Extraction of Palladium from Spent Nuclear Fuel Reprocessing Solutions

Abstract

New solvent systems for selective separation of palladium from nuclear wastes represent a prospective way to reduce the total waste volume and induce this metal’s extraction. For this purpose, the potential of modern green solvent room-temperature ionic liquid was assessed with diamide-type extractants based on N-heterocycles and S-donating thiodiglicolic acid. The N-donating heterocyclic extractants demonstrate structure-dependent high selectivity toward palladium in the presence of various impurity metals (such as Zr, Cs, Sr, Mo, Ce, Fe, and Cr) from spent nuclear fuel. Palladium is extracted into the organic phase quite selectively with a separation factor greater than a thousand for all extractants. Ionic liquid media are capable of selective palladium separation from platinum group metals and synergetically increase the selectivity of the extractants.

1. Introduction

The global demand for platinum and palladium is constantly increasing, and their production has almost doubled in recent years. Interest in palladium is growing along with the increasing technical need for it, and it is now one of the fastest growing financial assets in terms of value. Limited natural reserves and low content of platinum group metals in ores necessitate the involvement of various types of richer secondary raw materials in the processing. One of the sources of precious platinum group metals, which is of great commercial importance, is spent nuclear fuel (SNF); 1 ton of spent fuel accounts for 4–5 kg of palladium [1,2,3,4]. The extraction of palladium from spent nuclear fuel is currently the most promising option for developing this type of platinum metal raw material. Unfortunately, the composition of palladium released from spent fuel includes the beta-radioactive 107th isotope with a half-life of 6.5 × 10⁶ years, which is contained in quantities ranging from 14 to 16% [5]. When ¹⁰⁷Pd is removed from palladium contained in spent fuel, all remaining metal can be used in the same way as that extracted from natural raw material sources. Due to the lack of volatile compounds in palladium, the only technically feasible method for separating its isotopes today is electromagnetic separation [5]. At the same time, weakly radioactive ¹⁰⁷Pd can also be used as the basis of a catalyst in radiochemical production in the nuclear industry. Unlike natural ores, spent nuclear fuel is a renewable resource, the processing of which produces additional resources of technically important metals. Indeed, using rational spent fuel reprocessing technology, it would be possible to isolate the above valuable elements, but such technologies currently do not exist. However, a number of attempts for their creation have been undertaken [6,7,8,9,10,11,12]. The most common technological method in the SNF reprocessing scheme is extraction, which allows, in particular, to isolate noble metal ions from process solutions and concentrate them in a certain form [1]. Extractants for this purpose must be highly selective as well as resistant to hydrolysis and radiolysis under harsh technological conditions in order to be recyclable and reusable. However, the chromatographic separation, precipitation and chemical and electrochemical reduction of palladium, rhodium and ruthenium were also under investigation. Palladium can be sorbet on various ion-exchange resins directly from nitric acid media [13,14]. Precipitation can be considered as a method for palladium separation from high-level liquid nuclear wastes (HLW) after its chemical or electrochemical oxidation [15]. The electrochemical deposition also combines with exchange chromatography extraction [9].

Palladium forms anionic complexes in solutions of various acids. Being in anionic form, it forms stronger complexes with ligands with soft-donor N and S atoms. This behavior is characteristic of the entire series of noble metals. A number of extractants have been developed on this assumption. Previously, sulfur-containing organic extractants (sulfides) [12], nitrogen-containing extractants (alkylamines [7,8], oximes [10]) and hybride sulfur-oxygen-containing amides [6] were employed for palladium separation from HLW and SNF. Along with sulfur or nitrogen-containing mono- and bidentate extractants, diamides derivatives (malonamides [16], diglycolamides [17], dioxodiamides [18]) were used to extract platinum group metals from nitric acid solutions, simulators of spent fuel.

Therefore, solving the problem of reprocessing spent nuclear fuel and handling technogenic radioactive waste in order to isolate precious and other rare metals (such as palladium, ruthenium, rhodium and silver) from them is an urgent and promising scientific task.

Here, we report solvent extraction of PGM from model PUREX rafinate solution by hybrid “hard-and-soft” N,O-donating diamides in various solvents, including green ionic liquid solvent [BuMeIm]Tf₂N (IL). As a solvent, IL is actively proposed in hydrometallurgy [19,20] and PGM extraction [21,22]. As an extragent, we use diamides of 2,2′-bipyridyl-6,6′-dicarboxamides I (Figure 1). We also test the well-known tetraoctylthiodiglycolamide II, possessing high affinity for palladium and proposed for the separation of PGM from PUREX rafinates [2,6,18] for comparison with diamides I.

2. Materials and Methods

2.1. General Remarks

Caution! All manipulations with silver salts were carried out with protection from light.

The NMR spectra were measured with BRUKER AVANCE-600 MHz and AVANCE-400 MHz NMR spectrometers (Bruker Corporation, Billerica, MA, USA) at 24 °C. The IR spectra were recorded with a Varian 640t FTIR spectrometer (Varian Medical Systems, Palo Alto, CA, USA) with samples in KBr pellets. Metal concentrations were determined using the ICP spectrometers VISTA PRO Varian Optical Spectroscopy Instr. (Varian Inc., Palo Alto, CA, USA) and 725 ICP-OES Agilent (Agilent Technologies, Inc., Santa Clara, CA, USA) and the AAS spectrometer Spectr AA-240 FS Varian Optical Spectroscopy Instr. (Varian Inc., Palo Alto, CA, USA).

The diamides I solution was prepared according to the methods represented earlier [23]. Model solutions of PUREX process raffinates (composition represented in

Table 1. Composition of PUEREX raffinate solution (mg/L) in 1.22 M HNO₃.

Metal	Pd a	Ru a	Rh a	Ag	Cs	Ce	Cr	Fe	Mo	Sr	Zr
Concentration	12.21	4.82	7.50	3.50	13.25	34.0	39.50	0.26	88.50	4.10	97.5

^a—g/L.

) were prepared by dissolving the corresponding salts. The zirconium solution was prepared by boiling zirconyl nitrate in 9 M HNO₃ for 3 days. Solutions were prepared using the volume–weight method. The concentration of nitric acid solution was determined using potentiometric titration. The error in determining element ions in solution for atomic emission spectrometry is 1–7%, depending on the elements being determined, and in the case of atomic absorption spectrometry, the error is less than 2%, also depending on the elements being determined.

2.2. Synthesis of Amide II

2.2.1. Synthesis of Thiodiglycolic Acid Dichloride

A thioglycollic acid (15 g, 0.1 mol) and thionyl chloride (150 mL, 2.1 mol) solution was stirred for 20 h at room temperature. Unreacted SOCl2 was removed using vacuum distillation, leaving a dark oil. Yield: 13.09 g (0.07 mol), 70%. IR (KBr): 2925, 2856 (δ CH2), 1803 (δ COCl).

2.2.2. Synthesis of N,N,N′,N′-Tetraoctylthiodiglycolamide

In a 100 mL round-bottomed flask containing dioctylamine (10 g, 0.0448 mol) and triethylamine (8.686 g, 0,086 mol), 30 mL of DCM was added, followed by mixing. Thiodiglycolic acid dichloride (4.03 g, 0.0244 mol) was added dropwise and left for 12 h. The synthesis was divided using a separatory funnel with HCL and NaCl, then dried over Na₂SO₄ and divided on a column with DCM. Yield: 9.9469 g (0.0167 mol), 77% as a brownish slurry. 1H NMR (600 MHz, CHLOROFORM-d) δ ppm 0.85 (t, J = 6.65 Hz, 12 H, CH₃) 1.07–1.38 (m, 40 H, CH₂) 1.43–1.63 (m, 8 H, β-CH₂) 3.23 (t, J = 7.89 Hz, 4 H, trans-CH₂) 3.27 (t, J = 7.52 Hz, 4 H, cis-CH₂) 3.45 (s, 4 H, CH₂). IR (KBr): 2925, 2856 (δ CH₂), 1645 (δ CO), 1463 (δ CH₃). 13C NMR (600 MHz, CHLOROFORM-d): 13.59 (CH₃), 22.14 (CH₂), 26.41, 26.51, 28.72, 28.77, 28.83, 28.89 (Oct), 31.32 (CH₂), 31.32 (CH₂), 45.89 (trans-CH₂), 48 (sic-CH₂), 168 (CO).

2.3. Synthesis of Silver Complex with Amide Ic

Diamide Ic (100 mg, 0.19 mmol) was dissolved by heating in 3 mL of dry acetonitrile. A measure of 100 mg (0.59 mol) of AgNO₃ was added to the resulting solution. The mixture was boiled for 15 min and left at room temperature for a day. The resulting precipitate was filtered, washed on a filter with absolute acetonitrile and dried in air. A measure of 112 mg (0.165 mol, 87%) of white crystals were obtained. Mass spectrum (MALDI-TOF) m/z 614 [M-NO₃]⁺. 1H NMR (600 MHz, ACETONITRILE-d₃) δ ppm 1.01 (t, J = 6.79 Hz, 3 H) 1.09 (t, J = 6.65 Hz, 3 H) 2.53 (q, J = 7.52 Hz, 2 H) 3.87 (q, J = 6.69 Hz, 2 H) 7.04–7.09 (m, 2 H) 7.09–7.13 (m, 2 H) 7.26 (d, J = 4.95 Hz, 1 H) 7.69 (t, J = 6.69 Hz, 1 H) 7.88 (d, J = 4.95 Hz, 1 H).

2.4. Measurement of Extraction Coefficients

All extraction experiments were carried out in glass tubes with ground stoppers with a ratio of phase volumes of 1:1, at a temperature of 20 °C.

The distribution coefficients during extraction (D = [M]_org/[M]_aq) and the degree of extraction (E = 100% × [M]_org/M_total) were determined at constant extractant concentrations (0.01 mol/L in solvent) and initial metal concentrations in the experiment.

At least three independent experiments were performed for each concentration. The relative value of the confidence limits of the total error at n = 3, and the confidence probability p = 0.95 is δ = 0.33, taking into account the error in the analysis of the chemical composition in solutions, as well as the assessment of the random error and the non-excluded systematic component of the error during the experiment and sample preparation for analysis. Accordingly, the confidence interval of the presented results is 0.02 mg/L.

3. Results and Discussion

3.1. Extraction of Noble Metals from Industrial Waste Solution

Initially, we tested two individual solvents as diluents for the diamide ligands. The effect of nitrobenzene and chloroform on the extraction ability of ligand Ia with respect to noble metals was studied (

Table 2. Degrees of extraction of elements during extraction with 0.01 M solution of Ia in various diluents.

Diluents	Degrees of Extraction, %
	Pd	Ru	Rh	Ag	Cs	Ce	Cr	Fe	Mo	Sr	Zr
PhNO2	6.8	9.0	0	8.6	3.8	0	0	0	0	0	0
CHCl3	2.4	2.7	0	21.4	0	0	0	0	0	0	0
30 vol.% TBP in isopar M	26.6	17.3	0	0	7.5	8.8	1.3	7.6	12.4	7.3	30.8
0.04 M TOMAH in isopar M	5.2	4.8	0	0	7.5	0	0	2.5	4	0	2.1
0.04 M [BMIM]Tf2N in PhNO2	37.2	7.5	0	12.9	0	0	0	0	0	0	0

). The nitrobenzene is a structural analog of meta-nitrobenzotrifluoride (F-3), a popular industrial solvent [24,25]. As can be seen, the use of a 0.01 mol/L solution of ligand Ia in the above solvents as an extractant is quite selective for platinum-group metals (PGM). In the case of using chloroform, no impurity elements are extracted into the organic phase, except for silver. Similar behavior is observed in the case of using nitrobenzene, with the exception of cesium, which is extracted into the organic phase, apparently due to the influence of the crown-like fragment of ligand Ia, the coordination property of which is enhanced in nitrobenzene.

Simple solvent tests show the high potential of the diamide ligand Ia for PGM extraction. For comparison with the previous experiments, we tested the effect of popular radiochemical extraction mixtures, namely, 30 vol.% TBP in isopar M and 0.04 M trioctylmethylammonium nitrate (TOMAN) in isopar M (Table 2).

The use of 30 vol.% TBP in isopar M as a diluent increases the extraction ability of ligand Ia towards PGM. This is perhaps a consequence of the synergistic interaction of ligand Ia and TBP. However, in this case, impurity elements are also extracted, and the extraction of zirconium is comparable with the quantitative characteristics of PGM. So, further use of 30 vol.% TBP in isopar M required the initial separation of zirconium prior to PGM extraction.

The effect of 0.04 M TOMAN in isopar M on the extraction of PGM with ligand Ia. The extraction ability of 0.01 M Ia and 0.04 M TOMAN in isopar M towards PGM is inferior compared to the previous system. In addition, the non-selective extraction of PGM into the organic phase occurs, possibly due to the synergistic interaction of ligand Ia and TOMAN. Moreover, the cationic ammonium salt suppresses the extraction of most impurity elements, which goes into organic phase in the form of complex cation due to the increase in organic phase positive charge. So, the cationic liquids, such as the butylimidazolium salts—the popular green chemistry solvent – perhaps increase the selectivity of PGM isolation. To test this hypothesis, we investigated the effect of ionic liquid 0.04 M [BuMeIm]Tf₂N ([BMIM]Tf₂N) in nitrobenzene on PGM separation (Table 2). It should be noted that [BMIM]Tf₂N is insoluble in isopar M and is salted out into the second phase. Our hypothesis was brilliantly confirmed. Contrary to the nitrobenzene solvent, a solution of 0.04 M [BMIM]Tf₂N in nitrobenzene significant increasing in PGM extraction was observed. In this case, PGMs are extracted selectively into the organic phase, with the exception of silver, and the degree of extraction is much higher than when using TOMAN in isopar M, and is comparable to −30 vol.% TBP in isopar M.

Testing of the [BMIM]Tf₂N in different diluents and ionic liquid itself shows a weak affinity for metal cations, and only small quantities of PGM passes into the organic phase due to presence of PGM complex anions (

Table 3. Degrees of extraction of elements during extraction with 0.2 M solutions of [BMIM]Tf₂N in nitrobenzene and chloroform, and ionic liquid itself.

Solvents	Degrees of Extraction, %
	Pd	Ru	Rh	Ag	Cs	Ce	Cr	Fe	Mo	Sr	Zr
[BMIM]Tf2N in PhNO2	0	7.8	0	0	3.8	1.5	2.5	2.9	1.1	2.4	2.6
[BMIM]Tf2N in CHCl3	0	3.8	0	0	0	0	0	0	0	0	0
[BMIM]Tf2N	0.04	11.2	0	0	0	0	0	0	0	0	0

).

Thus, it was decided to carry out further studies on the extraction of PGM from a 1.22 mol/L nitric acid solution using [BMIM]Tf₂N in a mixture with nitrobenzene or chloroform as a solvent for the diamide ligands.

Almost complete extraction of Pd and Ag was observed for both extraction mixtures in nitrobenzene and chloroform (

Table 4. Degrees of extraction of elements during extraction with 0.2 M solution of Ia in 0.2 M [BMIM]Tf₂N in different diluents.

Diluents	Degrees of Extraction, %
	Pd	Ru	Rh	Ag	Cs	Ce	Cr	Fe	Mo	Sr	Zr
PhNO2	99.9	2.5	0	99.9	7.5	4.4	0	3.6	6.8	2.4	6.2
CHCl3	99.9	11.2	0	99.9	1.9	13.2	0	1.3	5.6	0	5.1

). It is clear that the high selectivity of both extraction mixtures occurs due to synergetic multiplication of both factors: the high affinity of diamide ligand Ia to Pd and Ag and suppression of complex cation distribution into organic phase by ionic liquid. The suppression of the Ru(III) and Cr(III) cation distribution was observed for the nitrobenzene-based system but not for Ce(III). Calculated distribution coefficients for both diluents represented in Figure 2.

The separation factors of palladium were calculated during extraction with a 0.2 M solution of Ia in 0.2 M [BMIM]Tf₂N in nitrobenzene and chloroform (

Table 5. Separation factors of palladium and other elements during extraction with 0.2 M solution of Ia in 0.2 M [BMIM]Tf₂N in different diluents.

Diluent	SFPd/M
	Pd/Ru	Pd/Rh	Pd/Ag	Pd/Cs	Pd/Ce	Pd/Cr	Pd/Fe	Pd/Mo	Pd/Sr	Pd/Zr
PhNO2	4.6 × 104	-	17	1.4 × 104	2.5 × 104	-	3.1 × 104	1.6 × 104	4.7 × 104	1.7 × 104
CHCl3	7.7 × 103	-	14	5.1 × 104	6.4 × 103	-	7.2 × 104	1.6 × 104	-	1.8 × 104

). When using diamide Ia in a mixture with [BMIM]Tf₂N in chloroform or nitrobenzene as an extractant, quantitative (>0.12 M for palladium—loading capacity) selective extraction of palladium with a separation factor over 10⁴ is possible.

The potential loading capacity of the extraction system was 0.2 M Ia in 0.2 M [BMIM]Tf₂N in nitrobenzene or chloroform. As a result of the experiment, it was revealed that the palladium contained in the solution (12.12 g/L or 0.11 M) passes into the organic phase by 99.97%. Accordingly, we assume that the capacity of 0.2 mol of organic ligand (assuming that the ligand is palladium saturated) contains 0.11 mol of palladium ions. And for 1 mol of ligand Ia, there is 0.55 mol of palladium ions, which is 58.3 g.

The next step of the work was to find the optimal conditions for efficient re-extraction of PGMs. As a rule, all soft organic ligands, in which nitrogen or sulfur atoms are donor centers during complex formation with metals, are efficiently extracted in low-acidity solutions. With an increase in the acidity of the solution, the extraction ability of these ligands sharply decreases. We tested 5 M of HNO₃ solution for back-extraction of PGM.

Table 6. Elemental composition of the initial extract in 0.2 M Ia and 0.2 M [BMIM]Tf₂N solution in nitrobenzene.

Pd, g/L	Ru, g/L	Rh, mg/L	Ag, mg/L	Cs, mg/L	Ce, mg/L	Cr, mg/L	Fe, mg/L	Mo, mg/L	Sr, mg/L	Zr, mg/L
12.2	0.12	>DL	3.5	>DL	1.5	>DL	9.5	6.0	0.1	6.0

>DL—below detection limits.

shows the composition of the extract in 0.2 M Ia in a 0.2 M solution of [BMIM]Tf₂N and nitrobenzene, directed to back-extraction with 5 mol/L HNO₃ solution.

The results of back-extraction are shown in

Table 7. Elemental composition of the aqueous solution obtained by re-extraction from extract of composition (Table 6) and the degree of extraction of elements during re-extraction.

The composition of the aqueous solution obtained via back-extraction
g/L		mg/L
Pd	Ru	Rh	Ag	Cs	Ce	Cr	Fe	Mo	Sr	Zr
11.95	0.11	0	2.5	0	0.25	0	3.25	4.0	0.1	3.35
Degree of extraction during stripping, %
Pd	Ru	Rh	Ag	Cs	Ce	Cr	Fe	Mo	Sr	Zr
97.9	93.7	-	71.4	-	16.7	-	34.2	66.7	100	55.8

.

As expected, a 5 M nitric acid efficiently re-extracts the target PGMs from the loaded organic phase.

From the presented data, the coefficient of purification (Kd) of palladium from impurities during its extraction from the model solution was calculated. The purification factor was calculated as the specific ratio of impurity to palladium after re-extraction and before extraction.

Calculation of Kd was carried out according to Formula (1):

$$K_d=\raisebox{1ex}{$\frac{{\left[Pd\right]}_{re}}{{\left[impurities\right]}_{re}}$}\!\left/ \!\raisebox{-1ex}{$\frac{{\left[Pd\right\}}_{start}}{{\left[impurities\right]}_{start}}$}\right.$$

(1)

Data for calculating Kd are presented in

Table 8. Element composition of the initial solution and aqueous solution after stripping.

Solutions	Element Composition of the Solution
	g/L			mg/L
	Pd	Ru	Rh	Ag	Cs	Ce	Cr	Fe	Mo	Sr	Zr	Total Impurities
Starting	12.21	4.82	7.50	3.5	13.25	34	39.5	262	88.5	4.1	97.5	12.86 g/L
Stripping	11.95	1.11	>DL	2.5	>DL	0.25	>DL	3.25	4.0	0.10	3.35	125 mg/L

>DL—below detection limits.

. The calculated Kd of palladium was above 100 (100.73).

Further, the exceptions of toxic solvent from the extraction mixture were tested. Ionic liquids usually possess good technological behavior: a proper density, viscosity, flashpoint, etc. Thus, the use of the [BMIM]Tf₂N itself as a diluent seems justified keeping in mind its affinity to PGM but not the impurities of the cations (Table 3). In new conditions, we probed two 2,2′-bipyridyl-2,2′-diamides, Ia and Ib, and also tetraoctyldiglicolamide II as a well-known extragent for PGM separation from HLW. The solution of Ia diamide in [BMIM]Tf₂N was effective toward PGM extraction (

Table 9. Degrees of extraction of elements during extraction with 0.2 M solution of diamides in [BMIM]Tf₂N.

Diamide	Degrees of Extraction, %
	Pd	Ru	Rh	Ag	Cs	Ce	Cr	Fe	Mo	Sr	Zr
Ia	99.9	42.5	0	98.5	7.5	5.6	0	0	0	0	9.8
Ib	99.9	51.4	0	98.5	7.5	0	0	0	0	0	15.7
II	99.7	21.6	0	0	0	0	0	0	88.3	0	93.9

): palladium and silver were quantitatively extracted. Moreover, higher ruthenium extraction was observed compared to nitrobenzene or chloroform diluents, where the extraction of ruthenium did not exceed 11%. However, the efficiency of selective extraction of palladium from the model solution is lower compared to experiments where a solvent was used, and the ionic liquid was used as an additive. Moreover, the effectiveness of metal separation depends on the backbone substituents in amide fragment of the molecule. The electron-withdrawing fluorine atom decreases the selectivity both within PGM and in relation to zirconium. The extraction ability of diamide II with respect to palladium is not inferior in efficiency to compounds Ia and Ib. It should be noted that the recovery of ruthenium is about 20% in [BMIM]Tf₂N media, which is comparable with the data obtained eelier for this diamide. However, it concedes the Ia and 1b in [BMIM]Tf₂N media in selectivity toward Zr and Mo. As a result, it is difficult to solve the problem of selective isolation of palladium from solutions in which molybdenum and zirconium ions are present.

Table 10. Separation factors for palladium and other elements during extraction with 0.2 mol/L solutions of Ia and Ib in [BMIM]Tf₂N.

Diamide	SF
	Pd/Ru	Pd/Rh	Pd/Ag	Pd/Cs	Pd/Ce	Pd/Cr	Pd/Fe	Pd/Mo	Pd/Sr	Pd/Zr
Ia	1.8 × 103	-a	19	1.6 × 104	2.2 × 104	-	-	-	-	1.2 × 104
Ib	1.3 × 103	-	20	1.7 × 104	-	-	-	-	-	7.3 × 103

a—Not applicable.

shows the separation factors for palladium and impurity elements when carrying out extraction under the conditions described above.

As can be seen from Table 10, palladium is extracted into the organic phase quite selectively with a separation factor greater than thousand for all ligands. So, the ionic liquid media are capable of selective palladium separation from PGM and increase the selectivity of the extractants.

3.2. Formation of Silver Complex with 2,2′-bipyridyl-6,6′-dicarboxamide

To understand the structure of the silver complex of diamides of type I in the concept of high extraction of this metal, we synthesized the corresponding complex from silver nitrate and model bipyridyl type ligand Ic (Scheme 1). The diamide Ic was chosen due to its moderate solubility in polar organic media compared to electron-donating diamide Ia.

Silver complex IcAg readily precipitates from the reaction mixture as colorless microcrystals with a good yield. As it possesses moderate solubility in acetonitrile, the 1H NMR spectrum was recorded to obtain information about the coordination of metal ions with the heterocycle. In the spectrum, there are several separate signals indicating the high symmetry of the complex. The chemical shifts of the complex significantly differ from that of the ligand Ic [23]. Signals of protons in pyridine moiety and in the methylene group of the ethylamide side chain are upfield shifted, proving the binding of the metal atom with heterocycle and amide oxygen atoms. So, silver atoms enter the cavity of the tetradentate ligand and bond to nitrogen and two oxygen atoms. It appears that the complex possesses an ion paired nature; the presence of coordinated silver cation is proved by MALDI-TOF mass spectrometry: only one peak corresponding to the [LM]⁺ composition is found using mass spectrometry.

4. Conclusions

Conditions have been found for the extraction purification of palladium from raffinates of the PUREX process using 2,2′-bipyridyl-6,6′-dicarboxamides in a solvent mixture of [BMIM]Tf₂N with chloroform or nitrobenzene. In these mixtures, palladium is selectively extracted into the organic phase with a separation factor >10⁴. The conditions for back-extraction of palladium using 5 M nitric acid were found; under these conditions, palladium is re-extracted from the loaded organic phase almost completely. The loaded capacity of the proposed extractant is 58.3 g per 1 mol of N,N′-diethyl-N,N′-bis(4-hexylphenyl)-[2,2′-bipyridine]-6,6′-dicarboxamide. The final purification factor of palladium from feeding solution ranged from 20 to 3000.

Data on the extraction of metals using solutions of 2,2′-bipyridyl-6,6′-dicarboxamides in PhNO₂ and CHCl₃ show that the reagents exhibit moderate coordination ability with respect to noble metals and do not bind impurity elements. We hypothesize that this property is due to the coordination of metal ions with tetradentate ligands. Chemical synthesis of the silver complex confirmed the possibility of the formation of the silver complex due to coordination with the “soft” donor centers of the ligand (heterocyclic nitrogen atoms). The “hard” donor centers of the ligand—the oxygen atoms of the carboxamide groups—are probably not close enough to each other to form stable complexes with “hard” impurity metals. The introduction of an additional reagent capable of forming complexes with metals, such as TBP, into the extraction system leads to a decrease in the extraction of noble metals by the dicarboxamide reagent and, at the same time, an increase in the extraction of impurity metals. The introduction of lipophilic cations (TOMAN, [BMIM]Tf₂N) into the extraction system leads to different results depending on the nature of the cation. TOMAN, similar to TBP, reduces the extraction of noble metals while increasing the extraction of Cs, Mo, Fe and Zr. The introduction of [BMIM]Tf₂N leads to a synergistic increase in the extraction of noble metals without leading to the extraction of impurity metals. At the same time, [BMIM]Tf₂N itself or its solutions in PhNO₂ and CHCl₃ selectively extract ruthenium, but not palladium and silver ([BMIM]Tf₂N and its solution in CHCl₃), or non-selectively bind impurity metals ([BMIM]Tf₂N in PhNO₂). Therefore, the selective joint quantitative extraction of palladium and silver with solutions of 2,2′-bipyridyl-6,6′-dicarboxamides in [BMIM]Tf₂N is synergistic; both dicarboxamides and the presence of [BMIM]Tf₂N are equally important for the appearance of this effect.