Calix [ 4 ] arene-Based Amino Extractants Containing n-Alkyl Moieties for Separation of Pd ( II ) and Pt ( IV ) from Leach Liquors of Automotive Catalysts

Pt, Pd, and Rh platinum group metals (PGMs) are utilized in automotive catalytic converters to reduce the level of harmful gas emissions. To evaluate the separation efficiencies of PGMs from the leach liquors of automotive catalysts, their extractabilities from both single-component solutions and leach liquors were determined using three p-dialkylaminomethylcalix[4]arenes (alkyl: n-hexyl~n-octyl; 3–5) as extractants, and the obtained results were compared with the data of acyclic p-di-n-octylaminomethyldimethylphenol (6) and those previously reported for macrocyclic calix[4]areneand thiacalix[n]arene-based amine and amide extractants. Various extraction parameters, including the contact time, HCl concentration, and concentrations of Pd(II) and Pt(IV) ions, were examined, and the distributions and stoichiometry ratios of the recovered Pd(II) and Pt(IV) species were calculated. All macrocyclic extractants 3–5 exhibited high and selective extractabilities of Pd(II) and Pt(IV) ions contained in the leach liquors of automotive catalysts. The E% values of 3–5 from the leach liquors were 94.2, 93.0, and 97.7 for Pd(II) and 91.7, 94.0, and 92.5 for Pt(IV), respectively. Acyclic compound 6 did not perform well even though higher ratios of extractant used. Furthermore, the results of reusability testing demonstrated relatively large extractability values obtained for these two metals even after five extraction/stripping cycles.


Introduction
Calix [n]arenes have been used in catalysis, molecular recognition, ion separation, sensors, and so on [1][2][3].Calixarene-based extractants have been widely utilized for the selective separation of alkali and rare earth metals, recovery of uranium ions from brine, and removal of heavy metals from aqueous solutions [4][5][6][7][8][9].By performing chemical modifications of the functional groups located at the calixarene "upper" and "lower" rims, many functionalized calix [4]arenes were synthesized, and their ability to extract precious metals was intensively studied.Furthermore, as compared with the commercially available extractants, certain polydentate-functionalized calix [n]arenes can efficiently and selectively extract platinum group metals (PGMs) from solutions.
Owing to their scarcity in the Earth's crust and eccentrically located resources, PGMs (Pd, Pt, Rh, Ru, Ir, and Os) represent valuable key components of many industrial applications.The demand for PGMs remains very high as they are widely used in automotive and industrial catalysts, electronics, jewelry, dental materials, and electric/hybrid vehicles [10][11][12].However, the existing imbalance between the limited supply of PGMs and their high demand causes significant price fluctuations.PGMs are mainly produced only in a few countries, including South Africa, Russia, USA, and Zimbabwe [12].Therefore, the development of a PGM separation method from secondary resources is very important in terms of their effective recycling and sustainable use (typical secondary resources of PGMs include spent automotive catalysts, industrial catalysts, and electronic scraps [13]).The recovery of PGMs from spent automotive catalysts is very important because it contains appreciable amounts of Pd, Pt, and Rh than those of any other secondary resources (their typical amounts range from 1-2 g for a small car and 12-15 g for a big truck) [14].In Japan, the extraction of PGMs from spent automotive catalysts is performed by combining various pyrometallurgical and hydrometallurgical processes [10,[15][16][17].Solvent extraction is a hydrometallurgical technique for the separation and purification of PGMs utilizing commercial extractants, such as 2-hydroxy-5-nonylacetophenone oxime (LIX-84A), di-n-octyl sulfide (DOS), and tri-n-butyl phosphate (TBP).LIX-84A and DOS are widely used for Pd(II) extraction and TBP for Pt(IV) extraction.Concomitantly, commercial extractants are oxidized upon extended contact with highly acidic aqueous phases, rendering them ineffective for metal separation [10,[18][19][20][21]. LIX-84A is susceptible to hydrolysis under acidic conditions [19].DOS is oxidized to di-n-octyl sulfoxide (DOSO) during extraction upon contact with oxidizing agents in the acidic aqueous phase [20].Additionally, when TBP extracts Pt(IV) from the aqueous phase, adjusting over 5 M HCl in aqueous phase is required because of protonating TBP [10,21].Further, they are characterized by relatively low selectivity, low stability in highly acidic media, and low extraction efficiency [10,21,22].Therefore, the development of novel PGM extractants with high selectivity, high extraction rates, high separation efficiency, and high stability in acidic media is currently ongoing in many research groups [23][24][25][26][27][28][29][30][31][32][33][34].
Recently, we have reported that n-dialkylamino extractants based on acyclic thiodiphenols exhibited high extractabilities of Pd(II) and Pt(IV) ions [33].In particular, their use resulted in rapid Pd(II) and Pt(IV) separation from the liquors of acid-leached automotive catalysts containing nine different metals (Pt, Pd, Rh, Al, Ba, Ce, La, Y, and Zr) with high selectivity through an ion-exchange extraction mechanism.Similarly, n-dialkylamino-modified calix [4]arene extractants (1: alkyl = butyl and 2: alkyl = amyl; Scheme 1) were developed by our group to study their extractabilities towards PGMs [34].By introducing gradually longer n-alkyl chains (alkyl = ethyl~amyl) in calix [4]arene skeletons, it has been observed that the extracted species consisting n-dialkylamino-modified calix [4]arene with chloro-PGM anions after PGM extraction showed an increase of solubility for chloroform as compared with the short alkyl chains having calixarene-based extractants with respect to our previous result [34].These results prompted us to a further synthesis of higher chain length alkyl amino derivatives based on calix [4]arene.Additionally, influence of alkyl chain length and structure on copper(II) extraction by 5-alkyl-2-hydroxybenzaldoximes has also been reported [35,36].The reports have shown that alkyl chain length affects the ability to function as useful extractants, that is, influence of an extraction rate of copper(II) and a solubility for hydrocarbon diluents.
In this work, the PGM extractabilities of three calix [4]arenes containing di-n-alkylamino moieties at the calixarene "upper" rims, p-di-n-alkylaminomethylcalix [4]arenes (n-hexyl~n-octyl = 3-5; Scheme 1), were investigated, and the effects of the structural properties of macrocyclic calix [4]arene-based n-dialkylamino extractants (3-5) and acyclic p-di-n-octylaminomethyldimethylphenol (6) on the extraction of Pt(IV) and Pd(II) ions from HCl media were examined.Additionally, Pd(II) and Pt(IV) extraction percentages of the previous reported macrocyclic calix [4]arene-and thiacalix[n]arene-based amine and amide extractants were compared with the extractants 3-5.Specifically, this study was aimed at determining the efficiencies of Pd(II), Pt(IV), and Rh(III) extraction from single-component PGM solutions and leach liquors of automotive catalysts, stripping of the extracted metal species, and reusability of the developed extractants.

Solvent Extraction of PGM by 3-6
Solvent extraction experiments were performed in triplicate at a temperature of 20 ± 1 • C to ensure that the obtained extractability values (E%) were reproducible within a range of ±5%.For this purpose, 1.0 mM CHCl 3 solutions of the extractants (10 mL) and 0.1 mM standard solutions containing single Pd(II), Pt(IV), and Rh(III) ions (10 mL) in 50 mL-glass bottles were used.The analyzed samples were shaken thoroughly at a speed of 300 rpm using a mechanical shaker for desirable time, and their aqueous and organic phases were separated in a separating funnel.The concentrations of Pd(II), Pt(IV), and Rh(III) ions in the aqueous phase were measured by inductively coupled plasma atomic emission spectrometry (ICP-AES; SPS-3000; Seiko Instruments Inc., Chiba, Japan).Various extraction parameters, such as the contact time, metal/extractant ratio, and concentrations of extractants and HCl, have been examined.The obtained E% values and distribution ratios of Pd(II) and Pt(IV) species D were calculated using Equations ( 1)-( 3): where [M] aq,init and [M] aq are the initial and final concentrations of metal ions in solution, respectively.

Solvent Extraction of Pd(II) and Pt(IV) Ions from Leach Liquors of Automotive Catalysts
Leached solutions were prepared by treating the automotive catalyst residue with HCl (11.6 M) and H 2 O 2 (1 vol %) aqueous solutions [37].The resultant mixture containing nine metal ions (Rh, Pd, Pt, Zr, Ce, Ba, Al, La, and Y) was diluted with water 25 times by volume for the purpose of liquid-liquid extraction (see Table S1).After dilution, the pH of the leach liquor was equal to 1.22.The concentration of each metal ion in solution was determined by ICP-AES.Subsequently, a chloroform solution of the extractants (10 mL, 1.0 mM) was added to 10 mL of the aqueous solution of metal ions.The resulting mixture was shaken at a speed of 300 rpm and temperature of 20 ± 1 • C for 60 min, and the concentrations of metal ions in the aqueous phase [M] aq were measured by inductively coupled plasma atomic emission spectrometry (ICP-AES; SPS-3000; Seiko Instruments Inc., Chiba, Japan).

Stripping of Pt(IV) and Pd(II) Ions from Extractant Organic Phases
Stripping of Pt(IV) and Pd(II) ions from the two PGM-containing organic phases of the extractants was conducted by mixing 10 mL of each phase with 0.1 M thiourea in 0.1 M HCl solution (10 mL) as a stripping reagent.The obtained mixture was shaken at a speed of 300 rpm and temperature of 20 ± 1 • C, and the concentrations of Pt(IV) and Pd(II) ions in the aqueous phase after stripping [M] aq were determined by ICP-AES.The stripping efficiency (S%) was calculated using the following equation: where [M] aq is the concentration of Pt(IV) and Pd(II) ions in the aqueous solution after stripping, and [M] org is the concentration of these ions in the organic phase before stripping.

Reusability of Extractants
A fraction of the resulting organic phase (10 mL) after stripping was added to 20 mL of water and shaken for 10 min at a speed of 300 rpm.The resulting layers were separated using a separating funnel, and five extraction-stripping cycles of the liquid-liquid extraction process were performed under the conditions identical to those utilized for the leach liquors of automotive catalysts.Finally, the concentrations of all metal ions in the aqueous phase [M] aq were determined by inductively coupled plasma atomic emission spectrometry (ICP-AES; SPS-3000; Seiko Instruments Inc., Chiba, Japan).Proton nuclear magnetic resonance ( 1 H NMR) spectra were recorded on a JEOL JNM-ECA500 (Tokyo, Japan) instruments at 20 ± 1 • C in CDCl 3 .Chemical shifts are quoted as parts per million (ppm) relative to the peak of tetramethylsilane (TMS) as an internal standard.ATR-FTIR spectra were measured at 4000-600 cm −1 using a Thermo Fisher Scientific Nicolet iS5 spectrophotometer (attenuated total refection (ATR) method, MA, USA) with diamond mode by 16 scans.The internal reflections of ATR accessory is a Thermo Fisher Scientific Nicolet iD5 ATR accessory. 1H NMR and ATR-FTIR spectra of the native extractants 3-5 and the acid-treated extractants 3-5 were measured directly by placing samples on the instruments.The sample preparation of the acid-treated extractants 3-5 for 1 H NMR and ATR-FTIR is given as follows: 10 mL of approximately 1 mM of 3-5 was diluted in CHCl 3 and 10 mL of 0.1 M HCl were shaken at a speed of 300 rpm and temperature of 20 ± 1 • C for 30 min.After that, organic phase was separated by a separating funnel, and then the CHCl 3 from organic phase were evaporated to give acid-treated extractants 3-5.

Results and Discussion
First, to examine the extraction behavior of extractants 3-5, extractions from three single-metal solutions (Pt, Pd, and Rh) in HCl media were performed at various contact times.These experiments were conducted using 1.0 mM solutions of extractants 3-5 and 0.1 mM standard solutions of Pd(II), Pt(IV), and Rh(III) ions in 0.1 M HCl. Figure S1 displays the effect of the contact time (0-60 min) on the extractions of Pd(II), Pt(IV), and Rh(III) ions with extractants 3-5.The extractabilities of all three single metal solutions reached equilibrium within 5 min, as was previously reported for extractants 1 and 2 [34].The apparent extraction efficiency of 3-5 is similar to that of 1 and 2. Extractants 3-5 showed low efficiency on Rh(III) extraction, as well as 1 and 2, because the chrolo-Rh(III) complex is an inert extractive species and is very stable in HCl media.Taking into account our previous work [34], Pd(II) extraction of 1-5 has almost same efficiency and differential ranges of the E% were 98.3-99.5%.Thus, extending alkyl chain lengths has not much effect on Pd(II) extraction.On the other hand, different phenomena, as is compared with Pd(II) extraction using 1-5, were observed on Pt(IV) and Rh(III) extractions.In the case of Pt(IV) extraction, differential ranges of the E% by 1-5 were 90.8-99.5%,but there is no regularity depending on the increases of the alkyl chain lengths of 1-5.In contrast, the regularity between 3-5 on Rh(III) extraction was found, however, the same efficiency on Rh(III) extraction of 1 and 2 was proven.The extending of the alkyl chain lengths led to increase of the E%; 9.6% in 3, 11.1% in 4, and 16.1% in 5.It might appear that the increase of efficiency on Rh(III) extraction attributes increases of hydrophobicity and basicity by extending alkyl chains in the extractants.
Furthermore, a comparison of the structural effects observed for the macrocyclic extractants 3-5 and acyclic compound 6 was performed.Figure S2 shows the extractability values obtained for each single-component PGM solution using macrocyclic extractant 5 and acyclic compound 6.Although the concentration of 6 was four times higher than that of 5, the extractabilities of 5 and 6 determined for the three PGMs were significantly different: 66.9% for Pd(II), 80.4% for Pt(IV), and 0% for Rh(III).These results suggest that the macrocyclic extractant 5 and PGMs formed highly stable chemical species owing to assembly of metal recognition sites by modification of the amino moieties located along the annular structures of the calixarene upper rims.
Generally, PGM extractability of amine-based extractants is affected by varying acid concentration [38].Thus, to elucidate PGM extraction efficiency as a function of HCl concentration is also very important from an industrial point of view.Therefore, in order to examine the dependence of PGM extractability on the HCl concentration, extractions from single-metal PGM solutions with HCl concentrations ranging between 0.1 M and 8.0 M were conducted for 60 min using 1.0 mM solutions of extractants 3-5 and 0.1 mM standard Pd(II), Pt(IV), and Rh(III) solutions.Figure 1 shows the E% values obtained for the three PGMs at various HCl concentrations.After increasing the HCl concentration from 0.1 M to 8.0 M, the E% magnitudes determined for Pd(II) moderately decreased; E% = 99.3-46.8% in 3, E% = 99.3-49.7% in 4, and E% = 99.0-42.3% in 5, at 0.1 M to 8.0 M HCl.On the other hand, those obtained for Pt(IV) decreased more rapidly; E% = 98.8-26.7% in 3, E% = 99.4-27.2% in 4, and E% = 90.4-18.8% in 5, at 0.1 M to 8.0 M HCl.In the case of Rh(III) extraction, the obtained E% values were not very large (similar behavior was observed for 2).The maximum E% values for of Pd(II) and Pt(IV) were obtained at an HCl concentration of 0.1 M, and the minimum ones at 8.0 M. In contrast, the maximum and minimum E% values determined for Rh(III) ions depended on the extractant composition.As was mentioned earlier, the HCl concentration produced a distinct effect on the Pd(II) and Pt(IV) extractions by 3-5, and the optimum condition for the extraction of these ions corresponded to the HCl concentration is 0.1 M.
A comparison of the E% for extractants 3-5 and calix [4]arene-and thiacalix [4]arene-based extractants previously reported in the literature for Pd(II) and Pt(IV) extraction from aqueous solutions is presented in Table 1 [24,34,39,40].Extractants 3-5 possess an enhanced E%, which confirms its increased selectivity toward Pd(II) and Pt(IV).Although various calix [4]arene-and thiacalix [4]arene-based extractants have been synthesized in the literature, the high efficiency was either Pd(II) or Pt(IV) extraction.Based on a literature review of thia-and calix[n]arene-based extractants, extractants 3-5 are confirmed to be the effective extractants for both Pd(II) and Pt(IV) from PGM included solutions comprising industrial wastes such as spent automotive catalysts.In order to determine the stoichiometry of extractants 3-5 treated with Pd(II) and Pt(IV), the related log [D] versus log [E] plots were constructed.Similar to extractants 1 and 2, the slopes obtained for Pd(II) and Pt(IV) ions and extractants 3-5 were close to each other.As shown in Figure S3, the overall slopes determined for extractants 3-5 were close to 1. Hence, the Pd(II)-extractant chemical species obtained after Pd(II) extraction likely exhibited a Pd(II)/extractant stoichiometry of 1:1.The three slope values obtained for the first five points located in the log [E] range between −5.5 and −4.0, were equal to 0.64 for 3, and 0.43 for 4 and 5.The other slopes determined for the last four points (in the log [E] range from −4.0 to −3.0) were 2.03 for 3, 2.06 for 4, and 2.00 for 5. From the constructed log-log plots, the existence of two different species with Pd(II)/extractant ratios of 2:1 and 1:2 in the organic phase can be suggested.Similar behavior was observed for the Pt(IV) extraction with extractants 3-5.In Figure S4, the slopes of the curves plotted for Pt(IV) species and three extractants were also close to 1; therefore, the produced species exhibited a Pt(IV)/extractant stoichiometric ratio of 1:1 in the organic phase.Furthermore, these curves can also be divided into two sections consisting of the first four and last five data points.The slopes of the first sections corresponding to the log [E] range from −5.5 to −4.0 were 0.64 for 3, 0.60 for 4, and 0.55 for 5.The slopes of the second sections lying in the log [E] range from −4.0 to −3.0 were 1.93 for 3, 1.87 for 4, and 2.21 for 5. Using the same logic, the existence of species with Pt(IV)/extractant stoichiometric ratios of 2:1 and a 1:2 in the organic phases can be assumed.
Further, in order to elucidate Pd(II) and Pt(IV) extraction mechanisms, native 3-5 and 0.1 M HCl-treated 3-5 were examined by means of 1 H NMR and ATR-FTIR spectroscopies.HCl extraction of 3-5 (10 mM) in CHCl 3 (10 mL) was carried out with a 0.1 M HCl (10 mL) for 30 min. 1 H NMR spectra of the native 3-5 and the acid-treated 3-5 were obtained as given in Figures S5-S7.Protons of methylene groups around nitrogen of amino moieties in 3-5 showed ~0.8 ppm downfield shift after being treated with 0.1 M HCl [41].On the other, it was very difficult to exactly obtain integrated values of each chemical shift of acid-treated 3-5 because the chemical shifts were broad.However, the shift supports that 3-5 extracted HCl after treating with 0.1 M HCl by comparing the NMR spectra of the native and the acid-treated 3-5.Additionally, ATR-FTIR spectra of the native and the treated 3-5 also support the HCl extraction by amino groups of 3-5 and were given in Figure S8.In the case of 0.1 M HCl-treated 3-5, new peaks appeared around 2780-2730 cm −1 , which might be attributed to a unique N-H stretching deriving from quaternary ammonium of extractants 3-5 [33].
From the log-log plots obtained for the Pd(II) and Pt(IV) extractions by 3-5 and the spectral analysis of 1 H NMR and ATR-FTIR after HCl extraction of 3-5, it can be concluded that the probable Pd(II) and Pt(IV) extraction mechanisms involve capturing one anionic Pd(II) or Pt(IV) chloro-complex by the two quaternary ammonium groups of cationic calix [4]arenes, according to Equations ( 5)- (7).The calixarene amino moieties of extractants 3-5 partially extract HCl molecules during the contact with the HCl media followed by exchanging Cl − ions with [PdCl 4 ] 2− or [PtCl 6 ] 2− complexes through the formation of an ion pair.

Extractant + 2HCl
USV Symbol Macro(s) Description In this work, acyclic compound 6 was also used for the PGM extraction from the leach liquors of automotive catalysts.Pd(II), Pt(IV), and Rh(III) extractions were performed using the leach liquors (10 mL) and 1.0 mM of extractants 3-5 (10 mL) over a period of 60 min.In the case of acyclic compound 6, its 4.0 mM solution was used for extraction because each calix [4]arene-based extractant contained four di-n-alkylaminomethyl-modified aromatic units.Figure 2 displays the E% values obtained for the nine metal ions present in the liquors.The E% values of extractants 3-5 measured after the extraction of Pd(II) and Pt(IV) ions were 94.2% and 91.7% for 3, 93.0% and 94.0% for 4, and 97.7% and 92.5% for 5, respectively.The E% magnitudes determined for the seven other metal ions present in the leach liquors of automotive catalysts were below 6%.The results of the Pd(II) and Pt(IV) extractions performed using extractants 3-5 were similar to those obtained for the PGM extractions from the leach liquors of automotive catalysts by extractants 1 and 2.Although acyclic compound 6 exhibited high recognition abilities for Pd(II) and Pt(IV) ions, the E% values of Pd(II) and Pt(IV) species were lower than the numbers obtained for macrocyclic extractants 3-5 (56.2% for Pd(II) and 63.0% for Pt(IV)).The observed difference can be attributed to the presence of regularly arranged amino moieties along the annular structures of the calixarene upper rims.In other words, the macrocyclic structure can form highly stable metal species by assembling the functional groups characterized by high metal recognition ability.Meanwhile, the macrocyclic extractants 1-5 obtained by the modification of alkyl chains (n-propyl to n-octyl) have no remarkable difference for the E% on Pd(II) and Pt(IV) extractions.Hence, it is obvious that hydrophobicity of an extractant is not responsible for the extraction efficiency.
macrocyclic structure can form highly stable metal species by assembling the functional groups characterized by high metal recognition ability.Meanwhile, the macrocyclic extractants 1-5 obtained by the modification of alkyl chains (n-propyl to n-octyl) have no remarkable difference for the E% on Pd(II) and Pt(IV) extractions.Hence, it is obvious that hydrophobicity of an extractant is not responsible for the extraction efficiency.
The stripping of the extracted Pd(II) and Pt(IV) ions from the organic phase was examined utilizing a 0.1 M thiourea/1.0M HCl mixed solution (10 mL) as a stripping reagent [23,26,30] and reusability of the used extractants 3-5 on Pd(II) and Pt(IV) extractions from the leach liquors of automotive catalysts.A total of five Pd(II) and Pt(IV) stripping/extraction cycles were performed.As shown in Figure 3, the S% values obtained for Pd(II) and Pt(IV) ions after the first stripping cycle were 99.9% and 96.3% for 3, 96.6% and 96.7% for 4, and 95.3% and 95.4% for 5, respectively.After this step, each organic phase containing extractants 3-5 (10 mL) was washed with distilled water (20 mL) to remove the unreacted thiourea species that had been transferred during the stripping process, and the extraction procedure was repeated.The E% magnitudes determined for the reused extractants and Pd(II) and Pt(IV) ions were 93.2% and 93.0% for 3, 90.1% and 90.7% for 4, and 97.5% and 92.5% for 5, respectively.After five extraction/stripping cycles, the results similar to those obtained after the first cycle were produced.Additionally, the accumulation of the other metals in the solvent after the five extraction/stripping cycles was within ~5% of each.The reusability check indicated that the reused extractants 3-5 could be utilized continuously for the separation of Pd(II) and Pt(IV) ions from PGM-containing secondary resources (including the leach liquors of automotive catalysts).The stripping of the extracted Pd(II) and Pt(IV) ions from the organic phase was examined utilizing a 0.1 M thiourea/1.0M HCl mixed solution (10 mL) as a stripping reagent [23,26,30] and reusability of the used extractants 3-5 on Pd(II) and Pt(IV) extractions from the leach liquors of automotive catalysts.A total of five Pd(II) and Pt(IV) stripping/extraction cycles were performed.As shown in Figure 3, the S% values obtained for Pd(II) and Pt(IV) ions after the first stripping cycle were 99.9% and 96.3% for 3, 96.6% and 96.7% for 4, and 95.3% and 95.4% for 5, respectively.After this step, each organic phase containing extractants 3-5 (10 mL) was washed with distilled water (20 mL) to remove the unreacted thiourea species that had been transferred during the stripping process, and the extraction procedure was repeated.The E% magnitudes determined for the reused extractants and Pd(II) and Pt(IV) ions were 93.2% and 93.0% for 3, 90.1% and 90.7% for 4, and 97.5% and 92.5% for 5, respectively.After five extraction/stripping cycles, the results similar to those obtained after the first cycle were produced.Additionally, the accumulation of the other metals in the solvent after the five extraction/stripping cycles was within ~5% of each.The reusability check indicated that the reused extractants 3-5 could be utilized continuously for the separation of Pd(II) and Pt(IV) ions from PGM-containing secondary resources (including the leach liquors of automotive catalysts).