1 Introduction

The platinum group element, ruthenium is produced in relatively large quantity during the fission of plutonium-rich nuclear fuels in fast breeder reactors. Ruthenium creates problems during the aqueous reprocessing of spent fuels by PUREX process as well as in the management of high level liquid waste (HLLW), owing to the formation of large number of complexes in nitric acid medium and also highly volatile radiotoxic RuO4. The complexity in the chemistry of Ru is because of the large range of its oxidation states from 0 to +8 and also −2. Considerable effort has been made to investigate the chemistry of ruthenium in nitric acid medium [1–6]. Ruthenium complexes formed in nitric acid can be represented by the general formula [RuNO(NO3)x(NO2)y(OH)z(H2O)5−x−y−z]3−x−y−z. Ruthenium can also form nitrosyl (RuNO) complexes in nitric acid, which have the ability to exchange ligands. The composition of the solution containing a mixture of ruthenium nitrosyl complexes depends on the concentration of acid, temperature, ruthenium species present initially and the time elapsed since its formation because of the inter-conversion of different nitrosyl complexes with different half times [7]. In the PUREX process, for the separation of U and Pu from the fission products by solvent extraction, some of the complexes of Ru (mainly trinitrato ruthenium nitrosyl complexes) are extracted into the organic phase, 30 % tributyl phosphate (TBP) along with U and Pu. Reduction of [Ru(NO)]3+ complexes to the lower oxidation state namely [Ru(NO)]2+ would facilitate the separation of Ru from U and Pu during PUREX process. Among the methods reported in the literature [8] for the separation of Ru from radioactive liquid waste, electrochemical method is advantageous, because this method is simple and requires neither the addition of external reagents nor complicates the reprocessing steps and waste management operation. In the electrochemical method, Ru can be separated either by deposition as metal on the cathode [9–14] or by oxidation to RuO4 at the anode [15–20]. Separation of Ru was quantitative in the electro-oxidation method [20]; however, separation percentage of Ru by electro-deposition technique was found to be very low and a large portion remained in the solution. In order to understand the low recovery of Ru by electro-deposition technique, it is desirable to study the behaviour of the lower oxidation states of Ru. The electrolytic reduction of ruthenium nitrosyl complexes had been investigated by several researchers in various electrolytes [21–26]. Nevertheless, there is no reliable report on the reduction of ruthenium nitrosyl complexes in nitric acid medium, except a single study on the oxidation of ruthenium nitrosyl complexes in nitric acid media [27]. In the present study, the reduction behaviour of ruthenium nitrosyl complex [RuII–NO+]3+ and its kinetics in nitric acid medium have been investigated using the potentiostatic electrolysis techniques, cyclic voltammetry (CV) and chronopotentiometry (CP), at Pt and glassy carbon (GC) working electrodes.

2 Experimental

For performing electrochemical measurements, a standard solution of ruthenium nitrosyl [RuNO]3+ containing 40 mM of Ru in 1 M nitric acid was prepared by diluting the commercially available ruthenium nitrosyl nitrate solution (1.7 % Ru in 9 M nitric acid, supplied by M/s. Arora Matthey Ltd, Kolkata). Nitric acid (AR grade, 69–71 %) used in all the experiments was supplied by M/s. Fischer Chemicals Ltd, Chennai. Electrochemical measurements on the reduction of ruthenium nitrosyl complex were conducted using the cell configuration shown in Fig. 1. A 1-mm diameter platinum wire (surface area (SA): 0.42 cm2) and a 3-mm diameter GC rod (SA: 0.44 cm2; supplied by M/s. Metrohm India) were used independently as the working electrodes. A Pt foil (SA: 1.2 cm2) and Ag/AgCl were used as the counter and reference electrodes, respectively. The GC working electrode was finely polished using alumina slurry on a microcloth pad. It was cleaned before and after polishing by ultrasonication in distilled water for about 300 s in order to remove the adhered particles. Further degreasing and rinsing were carried out in acetone and distilled water, respectively and was gently dried with a hot air blower. Ruthenium nitrosyl nitrate solution (25 ml) containing 5, 20 or 40 mM ruthenium in 1 M nitric acid served as the electrolyte. The voltammograms were recorded at different scan rates in the range 10–100 mV s−1. De-oxygenation of the test solutions prior to the scans was done by flushing the cell assembly with argon gas and the results were analysed after IR compensation. Likewise, chronopotentiograms were recorded by applying very small increments of current to the cell. Reduction of ruthenium nitrosyl complex, [RuNO]3+ was carried out by chronoamperometry and the reduced Ru species was characterized by UV–Visible spectrophotometry. Chemito Spectrascan, UV 2600 double beam UV–Vis spectrophotometer was used for recording the spectra. Bulk electrolysis was conducted in a divided electrolytic cell using glass frit as the diaphragm. Pt mesh of surface area, 34 and ~20 cm2 were used as the working and counter electrodes, respectively, along with the reference electrode. The potentiostatic and constant potential electro-oxidation studies were carried out using Comprehensive Autolab (Model PGSTAT-30; M/s. Eco-Chemie, the Netherlands) electrochemical system equipped with General Purpose Electrochemical Software.

Fig. 1
figure 1

Cell assembly and electrochemical work station used for electrochemical studies

Full size image

3 Results and discussion

3.1 Reduction potential and cyclic voltammograms of [RuNO]3+ complex in nitric acid medium

The cyclic voltammograms of 40 mM of Ru in [RuNO]3+ solution in 1 M nitric acid, recorded at the scan rate of 50 mV s−1 using platinum working electrode, at 298 K are shown in Fig. 2. The voltammograms were recorded, starting with the cathodic scan over the potential range 1.0 to −0.3 V against Ag/AgCl reference electrode. A distinct reduction peak (I c) observed at about 0.4 V and the oxidation peak (I a) at about 0.6 V (vs. Ag/AgCl) corresponded to the redox couple [RuNO]3+/[RuNO]2+. In addition to this peak, there was a surge in the cathodic current at 0.04 V, culminating into a small hump at about −0.16 V during the cathodic scan, which could be attributed to the reduction of [RuNO]2+ to Ru metal. The second oxidation peak observed at 0.95 V, when the CV was recorded at Pt working electrode over the potential region 1.2–0.0 V (vs. Ag/AgCl) is shown in Fig. 3. This oxidation peak could be ascribed to the oxidation of [RuNO]3+ to [RuNO]4+. Cyclic voltammograms of 5 and 20 mM of [RuNO]3+ solutions in 1 M HNO3 were also recorded, with the scan rate of 50 mV s−1. The reduction peak current was observed to increase with increasing concentration of ruthenium (Fig. 3).

Fig. 2
figure 2

Cyclic voltammogram of 40 mM ruthenium nitrosyl ([RuNO]3+) solution in 1 M HNO3; scan rate: 50 mV s−1; working electrode: Pt; temperature: 298 K

Full size image
Fig. 3
figure 3

Cyclic voltammograms of [RuNO]3+ solution ([Ru]: 5, 20 and 40 mM) in 1 M HNO3, recorded using 50 mV s−1 scan rate at Pt working electrode and at 298 K

Full size image

Ruthenium nitrosyl complexes with the general formula [RuNO(NO3)x(NO2)y(OH)z(H2O)5−x−y−z]3−x−y−z may be cationic, anionic or neutral complexes depending on the values of x, y and z. The NO ligand in the [RuNO]3+ complex is considered to bond formally as NO+ by donating an electron to Ru(III). The diamagnetism exhibited by [RuNO]3+ complexes ascertain this; hence, the complex is equivalent to a bivalent Ru(II) containing NO+ and the most probable description for the moiety would be [RuIINO+]3+ [28, 29]. The bonding electrons are partially donated from the filled metal d-orbitals to the empty ligand NO \(\pi^{ * }\) orbital (this process is known as metal-to-ligand π-bonding or back-bonding). Six coordinated ruthenium complexes are extremely stable because Ru2+ with the d6 configuration obeys the 18-electron rule, when octahedrally coordinated and additionally stabilized by π-bonding ligands. Based on molecular orbital (MO) analysis using the density functional theory (DFT) and from the reduction product analysis, it has been reported that during the electrochemical reduction of ruthenium nitrosyl complexes, the first site of reduction would be assigned to the NO+ to NO at the coordinated NO+ (Eq. 1) [30]. Further reduction might occur at the ruthenium centre and the possible reduction is Ru2+ to Ru metal (i.e. Ru2+ + 2e− = Ru0). It is also reported that the reduction potential for NO+/NO in Ru nitrosyl complexes is sensitive to the nature of trans ligand. As the π acidity of ligand increases, more nitrosonium character is imposed in the NO ligand and, thus, its reduction becomes easier [21–24, 30]. In the present work, the reduction wave (I c) and the corresponding oxidation wave (I a) were assigned to the reduction of [RuIINO+]3+ to [RuIINO]2+.

$$\left[ {{\text{Ru}}^{\text{II}} {\text{NO}}^{ + } } \right]^{ 3+ } +\, {\text{e}}^{ - } = \left[ {{\text{Ru}}^{\text{II}} {\text{NO}}} \right]^{ 2+ }.$$
(1)

Gomes et al. [24] reported that the NO+/NO reduction process in some ruthenium nitrosyl complexes is reversible within the time scale of experiments, suggesting the reasonable stability of some of the reduced complexes; however, Doro et al. [31] observed that this reduction process for some complexes is irreversible and it depends on the experimental conditions such as scan rate and temperature.

The peak corresponding to the reduction of ruthenium nitrosyl complex was validated by carrying out a controlled potential electrolysis of a solution containing 160 ppm of Ru in 1 M nitric acid at 0.4 V. The UV–Visible spectra of the electrolyte solutions recorded before and after electrolysis are reproduced in Fig. 4. Tfouni et al. [30] observed the spectra of ruthenium nitrosyl complexes to display a broad absorption band in the range 400–500 nm and they have attributed it to metal-to-ligand charge transfer (MLCT), RuII → NO transition; another higher absorption peak with high intensity in the range 300–350 nm, recorded by them was ascribed to d–d transition. In the present study also, similar observations were recorded. With increase in the duration of electrolysis, intensity of the characteristic absorption band owing to MLCT transition at 473 nm was found to decrease, which might be due to the destabilization of metal centre caused by the decrease in back-bonding because of NO+ reduction [32]. Increase in the intensity of the absorption band at about 340 nm and generation of new shoulders at 387, 370 and 358 nm were also observed as electrolysis proceeded.

Fig. 4
figure 4

UV–Visible spectra of [RuNO]3+ solution in 1 M HNO3 before and after electrolysis

Full size image

3.2 Estimation of kinetic parameters for the reduction of ruthenium nitrosyl complexes

If the rate of an electrochemical process is controlled by diffusion (mass transfer) and not by charge transfer kinetics, then the process is called as reversible process. In a reversible process, the charge transfer is rapid and the slow mass transfer is the rate determining step. The key criterion for a reversible charge transfer process is that the peak potential, E p is independent of scan rate (υ) and

$$E_{{{\text{p}},{\text{c}}}} - \, E_{{{\text{p}},{\text{a}}}} = 1. 8 5 7 {\text{ RT}}/\alpha nF,$$
(2)

where E p,c and E p,a are the cathodic and anodic peak potentials, respectively, R is the gas constant, T is the absolute temperature (K), α is charge transfer coefficient, n is the number of electrons involved in the charge transfer reaction and F is the Faraday constant.

The reversibility of the electrode processes corresponding to the peaks I c and I a in Fig. 2 was examined by evaluating the effect of scan rates on the current–potential behaviour, in the electro-reduction of ruthenium nitrosyl ions. Figure 5 represents all the scans from 10 to 100 mV s−1 for the reduction of 40 mM of [RuNO]3+ solution in 1 M nitric acid, within the selective potential window of 0.15–0.85 V at 298 K. This figure reveals that the cathodic peak potential, E p,c shifted significantly from 0.46 to 0.39 V (vs. Ag/AgCl) as the scan rate was increased from 10 to 100 mV s−1, which is not the characteristic of a reversible reaction. The peak parameters measured from the cyclic voltammograms in Fig. 5 are tabulated in Table 1.

Fig. 5
figure 5

Cyclic voltammograms of 40 mM of ruthenium nitrosyl solution in 1 M HNO3 recorded at Pt electrode with different scan rates (i.e. 10–100 mV s−1); T: 298 K

Full size image
Table 1 Peak parameters obtained from the CVs for 40 mM [RuNO]3+ solution in 1 M HNO3 recorded with Pt working electrode
Full size table

If the electrochemical process is controlled mainly by charge transfer kinetics, then the process is called as irreversible process. The important criterion for the irreversible charge transfer kinetics is the shift in the peak potential with scan rate. A relation between the cathodic diffusion peak current, i p,c and the scan rate, given by Bard and Faulkner [33] is

$$i_{{\text{p, c}}} = \text{0}\text{.496 }nFAC_{\text{0}} \sqrt {\frac{{D_{\text{0}} \text{(}\alpha \text{ }n_{\upalpha} \text{)}F\upsilon \text{ }}}{{\text{RT}}}},$$
(3)

where A is area of the electrode (in cm2), C 0 is the bulk concentration of electro-active species (mol cm−3), υ is the scan rate (mV s−1), D 0 is the diffusion coefficient (cm2 s−1) and n α is number of electrons transferred in rate determining step.

The value of diffusion coefficient of [RuNO]3+ ion calculated using Eq. (3) and from the slope of the plot of i p,c against square root of scan rate (Fig. 6) was 2.95 × 10−8 cm2 s−1. This very low diffusion coefficient value indicates the complexing nature of Ru(II). The existence of Ru(II) in different complexes instead of bare Ru(II) ions, limits the recovery of ruthenium from aqueous solutions. Further, Ru(II) undergoes two step reduction to Ru and the second reduction occurs in the decomposition range of the aqueous solution, which controls the separation of Ru by electro-deposition.

Fig. 6
figure 6

Plot of peak current against square root of scan rate for the reduction of 40 mM of Ru in [RuNO]3+ solution; [HNO3]: 1 M

Full size image

For the reduction of [RuNO]3+ ions in 1 M nitric acid medium, peak potential shift and broadening of peak shape with scan rate were observed in cyclic voltammetric runs. The difference between the cathodic and anodic peak potentials (ΔE p) increased with scan rate and the average of the peak potentials, 1/2(E p,c + E p,a) was more or less constant for different scan rates, indicating that the reduction process is quasi-reversible [34, 35].

The heterogeneous electron transfer rate constant, k s could be calculated using Eq. (4), which was proposed by Klingler and Kochi [36] based on peak separation.

$$k_{\text{s}} = \text{2}\text{.18 }\left[ {D_{\text{0}} \text{(}\alpha \text{ }n_{\upalpha} \text{)}\frac{\upsilon F}{{\text{RT}}}} \right]^{{{\raise0.7ex\hbox{$\text{1}$} \!\mathord{\left/ {\vphantom {\text{1} \text{2}}}\right.\kern-0pt} \!\lower0.7ex\hbox{$\text{2}$}}}} \text{exp}\left[ {\frac{{\alpha^{\text{2}} \text{ }nF}}{{\text{RT}}}\text{(}E_{\text{p}}^{\text{c}} - E_{\text{p}}^{\text{a}} \text{)}} \right].$$
(4)

According to Eq. (4), if [k s/(υ)1/2] > 0.11, the process is reversible. If 0.11 > [k s/(υ)1/2] > 3.7 × 10−6, then it is quasi-reversible. If 3.7 × 10−6 > [k s/(υ)1/2], the process is irreversible. The value of ks calculated for the scan rate of 10 mV s−1, using D 0 = 2.95 × 10−8 cm2 s−1 was 2.06 × 10−5 cm s−1 and the value of [k s/(ν)1/2] obtained for the same scan rate was 2.06 × 10−4 cm s(−1/2) V(−1/2). These values imply that the reduction of [RuNO]3+ ion is quasi-reversible.

3.3 Chronopotentiometric study on the reduction of ruthenium nitrosyl ions

The chronopotentiograms of 40 mM ruthenium in [RuNO]3+ solution in 1 M HNO3, recorded at Pt electrode for various applied constant current are shown in Fig. 7. The transient time (τ) in the chronopotentiogram is a measure of time elapsed between the attainment of constant potential and the time at which the concentration of electro-active species (resulting from diffusion) reaches zero at the electrode. A relation between the applied current and transient time is given by Sand’s equation [33], which enables the determination of diffusion coefficient.

$$i\tau^{{{\raise0.7ex\hbox{$\text{1}$} \!\mathord{\left/ {\vphantom {\text{1} \text{2}}}\right.\kern-0pt} \!\lower0.7ex\hbox{$\text{2}$}}}} = \frac{{nFA\text{ (}D_{\text{0}} \pi \text{)}^{{{\raise0.7ex\hbox{$\text{1}$} \!\mathord{\left/ {\vphantom {\text{1} \text{2}}}\right.\kern-0pt} \!\lower0.7ex\hbox{$\text{2}$}}}} }}{\text{2}}C_{\text{0}}.$$
(5)

The D 0 value of 1.1 × 10−8 cm2 s−1 derived using Eq. (5) at room temperature is in reasonably good agreement with the value of diffusion coefficient calculated from cyclic voltammetric results.

Fig. 7
figure 7

Chronopotentiograms recorded for [RuNO]3+ ions in 1 M HNO3 at different applied currents; [Ru]: 40 mM; WE: Pt; CE: Pt; RE: Ag/AgCl

Full size image

3.4 Reduction of [RuNO]3+ moiety using glassy carbon (GC) working electrode

The cyclic voltammograms recorded for a solution of [RuNO]3+ containing 40 mM Ru in 1 M HNO3 at different scan rates, using GC as the working electrode at 298 K are shown in Fig. 8. The redox behaviour exhibited by [RuNO]3+ in 1 M HNO3 at GC electrode was similar to that at Pt electrode. However, the reduction wave for [RuNO]3+ complex was observed to occur at lower cathodic potentials (0.9 V vs. Ag/AgCl) with GC working electrode than that of 0.45 V versus Ag/AgCl with Pt WE, for the scan rate of 10 mV s−1. The peak parameters measured from the cyclic voltammograms using GC as the working electrode are listed in Table 2.

Fig. 8
figure 8

Cyclic voltammograms of 40 mM [RuNO]3+ solution in 1 M nitric acid recorded using GC electrode with different scan rates at 298 K; CE: Pt; RE: Ag/AgCl

Full size image
Table 2 Peak parameters obtained for 40 mM [RuNO]3+ solution in 1 M HNO3 from the CVs recorded with GC working electrode
Full size table

The peak potential shift and broadening of peak shape with scan rate in CV, observed using GC electrode were similar to the behaviour of the CVs at Pt electrode. The difference between the cathodic and anodic peak potentials (ΔE p), increased with scan rate and the average of the peak potentials, ½ (E p,c + E p,a) was nearly constant for different scan rates, suggesting that the process could be quasi-reversible. The value of diffusion coefficient calculated on the basis of Eq. (3) was 1.65 × 10−8 cm2 s−1 for the reduction of [RuNO]3+ moiety at GC electrode, which is in good agreement with the value of 2.95 × 10−8 cm2 s−1 obtained at Pt electrode.

The potential versus time transients for the reduction of [RuNO]3+ ions at GC electrode is shown in Fig. 9. Using Eq. (5), D 0 of [RuNO]3+ at GC electrode was determined to be 1.57 × 10−8 cm2 s−1, which is in excellent agreement with the value of 1.1 × 10−8 cm2 s−1, obtained at Pt electrode using Eq. (5). The heterogeneous electron transfer rate constant, k s (6.73 × 10−5 cm s−1), obtained at GC electrode also indicates the reduction of [RuNO]3+ in 1 M nitric acid to be quasi-reversible.

Fig. 9
figure 9

Chronopotentiograms of 40 mM [RuNO]3+ solution in 1 M HNO3 recorded at GC electrode with different applied current

Full size image

The value of heterogeneous electron transfer rate constant (k s), estimated using Klingler and Kochi equation [36] and the diffusion coefficient values determined from the results of CV and CP for the reduction of [RuNO]3+ species in 1 M HNO3 at the working electrodes Pt and GC are compared in Table 3. The data in Table 3 reveal the redox behaviour of [RuNO]3+ solution in 1 M HNO3 to be the same with both electrodes.

Table 3 Comparison of kinetic parameters derived for the reduction of [RuNO]3+ species in 1 M HNO3 at Pt and GC working electrodes
Full size table

4 Conclusions

The reduction behaviour of [RuNO]3+ species in nitric acid medium was investigated by CV and CP using Pt as well as GC working electrodes. Quasi-reversible, one electron reduction of [RuNO]3+ ion was observed at both electrodes. The diffusion coefficient (D 0) of [RuNO]3+ species, determined by CV and CP techniques was in the order of 10−8 cm2 s−1 and the heterogeneous electron transfer rate constant (k s) for the reduction of [RuNO]3+ was estimated to be in the order of 10−5 cm s−1 using Klingler and Kochi equation. The very low diffusion coefficient value revealed the complexing nature of Ru(II). As Ru(II) exists as different complexes instead of bare Ru(II) ions in nitric acid medium, the separation percentage of Ru from ruthenium nitrosyl nitrate solution by electro-deposition technique will be very low. Hence, it is advantageous to separate and recover Ru as RuO4 from liquid waste, by electro-oxidation method.