The Extraction Mechanism of Zirconium and Hafnium in the MIBK-HSCN System

Abstract

The extraction of zirconium (Zr) and hafnium (Hf) in methyl isobutyl ketone (MIBK)—thiocyanic acid (HSCN) system has been widely used in the production of nuclear-grade zirconium and hafnium in industry, while the extraction mechanism was not adequately studied. In this study, the extraction and stripping equilibrium of Zr and Hf in the MIBK-HSCN system was studied. The results showed that elevated HCl concentration can increase the distribution ratio of SCN⁻ and decrease that of Zr/Hf in organic phase. In the stripping process, HCl concentration and the Organic/Aqueous (O/A) phase ratio played important roles. The mechanism of the extraction reaction was discussed by considering the stoichiometric relationship of possible reaction equations and corresponding equilibrium constants. The results indicated that SCN⁻ could be extracted into MIBK as HSCN·MIBK. Meanwhile, SCN⁻ could also be extracted into MIBK by complexing with metal (Zr or Hf). The molar ratios of MIBK to the complexes of Zr and Hf have been found to be 5.34 and 5.03, respectively. With the increase in the initial concentration of HCl in the aqueous phase, the complexation molar ratios of SCN⁻ to Zr and Hf increased first and then decreased, and so do the extraction equilibrium constants, which might be due to the extraction competition of HSCN and metal complexes.

1. Introduction

Zr and Hf are widely used in many fields, benefiting from their unique properties, such as ferrous and non-ferrous metallurgy, ceramics, refractories, glass, and industrial catalysis. In the nuclear chemical industry, Zr has a low effective capture cross-section for thermal neutrons, making it an attractive construction material for nuclear reactors in the nuclear chemistry sector. Since Hf has a high thermal neutron absorption capacity [1], its presence in Zr reduces the material’s suitability for use in nuclear processes by increasing the effective cross-section for thermal neutron capture. As a nuclear material, the content of Hf in Zr cannot be higher than 100 mg·kg⁻¹ [2].

Zr and Hf typically coexist in nature as they have similar chemical properties. There are currently a number of techniques available for separating Zr and Hf, such as fractional crystallization [3], selective liquid extraction [4], ion exchange [5,6], extractive distillation [7], liquid membrane [8], and so on. Benefiting from its high production capacity, low cost, ease of automation, safe and quick operation, and complete separation of Zr and Hf, solvent extraction is the most frequently used method [9]. Among the existing solvent extraction systems of Zr and Hf, the MIBK-HSCN system [10], TBP-HNO₃ system [11], and N235-H₂SO₄ [12] are widely used. The MIBK-HSCN system has a shorter separation process and consumes much less organic extractant than other systems because it preferentially extracts Hf (which typically has a relatively small proportion of components) from the aqueous phase.

The studies of separation between Zr and Hf with MIBK-HSCN were widely focused on process optimization. Fischer et al. investigated the distribution of zirconium and hafnium in the aqueous and MIBK phases [9]. They discovered that the distribution ratio decreased by 2% for every 1 °C rise in temperature in the range of 15–35 °C. Additionally, they proposed that the separation factor of zirconium and hafnium could be increased by increasing SCN⁻ concentration and decreasing HCl concentration [9]. Otsuka investigated the effect of the concentrations of SCN⁻ and H⁺ in the aqueous phase and the concentration of HSCN in the organic phase on the distribution ratio and separation factor of metals. Based on the study, they provided a liquid-liquid separation procedure for producing nuclear-grade zirconium [13]. Zarpelon et al. discussed the range of free acid value in the aqueous phase and SCN⁻ concentration in the organic phase when the MIBK-HSCN system had the best separation effect for Zr and Hf [14]. Lu et al. studied the effect of the concentration of H⁺ and (NH₄)₂SO₄ on the extraction separation of zirconium and hafnium in the MIBK-HSCN system [15]. However, the literature on the extraction mechanism of Zr/Hf in the MIBK-HSCN system is sparse.

Recently, studies on the separation mechanism of Zr and Hf in some new systems have been paid more attention, including tributyl phosphate (TBP)-HNO₃ system, Cyanex572-HNO₃ system, N, N-di-2-ethylhexyl diglycolamic acid (D2EHDGAA)-HCl system, Aliquat 336-HSCN system and so on [16,17,18,19]. As for the MIBK-HSCN system, some researchers considered that the complex molar ratios of MIBK/Metal (Zr or Hf) and SCN⁻/Metal (Zr or Hf) are 2:1 and 4:1, respectively [20,21]. Krishna [21] found out that the Hf complex has larger interaction energy with the ligands than the Zr complex through the molecular dynamics simulation method based on the assumption of SCN⁻/Metal (Zr or Hf) equal to 4:1. However, some studies showed that Zr⁴⁺/Hf⁴⁺ have a strong affinity with OH⁻ and could form different hydrolysis complexes [22,23]. Stern et al. [24] discovered that Zr⁴⁺ is capable of forming hydrolysis complexes in aqueous solutions even when the H⁺ concentration exceeds 1 mol/L through density functional theory. Sommers [25] found out that Zr/Hf and SCN⁻ in the organic phase can both form oxo-centered tetramer cluster with a core formula of [OM₄(OH)₆(NCS)₁₂]⁴⁻ under two certain extraction conditions through single-crystal X-ray diffraction, where SCN⁻/Metal (Zr or Hf) equals 3:1. Besides, Zr could form another cluster, where SCN⁻/Zr was 6/5. These findings suggest that the solution chemistry can significantly impact the metal complex form across a wide range.

Ions and molecules present in an aqueous solution interact either with ions with opposite charges or with neutral molecules. These interactions can be described in terms of Debye-Huckel theory or chemical equilibrium. However, these theories are more applicable to the case of diluted solutions than industrially produced solutions, which are mostly very concentrated and contain high ionic strength and high acidity. The mechanisms that occur in many solvent extraction industrial systems are difficult to characterize. The mechanism by which species are extracted can be significantly affected by many variables in the system. Therefore, each individual variable and the way in which each variable affects the extraction mechanism must be evaluated. Besides, the HSCN and metal complexes can both be extracted into MIBK, which increases the complexity of the extraction process of Zr/Hf in the MIBK-HSCN system. In such complex extraction system, understanding solution chemistry at atomic level under a wide range of operating conditions is crucial for the separation process of Zr and Hf.

Herein, MIBK was specifically employed for the extraction and separation of the trace element Hf from an highly concentrated Zr aqueous solution (˃90 g·L⁻¹). The extraction mechanism of Zr and Hf in the MIBK-HSN system under different operational conditions was studied. Firstly, the rules of the HSCN distribution in MIBK and the separation of Zr and Hf were explored. Secondly, the specific stoichiometric constants of the reaction equations and the possible metal complex forms in the two phases in the extraction of Zr and Hf with the MIBK-HSCN system were investigated using slope analysis. Finally, the competition between HSCN and metal complexes in the extraction process was observed and discussed.

2. Materials and Methods

2.1. Sample and Reagents

The zirconium oxychloride (Zr(Hf)OCl₂·8H₂O, 99.9%) used in this study was provided by Shenzhen SinoHf Technology Co., Ltd. (Shenzhen, China), in which the total content of Zr(Hf)OCl₂·8H₂O is 99.9% and the content of Hf is 7.8% in the count of the whole mass of Zr and Hf. The extractant used in this study was methyl isobutyl ketone (MIBK, AR, 99.5%), also provided by Shenzhen SinoHf Technology Co., Ltd. The structure of MIBK is shown in Section 3 of supporting information for reference. Inorganic reagents including NH₄SCN (AR, 98.5%), HCl (AR, 37 wt%), and Ce(SO₄)₂ (AR, 98%) were purchased from Sinopharm Group (Holding) Co., Ltd. (China). All aqueous solutions were prepared using de-ionized water, and the concentration of MIBK in the experiments in this study was changed by diluted with kerosene (boiling point range: 190–250 °C).

2.2. Experimental

In this study, all extraction and stripping experiments were carried out using separation funnels at room temperature (25 ± 0.5 °C). The aqueous feed was prepared by dissolving Zr(Hf)OCl₂⋅8H₂O, HCl, and NH₄SCN in de-ionized water. The feed-extractant mixture was shaken for 20 min to make sure that the reaction reached equilibrium and then left for another 10 min to completely separate the two phases. To study the effects of the concentrations of MIBK (1–3 M) and H⁺ (0–2 M) on the extraction of MIBK, SCN⁻ and HCl were respectively diluted in kerosene and de-ionized water to a certain concentration. Under the same dilution method, MIBK, NH₄SCN, and HCl were diluted to a certain concentration, respectively, to study the effects of MIBK (4–8 M), SCN⁻ (2.5–3.5 M) and H⁺ (0–1.5 M) concentrations on the extraction of Zr and Hf. In addition, to develop the equilibrium curves of SCN, Zr, and Hf under certain experimental conditions, NH₄SCN (0–3 M), Zr (0–0.4 M), and Hf (0–0.01 M) were diluted to a certain concentration, respectively.

The concentrations of the metals in the aqueous phase were measured by Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES, Plasma Quant PQ 9000 Elite, Analytick-jena, Germany), and the concentration of metal in the loaded organic phase was calculated by mass balance. The concentration of H⁺ of the aqueous phase when the extraction reaction reached equilibrium was measured with a multi-parameter tester (SevenExcellence, S400-K, METTLER TOLEDO, America). The organic phase was analyzed by Fourier transform infrared (FT-IR) spectroscopy using a Nicolet iS5 instrument, in a range of 4000–400 cm⁻¹. The molecular weight of the extracted complexes was determined using electrospray ionization-mass spectrometry (ESI-MS, Q Sight 120, PerkinElmer, America).

The concentration of SCN⁻ ($C_{\left({SCN}^{-}\right)}$) was measured by titration using ceric sulfate solution, the equipment used for titration was an acid burette. The ceric sulfate solution ($C_{\left({{Ce(SO}_4)}_2\right)}$) for the titration was prepared by dissolving precisely weighed ceric sulfate in sulfuric acid solution (containing 28 mL concentrated sulfuric acid) and then diluting to 1000 mL. Sodium oxalate was used to calibrate the concentration of the cerium sulfate solution. Aqueous solutions for analysis were diluted with de-ionized water, organic solutions were diluted with isopropyl alcohol. The solution to be tested was prepared by diluting by a certain factor (x) and then precisely adding 1 mL diluted solution into 50 mL HCl solution (the volume ratio of concentrated HCl to H₂O is 1:4). The concentration of SCN⁻ could be calculated by the volume (V) of ceric sulfate solution consumed during titration, as shown in Equation (1).

$$C_{\left({SCN}^{-}\right)}=\frac{x·C_{\left({{Ce(SO}_4)}_2\right)}·V}{6}$$

(1)

Distribution ratio (D) was defined as the ratio of the concentration of species in the organic phase to that in the aqueous phase when the extraction reaction reached equilibrium. The distribution ratio and separation factor (SF) were calculated using Equations (2) and (3), where the “aq” and “org” subscripts indicate the aqueous phase and the organic phase, respectively.

$$D=\frac{{\left[M\right]}_{\left(org\right)}}{{\left[M\right]}_{\left(aq\right)}}$$

(2)

$$SF=\frac{D_{Hf}}{D_{Zr}}$$

(3)

where D_Hf and D_Zr are the distribution ratios of Hf and Zr, respectively.

2.3. Stoichiometry and Calculation of Equilibrium Constants

Since the change of ionic strength will affect the activity coefficient, to avoid the error caused by the change of ionic strength on the research results, ammonium chloride is used to keep the ionic strength consistent under different experimental conditions. Furthermore, the concentration of the extractable species is roughly equal to the total concentration of the metal in the aqueous phase, and the influence of the rare impurity on the extraction was ignored in the study. These assumptions imply that the form of metal complexes extracted into the organic phase is determined under a specific condition. In order to determine the extraction stoichiometry of Zr and Hf and to explore the competitive extraction relationship between Zr, Hf, and HSCN, the relationship between the distribution ratio and the concentration of MIBK, SCN⁻, and HCl was studied by slope analysis [16,17,18,19].

3. Results and Discussion

To improve the separation of Zr and Hf, the impact of various experimental conditions on the extraction of Zr, Hf, and HSCN was studied. Meanwhile, the reaction equations and the corresponding equilibrium constants in the extraction process were investigated.

3.1. Extraction of SCN⁻

We learned through exploratory experiments that SCN⁻ can be extracted into the organic phase even though no metal ions are present in the aqueous phase. Considering the only two nonmetallic cations, i.e. H⁺ and NH₄⁺, SCN⁻ might be extracted into the organic phase as in the form of H⁺_α NH₄⁺_1−α SCN. Herein, the effects of H⁺ and MIBK concentrations on the extraction of SCN⁻ were investigated, and the molar ratios of SCN⁻/MIBK and H⁺/SCN⁻ in the extraction were established.

Equations (4) and (6) were proposed to describe the reaction of H⁺/SCN⁻ and MIBK/H_αSCN^α−1 in the HCl medium, and K₁ and K₂ were defined as the equilibrium constants for Equations (4) and (6), which were shown in Equations (5) and (7).

$$\alpha {H^{+}}_{\left(aq\right)}+{{SCN}^{-}}_{\left(aq\right)}\leftrightharpoons {H_{\alpha }{SCN}^{\alpha -1}}_{\left(aq\right)}$$

(4)

$$K_1=\frac{\left[{H_{\alpha }{SCN}^{\alpha -1}}_{\left(aq\right)}\right]}{{\left[{H^{+}}_{\left(aq\right)}\right]}^{\alpha }\left[{{SCN}^{-}}_{\left(aq\right)}\right]}$$

(5)

$${H_{\alpha }{SCN}^{\alpha -1}}_{\left(aq\right)}+\beta {MIBK}_{\left(free\right)}\leftrightharpoons {H_{\alpha }{SCN}^{\alpha -1}·\beta MIBK}_{\left(org\right)}$$

(6)

$$K_2=\frac{\left[{H_{\alpha }{SCN}^{\alpha -1}·\beta MIBK}_{\left(org\right)}\right]}{\left[{H_{\alpha }{SCN}^{\alpha -1}}_{\left(aq\right)}\right]{\left[{MIBK}_{\left(free\right)}\right]}^{\beta }}$$

(7)

where α and β represented the molar ratio of H⁺/SCN⁻ and MIBK/H_αSCN^α−11, respectively.

The slope method is a classical method to investigate stoichiometry and mechanisms in the extraction thermodynamics [17,19,26]. The concentration of free extractant should be theoretically used in stoichiometric analysis using the slope method. However, the relevant studies usually ignored the consumption of extractants in the extraction process and assumed that the free extractant is equal to the total extractant, which might have adverse impact on the result accuracy. Herein, the extractant consumed was taken into account, and stoichiometry and equilibrium constants of Equations (4) and (6) were calculated using the Levenberg-Marquarde method.

Combined Equations (2), (5) and (7), $log{D}_{\left(SCN\right)}$ can be expressed as Equation (8), the specific calculation process could be seen in Supporting Information in Section 1.

$$log{D}_{\left(SCN\right)}=log{K}_1+log{K}_2-log\left({1+K}_1{\left[H_{\left(aq\right)}^{+}\right]}^{\alpha }\right)+\alpha \mathrm{l}\mathrm{o}\mathrm{g}\left[H_{\left(aq\right)}^{+}\right]+\beta \mathit{log}(\left[{MIBK}_{\left(org\right)}\right]-\frac{\beta D_{\left(SCN\right)}}{1+D_{\left(SCN\right)}}\left[SCN\right])$$

(8)

According to Equation (8), we used experimental data to fit K₁, K₂, α and b, and the data used were shown in

Table 1. Experimental data of SCN extraction with changing MIBK and HCl concentration.

[MIBK]	[H+(aq)]	D(SCN)
1	1.698	0.194
1.4	1.667	0.288
1.8	1.622	0.383
2.6	1.288	0.614
3	1.259	0.731
4	1.202	1.064
5	1.072	1.425
8	1 × 10−5	0.011
8	0.055	0.271
8	0.076	0.319
8	0.132	0.486
8	0.178	0.663
8	0.191	0.646
8	0.263	0.874
8	0.363	1.106
8	0.398	1.289

The standard uncertainties of temperature and pressure are u(T) = 0.1 K and u(P) = 1 kPa, respectively. The relative standard uncertainty of [MIBK], [H⁺_(aq)], and D_(SCN) are 0.025, 0.05, and 0.01, respectively.

, where [H⁺] represents the equilibrium concentration of H⁺.

3.1.1. Effect of MIBK Concentration on the SCN⁻ Extraction

In this part, the distribution of SCN⁻ in the two phases with different MIBK concentrations was investigated. The distribution ratios of SCN⁻ in MIBK with different concentrations are presented in Figure 1, which indicates that the distribution ratio of SCN⁻ increased with the increase in MIBK concentration.

3.1.2. Effect of HCl Concentration on the SCN⁻ Extraction

In this part, we studied the effect of H⁺ concentration on the complexation ratio of H⁺/SCN⁻ in the organic phase. The extraction equilibrium of SCN⁻ under different HCl concentrations in the aqueous phase was shown in Figure 2, which indicated that the distribution ratio of SCN⁻ increased with the increasing concentration of HCl in the aqueous phase.

3.1.3. Stoichiometry and Equilibrium Constants

The values of K₁, K₂, α, and b were obtained by fitting Equation (8) using the least square method in Python based on the extraction equilibrium data listed in Table 1. The experimental and calculated values of the distribution ratio under different conditions fit well, as shown in Figure 3; R² and RMSE were 0.9957 and 0.0113, respectively. The values of K₁, K₂, α, and β obtained by fitting were 8.741, 0.033, 0.92, and 1.22, respectively. The complex molar ratios of H⁺/SCN⁻ and MIBK/HSCN in the organic phase were both almost 1:1, which indicated that SCN⁻ was extracted into the organic phase as a form of HSCN and form a complex of HSCN·MIBK. The large value of K₁ indicated that the HSCN is a weak acid, which is consistent with the conclusion of the study by Wang et al. [27].

3.1.4. FT-IR and ESI-MS Analysis

The FT-IR spectra of the organic phase before and after extraction were studied to elucidate the combination form of MIBK and HSCN in the extraction process, which were presented in Figure 4. The peak at 2027 cm⁻¹ was ascribed to C≡N triple-key telescopic vibration, while the peaks at 1730 cm⁻¹ for pure MIBK and at 1698 cm⁻¹ for MIBK with HSCN were attributed to C=O double-key telescopic vibration. The change in MIBK FT-IR spectrum after extraction of HSCN included the addition of the C≡N vibration peak at 2027 cm⁻¹ and the shift of the C=O vibration peak from 1730 cm⁻¹ to 1698 cm⁻¹. The red shift of the C=O peak might be due to the hydrogen bonding between C=O and HSCN, indicating that the possible binding site of MIBK with HSCN is located proximal to the C=O double bond.

The combination form of MIBK and HSCN in the extraction process was also determined using the ESI-MS spectrum, as illustrated in Figure 5. The signal at m/z = 58.2, 99.2, and 157.7 in the negative ion spectrum of the HSCN-loaded organic phase could be assigned to the HSCN, MIBK, and HSCN·MIBK, respectively. The results agree with the discussion in Section 3.1.1 and Section 3.1.2.

3.2. Extraction of Zr and Hf

The complex reactions of Zr/Hf with SCN⁻ in the aqueous phase and the extraction reactions of metal complexes are discussed in this section. Recent studies showed that there is no evidence for the presence of MO²⁺ in the aqueous solution [28,29,30,31]. Thus, the stable uncomplexed form of Zr/Hf is considered as M⁴⁺ ion. The reaction of Zr and Hf with SCN⁻ in the aqueous phase was expressed in Equation (9), and K₃ was defined as the equilibrium constant of Equation (9), which was shown in Equation (10). Equation (11) describes the reaction of MIBK with a metallic complex in the MIBK-HSCN extraction system. K₄ was defined as the equilibrium constant of Equation (11), as expressed by Equation (12).

$${M^{4+}}_{\left(aq\right)}+{\delta {SCN}^{-}}_{\left(aq\right)}\leftrightharpoons {{M\left(SCN\right)}_{\delta }^{4-\delta }}_{\left(aq\right)}$$

(9)

$$K_3=\frac{\left[{{M\left(SCN\right)}_{\delta }^{4-\delta }}_{\left(aq\right)}\right]}{\left[{M^{4+}}_{\left(aq\right)}\right]{\left[{{SCN}^{-}}_{\left(aq\right)}\right]}^{\delta }}$$

(10)

$${{M\left(SCN\right)}_{\delta }^{4-\delta }}_{\left(aq\right)}+\gamma {MIBK}_{\left(org\right)}\leftrightharpoons {{M\left(SCN\right)}_{\delta }^{4-\delta }·\gamma MIBK}_{\left(org\right)}$$

(11)

$$K_4=\frac{\left[{{M\left(SCN\right)}_{\delta }^{4-\delta }·\gamma MIBK}_{\left(org\right)}\right]}{\left[{{M\left(SCN\right)}_{\delta }^{4-\delta }}_{\left(aq\right)}\right]{\left[{MIBK}_{\left(org\right)}\right]}^{\gamma }}$$

(12)

where δ represents the molar ratio of SCN⁻/M (Zr or Hf), and γ represents the complexation molar ratio of MIBK/${\mathrm{M}\left(\mathrm{S}\mathrm{C}\mathrm{N}\right)}_{\mathsf{\delta }}^{4-\mathsf{\delta }}$, the metal complex extracted from the aqueous phase was largely present as a single species [9].

Combined Equations (2) and (12), $D_{\left({M\left(SCN\right)}_{\delta }^{4-\delta }\right)}$ can be written as Equation (13).

$$D_{\left({M\left(SCN\right)}_{\delta }^{4-\delta }\right)}=K_4·{\left[{MIBK}_{\left(org\right)}\right]}^{\gamma }$$

(13)

$${logD}_{\left({M\left(SCN\right)}_{\delta }^{4-\delta }\right)}=log{K}_4+\gamma log\left[{MIBK}_{\left(org\right)}\right]$$

(14)

After a few substitutions, the specific calculation process could be seen in Supporting Information in Section 2, Equations (15) and (16) could be obtained.

$$D_{\left(M\right)}={K_4·{\left[{MIBK}_{\left(org\right)}\right]}^{\gamma }·K}_3·{\left[{{SCN}^{-}}_{\left(aq\right)}\right]}^{\delta }$$

(15)

$$log{D}_{\left(M\right)}=log{K}_4+\gamma log\left[{MIBK}_{\left(org\right)}\right]+log{K}_3+\delta \mathit{log}\left[{{SCN}^{-}}_{\left(aq\right)}\right]$$

(16)

Using Equations (14) and (16), The stoichiometric and equilibrium constants can be determined by logarithmic values of the species distribution ratio versus that of the variables. The slopes of the curves fit by experimental data presented the stoichiometry of the reaction, and the equilibrium constants could be calculated using the intercepts.

3.2.1. Effect of MIBK Concentration on the Zr and Hf Extraction

The influence of MIBK concentration on the distribution ratio and separation factors of Zr and Hf was presented in Figure 6, which indicated that increasing MIBK concentration would increase the distribution ratios of metallic species in the organic phase. However, the separation factor between Hf and Zr decreased with the increasing MIBK concentration in the experimental range. It should be noted that the change of metal concentration in the aqueous phase before and after extraction was lower than the detection limit of ICP-OES when the concentration of MIBK in the organic phase is low, which made some of the data unavailable, thus the calculation range of separation factor was 6–8 mol/L of MIBK concentration.

The relationship between ${logD}_{\left({M\left(SCN\right)}_{\delta }^{4-\delta }\right)}$ and $log\left[MIBK\right]$ as shown in Figure 7, where the slope of the lines represented the stoichiometric coefficient of Equation (11). As shown in Figure 7, the complexation molar ratio of MIBK/Hf and MIBK/Zr was 5.03 and 5.34, respectively. The values of K₄ for Hf and Zr could be estimated according to the equation shown in Figure 8, which were 2.692 × 10⁻⁵ and 3.311 × 10⁻⁶, respectively. The results showed that the equilibrium constant of Hf was about ten times higher than that of Zr, indicating that the extraction efficiency of the MIBK-HSCN system for Hf is much higher than that of Zr under experimental conditions. The result confirmed that extraction of Hf is preferred, and more significantly, the difference in extraction efficiency between Zr and Hf in MIBK-HSCN system was quantified.

3.2.2. Effect of SCN⁻ Concentration on the Zr and Hf Extraction

In this part, the effect of SCN⁻ concentration on extraction separation of Zr and Hf was investigated, and the possible molar ratios of SCN⁻/Zr and SCN⁻/Hf in solvent extraction at different concentrations of HCl were discussed.

The distribution ratios and separation factors of Zr and Hf at different SCN⁻ initial concentrations were shown in Figure 8, which indicated that with the increase in SCN⁻ concentration, the distribution ratios of Zr and Hf increased. This might be due to that the increased concentration of SCN⁻ separation factor between Zr and Hf changed little with the increasing SCN⁻ concentration.

Since γ and K₄ were already obtained in Section 3.2.1, combined with Equation (16), δ and K₃ can be obtained from the data in Figure 9. The relationships between $log{D}_{\left(M\right)}$ and $\mathit{log}\left[{{SCN}^{-}}_{\left(aq\right)}\right]$ under different HCl concentrations were shown in Figure 9a–e. The influence of the initial HCl concentration in the aqueous phase on the molar ratios of SCN⁻/metal (Zr or Hf) is shown in Figure 9f. At different initial concentrations of HCl, the complexation molar ratio of SCN⁻/Metal (Zr or Hf) and the equilibrium constants of the extraction reaction are shown in

Table 2. The complexation molar ratios (δ) of SCN⁻/metal (Zr or Hf) and the equilibrium constants (K₃) of extraction reaction at different initial HCl concentrations.

HCl Concentration (M)	δ (Zr)	δ (Hf)	K3 (Zr)	K3 (Hf)
0	3.34	3.30	1.311	0.704
0.5	4.44	4.52	9.279	7.370
1	4.77	4.80	5.721	6.273
1.5	4.46	4.13	3.780	4.241
2	2.97	2.85	2.497	3.292

.

As shown in Figure 9a, the molar ratios of SCN⁻/Zr and SCN⁻/Hf were around 3.3:1 with 0 M initial HCl concentration, which is close to that in the study of Sommers et al., who found that both Hf and Zr form an oxo-centered tetramer cluster with a core formula of [OM₄(OH)₆(NCS)₁₂]⁴⁻ [25]. As shown in Figure 9b–e, the complex molar ratio of SCN⁻/metal (Zr or Hf) extracted by MIBK increased with the increase in the initial HCl concentration (when HCl concentration was lower than 1 M) but decreased significantly with the increase in the initial concentration of HCl (when HCl concentration was higher than 1 M). This indicated that the composition of the extractable metal complex changed at different initial HCl concentrations. When the initial HCl concentration was lower than 1 M, the smaller molar ratio of SCN⁻/metal (Zr or Hf) might be caused by the hydrolysis of Zr⁴⁺ and Hf⁴⁺ [22,23,24]. Zr and Hf have a strong affinity for OH⁻ when the H⁺ concentration in the aqueous phase is low. A study showed that Hf(OH)³⁺ is present in the solution with a pH value of lower than 1 [32]. However, the decrease of the molar ratios of SCN⁻/metal (Zr or Hf) at the range of 1–2 M for initial HCl concentration might be caused by the competition between the formation of HSCN and metal complexes [15]. The dissociation of HSCN was incomplete at high H⁺ concentration [33], so SCN⁻ is more likely to be extracted into MIBK in the form of HSCN·MIBK with higher HCl concentration. Fischer et al. reported that Zr/Hf could complex with other anions except for SCN⁻ [9]; thus, incomplete dissociation of HSCN might decrease the complexation molar ratio of SCN⁻/metal (Zr or Hf).

In addition, as shown in Figure 9f, the molar ratios of SCN⁻/metal (Zr or Hf) were similar under the same experimental conditions, indicating that the influence of the concentration of SCN⁻ on the separation of two metals was insignificant [19]. This is consistent with the separation factor shown in Figure 8, which showed that the separation factor is less affected by SCN⁻ concentration.

Based on the intercepts of the lines in Figure 9, the corresponding values of K₃ for Hf and Zr were calculated and shown in Table 2. The stability constants of Hf complexes are generally lower than those for Zr in H₂SO₄ solution because the charge density of the two metals is different [34], but in the MIBK-HSCN system, the acid concentration played an important role in the complex stability of Zr/Hf and SCN⁻. The complex stability of SCN⁻ and Hf was greater than that of SCN⁻ and Zr when the HCl concentration was equal to or greater than 1 M, and the opposite is true when the concentration was lower than 1 M. In the traditional concept, the selective extraction of Hf is more favorable than Zr in thiocyanate media because the SCN⁻ has a stronger tendency to form complexes with Hf than with Zr [35]. This was thought to be the result of the binding energy of SCN with Zr is higher than that of Hf. The conclusion was obtained under conditions with high H⁺ concentration, as the optimized H⁺ concentration was higher than 1 M. The results were inconsistent with their conclusion when the HCl concentration was lower than 1 M. The discrepancy might be caused by the possible effect of H⁺ on the charge density of Zr/Hf in the aqueous solution. In a solution with a low H⁺ concentration, Zr⁴⁺ might have a larger charge density, which leads to formation of a more stable complex with SCN⁻.

When the initial concentration of HCl was greater than 0.5 M, K₃ (Zr or Hf) decreased with the increase in the concentration of HCl, which indicated that the stability of extractable metal complexes was declined. This is consistent with the trend that the distribution ratio of metals decreased with the increase in initial HCl concentration when the initial concentration of HCl was higher than 0.5 M, which will be described in Section 3.2.4. Referring to the study of Tan et al. [36], the decrease of K₃ (Zr or Hf) was considered to be caused by the changes in metal complexes formed at different HCl concentrations, which might be due to the competition between the formation of HSCN and metal complexes described above [15]. Since the equilibrium constant for Zr was more sensitive to the change in HCl concentration, the influence of HCl concentration on the complexation reaction of SCN⁻/Zr was more significant than that of SCN⁻/Hf. Thus, a proper concentration of HCl plays an important role in the separation of the two metals.

3.2.3. FT-IR Analysis

The FT-IR spectra of the organic phase before and after the extraction of ZrOCl₂ were analyzed to study the changes in the organic phase during the extraction process, as illustrated in Figure 10.

The obvious change in the organic phase after extraction is the red shift of C≡N vibration peak from 2027 cm⁻¹ to 2021 cm⁻¹, which may be due to the formation of a coordination bond between C≡N and metal. Therefore, the potential binding site of metal to HSCN is the vicinity of C≡N triple bond. The results agree with the study of Sommers et al. [25].

3.2.4. Effect of HCl Concentration in the Zr and Hf Extraction

In this part, the extraction equilibrium of Zr and Hf under different HCl concentrations in the aqueous solution was discussed. The extraction equilibrium data were obtained by changing ZrOCl₂·8H₂O concentration in the aqueous solution with a certain HCl concentration.

The effect of HCl concentration on the distribution of Zr and Hf was shown in Figure 11, where [Metal_(aq)] and [Metal_(org)] represents the equilibrium metal concentration in the aqueous phase and the organic phase, respectively. It could be observed from Figure 12 that the distribution ratios of Zr and Hf decreased with the increasing initial HCl concentration at the range of 0.5 to 1 M. This is in agreement with the results of competition between the formation of HSCN and metal complexes mentioned in Section 3.2.2. However, the increase in HCl concentration had little effect on the distribution of Hf and had a negative effect on the distribution of Zr in the organic phase at the range of 0 to 0.5 M for HCl concentration. Moreover, the change in HCl concentration had a greater influence on the distribution of Zr compared to that of Hf.

The effect of HCl concentration on the distribution ratio and the separation factor when NH₄SCN concentration was 2.5 M was studied in detail, and the results are presented in Figure 12. The calculated values of the distribution ratio of Zr and Hf were obtained according to Equation (16), which fit well with the experimental values of the distribution ratio of Zr/Hf, where R² was equal to 0.9997/0.9995. It could be seen from Figure 12 that with the increase in initial HCl concentration, the distribution ratio and separation factor increased first and then decreased. The change in distribution ratio agreed with the equilibrium constants of metal complexes formation, as shown in Table 2. Combined with the separation factor, it can be considered that when the initial concentration of HCl was 1 M, a higher separation factor and more stable metal species to be extracted can be obtained. The maximum extraction amount of Zr/Hf was obtained at 0.5 M HCl concentration, and the extraction amount of Zr or Hf decreased with the increase in HCl concentration in the range of 0.5–1.5 M HCl concentration, which might be caused by the competition between the formation of HSCN and the formation of metal complexes mentioned in Section 3.2.2. As the concentration of HCl increased, more SCN⁻ in the form of HSCN can be presented. The complexing process of Zr⁴⁺/Hf⁴⁺ with SCN⁻ must go through the dissociation of HSCN. Since HSCN is a weak acid, the hydrolysis of HSCN would restrain the complexation of Zr⁴⁺/Hf⁴⁺ with SCN⁻ [27].

3.3. Stripping of Zr and Hf from Loaded MIBK

Stripping of Zr and Hf from the loaded organic phase is necessary for separation between Zr and Hf. Thus, the stripping of Zr and Hf under different HCl concentrations and the different Organic/Aqueous (O/A) phase ratios were studied. The concentrations of Zr and Hf in the metal-loaded organic phase were 0.19 and 0.012 M, respectively. The effect of HCl concentration on the stripping ratio of Zr and Hf is shown in Figure 13. The results showed that Zr was preferential stripped by the HCl solution. The stripping ratios of Zr and Hf increased with the increase in HCl concentration, and when the HCl concentration was 3 M, the stripping ratios of Zr and Hf were 97.5% and 88.1%, respectively.

The effect of phase ratios on the stripping ratio of Zr and Hf is shown in Figure 14. As shown in Figure 14, with the increase in the volume ratio of the aqueous phase, the stripping ratio of Hf increased, and that of Zr first increased and then decreased, which was similar to the study of Wu et al. [26]. The separation efficiency between Zr and Hf was highest when the Organic/Aqueous (O/A) phase ratio was 1:1.

4. Conclusions

The extraction mechanism of Zr/Hf in the MIBK-HSCN system was investigated in this study. The results showed that SCN⁻ can be extracted into MIBK without metal as a form of HSCN·MIBK, whose possible binding site is located proximal to the C=O double bond. With higher HCl concentrations, the distribution ratio of SCN⁻ in MIBK is higher. In the extraction of Zr/Hf, the metal complexes can be extracted into MIBK with a stoichiometric constant of about 5:1 of MIBK/metal for Zr and Hf, while the equilibrium constant of Hf is about ten times larger than that of Zr. As for complexes of metal and SCN⁻, the molar ratio of SCN⁻/metal (Zr or Hf) and complexation equilibrium constants increased first and then decreased significantly with the increase in the initial concentration of HCl. The FT-IR spectra indicated that the metal and SCN⁻ might be complexed by the formation of a coordination bond between C≡N and metal. In the stripping process of Zr and Hf, the stripping ratio of Zr and Hf both increased with the increasing HCl concentration. However, with the increase in the aqueous volume ratio, the stripping ratio of Hf increased, and that of Zr first increased and then decreased.

Interpretation of the extraction processes of Zr/Hf shows that there existes competition between the formation of HSCN and metal complexes. When the HCl concentration is low, metal complexes are formed preferentially. It can be seen that the concentration of HCl NH₄SCN has a prominent influence on the separation factors of Zr/Hf.