather large limits of error on the potentials given in Table 111. ~~~~
~
~~
Table I11 : Potentials Derived from Polsrograms EO, va
Half-reaction
R u ~ +== Ru3+
+ e-
+
-0.21 0.01 0.10 0.33
+ e+ e-
RuClz = RuClz+ R u c k = RuCh Ru(II)(satd LiC1) R u C ~ & ~ -e-
=
The estimated uncertainty for the potentials reported is 3~0.03v (see text).
Discussion From the polarographic data, some interesting conclusions can be made concerning the species present in solutions of ruthenium(I1) in the presence of high chloride concentrations. From the data of Fine,I6 an approximate value for the equilibrium constant for the reaction Ru3+
+ 6C1-
RuC1e3-
(8)
is found to be of the order of This value was found using the methods given by Rossotti and Ross0tti.l’ In solutions where the concentration of chloride is greater than 10 M , it has been shown that Ru(111) is present as RuC&,~-only.16 Using our poten-
The Journal of Physical Chemistry
tials for the Ru(I1)-Ru(II1) couple in both noncomplexing and saturated lithium chloride solutions, together with the equilibrium constant for reaction 8, K,, = for the equilibrium Ru’+
+ 6C1-
=
RuCle4-
(9) This calculation was based on the assumption that the ruthenium(II1) in saturated lithium chloride solution was reduced to the hexachlororuthenate(I1) ion. While the equilibrium constant calculated for reaction 9 is not highly accurate, the extremely small value strongly suggests that the hexachlororuthenate(I1) ion cannot have stable existence in aqueous solutions of even very high chloride concentrations. The reliability of the extrapolated potential obtained in Figure 1 is deserving of some comment. As can be seen from the plot, quite a lengthy extrapolation was required. In spite of this, we feel that this potential is valid to h0.8 mv because of the excellent agreement between the experimental and the calculated values of the potential at higher ionic strengths. At the present time, unfortunately, there is no way of accurately relating the aqueous ions to the free metal, so precise values of the enthalpies and free energies of formation of these ions cannot be calculated. However, we did not observe any disproportionation of the RuZ+ion in any of our experiments. In chloride solution, ruthenium(I1) has been observed to disproportionate by many investigators. Since the disproportionation of ruthenium(I1) was one of the main factors used by Latimers in estimating the potential of the Ru-Ru2+ couple as -0.45 v, it seems likely that this value is too negative by at least 0.2 v.
Acknowledgments. The authors wish to thank the National Science Foundation for its financial support of this work on Grant No. GP-1584. We also thank Professor J. W. Cobble of Purdue University for his advice and encouragement while this work was in progress. (13) H. H. Cady and R. E. Connick, J. Am. Chem., Soc., 80, 2646 (1958). (14) D. A. Fine and R. E. Connick, ibid., 8 2 , 4187 (1960). (15) D. A. Fine and R. E. Connick, ibid., 83, 3414 (1961). (16) D. A. Fine, Ph.D. Thesis, University of California, Berkeley, Calif.; UCRL-9059 (1960). (17) F. J. C. Rossotti and H. Rossotti, “The Determination of Stability Constants,” McGraw-Hill Book Co., Inc., New York, N. y., 1961, p 110.
