Article pubs.acs.org/JPCB
Selective Crystallization of Phosphoester Coordination Polymer for the Separation of Neodymium and Dysprosium: A Thermodynamic Approach Yuiko Tasaki-Handa,*,† Yukie Abe, Kenta Ooi, Hirokazu Narita, Mikiya Tanaka, and Akihiro Wakisaka Environmental Management Research Institute, National Institute of Advanced Science and Technology (AIST), 16-1 Onogawa, Tsukuba 305-8569, Ibaraki, Japan S Supporting Information *
ABSTRACT: Thermodynamics of the formation of coordination polymers (CPs) or metal−organic frameworks (MOFs) has not been focused on, whereas many CPs or MOFs have been synthesized in a solution. With a view of separating Nd3+ and Dy3+ in an aqueous solution, we demonstrate that crystallization of the CPs of Nd3+ and Dy3+ based on dibutyl phosphoric acid (Hdbp) can be thermodynamically described; crystallization yields of [Ln(dbp)3] (Ln = Nd or Dy) complex are predicted well using a simple calculation, which takes the apparent solubility products (Kasp) for [Ln(dbp)3] and the acid dissociation constant of Hdbp into account. The Kasp values of [Nd(dbp)3] and [Dy(dbp)3] are experimentally determined to be (1.3 ± 0.1) × 10−14 and (2.9 ± 0.4) × 10−18 M4, respectively, at 20 °C. The ratio of these Kasp values, that is, ca. 4500, is significantly larger than the ratio of the solubility products for inorganic salts of Nd3+ and Dy3+. Therefore, Nd3+ and Dy3+ are selectively crystallized in an aqueous solution via the formation of CPs. Under optimized conditions, Dy3+ crystallization is preferable, whereas Nd3+ remains in the solution phase, where the ratio of the Dy molar content to the total metal content (i.e., Nd + Dy) in the crystal is higher than 0.9. The use of acids, such as HCl or HNO3, has no practical impact on the separation in an aqueous solution.
■
INTRODUCTION Coordination polymers (CPs) or metal−organic frameworks (MOFs) are types of extended crystals that are formed by metal ions and multidentate organic ligands; in recent years, such crystals have received much attention owing to their applications in gas separation, gas storage, catalysis, and sensing.1−4 Despite the binding preference of the ligand to the metal ions being involved in the crystallization of CPs or MOFs, most studies have not emphasized this fact; a lot of effort has been devoted to obtaining single crystals with high-porosity frameworks for guest molecule storage.5−7 Only a few studies have focused on the different crystallization yields of MOFs. Wang et al.8 synthesized MOFs comprising 2−10 different transition metals from their mixtures, and reported that the ratios of the metal ions in MOFs differ from those of the original solutions. Zhao et al.9 reported fractional crystallization of lanthanide ions as a camphorate MOF, in which the ratios of, for example, Y in Y−Yb and Y−Gd mixed systems were 36−37 and 67−70%, respectively. These studies indicate that the resulting molar ratio of metals in a crystal can vary depending on the coexisting metal ions. The selective precipitation or crystallization of metal ions, for which inorganic anions are conventionally used, has often suffered from low selectivity, and, as a result, has been replaced by ion-exchange or liquid−liquid extraction methods. To enhance the precipitation or crystallization selectivity, it is principally effective to use an anion with high selectivity against the target © 2016 American Chemical Society
metal ions. We conceived the idea that the formation of CPs or MOFs may enable the selective crystallization of metal ions because acidic organic ligands often display a higher selectivity for metal ions over conventional inorganic anions: indeed, they play a key role in the selectivity of ion-exchange and liquid− liquid extraction systems. If the selective crystallization of metal ions was feasible, it would mean that (1) a wide variety of organic ligands could be applied to form CPs or MOFs and (2) organic solvents would not necessarily be required for this process, that is, it would provide a more environmentally friendly method than conventional liquid−liquid extraction. From a separation point of view, selective and effective crystallization most importantly requires (1) a ligand that highly and selectively coordinates with target metal ions, (2) environmentally friendly solvent, for example, water, and (3) optimization of the solution composition. Point 1 is basically in agreement with the preference for mononuclear complexing. Point 2 may be difficult to achieve because CPs or MOFs are generally synthesized in organic solvents, such as N,Ndimethylformamide and dimethylacetamide. However, any solvent, including water, can technically be employed if the target ligand is soluble in it. Point 3 concerns the analysis of the Received: September 19, 2016 Revised: November 14, 2016 Published: November 14, 2016 12730
DOI: 10.1021/acs.jpcb.6b09450 J. Phys. Chem. B 2016, 120, 12730−12735
Article
The Journal of Physical Chemistry B
and N in the precipitates were determined using an organic elemental analyzer (2400 II, PerkinElmer). The final c(H+) values of the filtrates were determined from the electric potential using a pH meter (pH meter 744, Metrohm) when ci(H+) was less than 0.1 M. For the solutions with ci(H+) greater than 0.2 M, c(H+) was considered to be the same as ci(H+).
reaction based on the solution equilibria. Such an analysis would be easier if water was used as a solvent. In the previous study, we studied the selective crystallization of Nd and Dy as a CP based on di-(2-ethylhexyl) phosphoric acid (Hdehp; PO(OH)(CHCH(C2H5)(CH2)3CH3)2).10,11 The mutual separation of Nd and Dy remains a challenging issue because of their similar chemical properties, in spite of an increasing demand for them as the major component of Nd magnets.12−15 Therefore, we focused on the formation of a CP based on phosphoester because phosphoesters are known to be the most effective substances with regard to separating lanthanide ions;13,16,17 additionally, they are bidentate ligands and form CPs.18−20 Under optimized conditions, a CP for which the molar ratio of Dy to the total metal (i.e., Nd + Dy) was over 97% was obtained. However, a major defect of this system was that an ethanol−water mixture was used to dissolve Hdehp. The drawbacks of this are not only the potential environmental impact, but also the difficulty of the analysis of the solution equilibria. In the present study, we used dibutyl phosphoric acid (PO(OH)(CH2)3CH3; Hdbp) in anticipation of its higher solubility in water because Hdbp has shorter alkyl chains than Hdehp. Hdbp is a phosphoester and should have a high selectivity for intralanthanide ions. The solution equilibria were analyzed by characterizing the crystalline and solution phases, which enables to optimize the separation condition. Furthermore, the effects of HCl and HNO3 were compared because HNO3 offered better separation in the previous study.11
■
RESULTS AND DISCUSSION Determination of Reaction Equilibria. HCl Media. Figure 1 shows the crystallization yields of Nd and Dy in HCl media as a
Figure 1. Change in crystallization yields of Nd and Dy with ci(H+).
function of ci(H+); the yields were calculated based on the change in the concentration of Nd3+ and Dy3+ in the solution phases. The crystallization yields are around 95% under the given conditions and decrease with an increase in ci(H+). This indicates that the acid dissociation of Hdbp is involved in the crystallization.
■
EXPERIMENTAL SECTION Formation of both the Nd and Dy CPs. Hdbp was converted to a Na salt to enhance its solubility: a Nadbp (∼0.3 M) aqueous solution was prepared by mixing NaOH and Hdbp in equimolar amounts. Both the Nd and Dy CPs were formed by mixing Nadbp, LnX3 (Ln = Nd or Dy, X = Cl or NO−3 ), HX, and NaX aqueous solutions at 20 °C. The initial concentrations of dbp− (ci(P)) and LnX3 (ci(Ln3+)) were 0.06 ± 0.002 and 0.02 ± 0.003 M, respectively. The initial concentration of H+, ci(H+), was varied between 0.05 and 0.09 M and between 0.2 and 0.5 M for Nd3+ and Dy3+, respectively, by adding HX. NaX was added to adjust the ionic strength of the solutions to 0.6 ± 0.02. The solutions were filtered with a polytetrafluoroethylene membrane (0.2 μm) 14 days after their preparation, which was sufficient to achieve equilibrium (the time required was 7 days, see Figure S1). NdCl3, DyCl3, Nd(NO3)3, and Dy(NO3)3 were purchased as hexahydrates (>99.7%) from Wako Pure Chemical Industries, Ltd., and Hdbp (>97%) was purchased from Sigma-Aldrich. Selective Crystallization of Dy from Nd−Dy Mixture. The procedure was basically the same as described above. The ci(Nd3+) and ci(Dy3+) values were both 0.021 ± 0.003 M, and ci(P) and ci(H+) were varied between 0.060 and 0.064 M and between 0.018 and 0.039 M, respectively. The solutions were filtered 7 days after preparation. Characterization of the CPs and Solution Phases. The filtrates (3−5 mL depending on the concentration of the species) were decomposed with a mixture of HNO3 (60%, 3 mL) and H2O2 (30%, 0.3 mL) by microwave irradiation. The concentrations of P, Nd, and Dy were then determined by inductively coupled plasma-optical emission spectroscopy (ICP-OES; ICPE-9000; Shimadzu). The precipitates were characterized by scanning electron microscopy (SEM; JSM-6010LA, JEOL) and powder X-ray diffraction (PXRD; D2PHASER, Bruker). The contents of C, H,
Hdbp ⇌ H+ + dbp−
(1) 3+
The UV−vis spectra revealed that Ln in the solution phase is present in a hydrated form (Figure S2). Figure 2 shows the SEM micrographs of the Nd and Dy CPs. Regardless of c(H+), the Nd precipitates are rodlike, 3−6 μm thick, and 10−70 μm long; whereas, the Dy precipitates are fibrous, ∼1 μm thick, and over 30 μm long. The contents of C and H in [Ln(dbp)3], shown in Table 1, reveal that the chemical formula of the crystals is [Ln(dbp)3] (Ln = Nd or Dy). The attenuated total reflectance Fourier transform infrared (ATRFTIR) spectra (Figure S3) confirmed the absence of water molecules. Stoichiometry requires that Ln3+ in [Ln(dbp)3] should be coordinated with six oxygen atoms of dbp−.18 Figure 3 shows the PXRD patterns, revealing that [Nd(dbp)3] and [Dy(dbp)3] are crystalline. Dibutyl chains may be well stacked in [Ln(dbp)3] compared with the 2-ethylhexyl chains in [Ln(dehp)3]. Despite the obvious differences in their morphologies (Figure 2), the PXRD patterns of [Nd(dbp)3] and [Dy(dbp)3] are substantially the same; in both the patterns, the strongest peak is located at 2θ ∼ 7° and four or five moderate peaks are found at 2θ ∼ 12, 14, 18, and 20° (Table S1). In addition, the PXRD patterns of both are very similar to that of [Ln(dehp)3] reported in a previous study, on the basis of which we predict that the structure consists of hexagonally packed linear chains in which the lanthanide ions are connected by O− P−O bridges, as shown in Figure 4.18 Although the unit cell cannot be indexed owing to a few peaks, the intense peak at 2θ ∼ 7° (d = 12.7 Å) is likely to attribute to the distance between the chains. The fibrous and rodlike morphologies of [Ln(dbp)3] (Figure 2) may be explained by growth in the main-chain 12731
DOI: 10.1021/acs.jpcb.6b09450 J. Phys. Chem. B 2016, 120, 12730−12735
Article
The Journal of Physical Chemistry B
Figure 3. PXRD patterns of [Nd(dbp)3] (A) and [Dy(dbp)3] (B).
To confirm whether NO−3 affects separation in the present study using Hdbp, both the crystalline and solution phases obtained in HNO3 were characterized. As shown in Table 2, the C and H contents of the crystals are almost the same as those for [Ln(dbp)3] obtained in HCl media, that is, nitrogen was not detected. Therefore, the chemical composition of the crystals is [Nd(dbp)3] and [Dy(dbp)3], in common with HCl systems. This is supported by the fact that there were no differences in SEM micrographs of the crystals from HCl and HNO3 systems (Figure S4). This finding is not consistent with that of the previous study using Hdehp. A difference in the alkyl chain length, solvent (i.e., ethanol−water or water), or crystallinity will plausibly affect the association of NO−3 with Nd3+. Figure 5 compares the absorption spectra of the filtrates in HCl and HNO3 media. The band around 575 nm corresponds to the 4I9/2 →4 G5/2 and 2G7/2 transitions of Nd3+, which are sensitive to the coordination environment.21 The band at 575 nm obtained in the HCl system is attributed to hydrated Nd3+, and that around 580 nm to [Nd(NO3)]2+. The ratio of the absorbance at 580 nm to that at 575 nm (Rabs) is 0.78, which is the same as that measured at 25 °C by Rao and Tian.22 Therefore, we consider that Kc = c(Nd(NO3)2+)/{c(Nd3+)· c(NO−3 )} ≈ 0.64, which is the value reported by Rao and Tian. Because the effect of NO−3 was very small even when Rabs was about 1.2 for the Hdehp system, Rabs = 0.64 may be too small to affect the solution equilibria. We found that the presence of ethanol may enhance the association between Nd3+ and NO−3 ; Rabs for an 80:20 (vol %) ethanol−water mixture, including Nd3+, dbp−, and NO−3 , was 1.1 (Figure S5). We propose that it is not the structure of the ligand but the presence of ethanol that significantly affects the association of NO−3 with Nd3+, which enhances separation. However, we have not investigated the effect of ethanol in more detail because the aim of this study was to separate Nd3+ and Dy3+ in an aqueous solution. We conclude that the reactions given by eqs 1 and 2 should be taken into account in aqueous media, 2 irrespective of the acid type. Estimation of Selectivity. The solubility product, Ksp = a(Ln3+)·a(dbp−)3 (a: activity), is a key parameter in separation
Figure 2. SEM micrographs of [Nd(dbp)3] (A, scale bar = 50 μm) and [Dy(dbp)3] (B, scale bar = 20 μm).
Table 1. C and H Contents in [Ln(dbp)3] Obtained in HCl Mediaa,b [Nd(dbp)3] +
[Dy(dbp)3] +
c(H ) (M)
C (%)
H (%)
c(H ) (M)
C (%)
H (%)
0.057 0.089
37.2 37.3
7.20 7.18
0.21 0.50
37.2 37.2
7.20 7.20
a Calcd for [Nd(dbp)3]: C, 37.35; H, 7.052. bCalcd for [Dy(dbp)3]: C, 36.48; H, 6.889.
direction. In addition, the difference in the crystal thickness of [Nd(dbp)3] and [Dy(dbp)3] may be affected by aggregation of the chains. From these results, the following reaction should be taken into account when analyzing the crystallization process + Ln 3hyd + 3dbp− ⇌ [Ln(dbp)3 ]
(2)
HNO3 Media. In our previous study using Hdehp in an ethanol−water mixture,10,11 the presence of NO−3 enhanced the separation performance. We concluded that this was caused by a large difference in the association of NO−3 with Nd3+ and Dy3+; theoretical and experimental studies have revealed that NO−3 shows a stronger affinity for Nd3+ than Dy3+.21 Although a detailed analysis was impossible because we used an ethanol− water mixture, the formation of [Nd(NO3)]2+ in the solution phase may inhibit crystallization of Nd, whereas the replacement of dehp− with NO−3 in the crystals may enhance the solubility of Nd crystals, resulting in a better separation of Nd3+ compared with Dy3+ in HNO3 than that in HCl media. 12732
DOI: 10.1021/acs.jpcb.6b09450 J. Phys. Chem. B 2016, 120, 12730−12735
Article
The Journal of Physical Chemistry B
Figure 4. Schematics of the structure of [Ln(dbp)3].18
Equations 3 and 4 yield
Table 2. C, H, and N Contents in [Ln(dbp)3] Obtained in HNO3 Media [Nd(dbp)3] +
⎛ 1 + α ⎞3 c(Ln 3 +) = K spa(Ln) ·⎜ t ⎟ ⎝ c (P) ⎠
[Dy(dbp)3] +
c(H ) (M)
C (%)
H (%)
N (%)
c(H ) (M)
C (%)
H (%)
N (%)
0.034 0.062
37.5 37.3
7.32 7.24
0.00 0.00
0.21 0.51
36.3 36.5
7.05 7.06
0.00 0.00
(5) −
where α = c(H and c (P) = c(Hdbp) + c(dbp ). The Kaa value at 20 °C was estimated to be (4.0 ± 0.5) × 10−3 M−1 by titration (Figure S5), and ct(P) was determined by ICP-OES after acid digestion of the solution. According to eq 5, the plots of c(Ln3+) versus {(1 + α)/ct(P)}3 (Figure 6) give Kasp(Nd) = (1.3 ± +
)/Kaa
t
Figure 6. Change in c(Ln3+) with {(1 + α)/ct(P)}3.
0.1) × 10−14 M4 and Kasp(Dy) = (2.9 ± 0.4) × 10−18 M4 at 20 °C. The reason that Kasp(Dy) is smaller than Kasp(Nd) is likely because of the stronger electrostatic interaction of Dy3+ with dbp−, caused by its smaller ionic radius, compared with Nd3+. It should be noted that the Kasp(Nd)/Kasp(Dy) ratio of ca. 4500 (3600−5600, if the deviation of Kasp is taken into account) is remarkably large as far as Nd3+ and Dy3+ are concerned. For example, the ratios of the Kasp values for phosphorous salt,23 oxalate,24 and hydroxide25 are estimated to be 0.1, 1.5, and 500, respectively. The high selectivity of the phosphoester, Hdbp, against intralanthanide ions is likely to be reflected in the result, although stability coefficients of Hdbp complex with Nd3+ and Dy3+ have not been reported. The formation of [Ln(dbp)3] has the potential to crystallize Dy3+ selectively over Nd3+ in an aqueous solution. Fractional Crystallization. The c(Nd3+) and c(Dy3+) values in their mixture systems are calculated using Kasp(Nd), Kasp(Dy), and Kaa values. Considering eq 3 for both Nd3+ and Dy3+ and the mass balance of phosphor, the following equation related to X = c(Nd3+) + c(Dy3+) can be derived as
Figure 5. Absorption spectra of Nd3+ in HCl and HNO3 media. The broken blue lines are the differential spectra.
by crystallization. In this study, the apparent solubility product (Kasp) when the ionic strength is 0.6 is defined as follows K spa(Ln) =
K sp(Ln 3 +) y(Ln 3 +)·y(dbp−)3
= c(Ln 3 +)·c(dbp−)3
(3)
where c is the concentration in the solution and y is the activity coefficient. The apparent acid dissociation constant for Hdbp (cf. eq 1) is similarly defined as K aa =
c(H+) ·c(dbp−) c(Hdbp)
(4) 12733
DOI: 10.1021/acs.jpcb.6b09450 J. Phys. Chem. B 2016, 120, 12730−12735
Article
The Journal of Physical Chemistry B
ci(P) deviates from this point, Nd3+ is unfavorably crystallized or Dy3+ remains in the solution phase. The experimental results of c(Nd3+) and c(Dy3+) in HCl and HNO3 are represented by triangles and circles, respectively, in Figure 7A,B. Both c(Nd3+) and c(Dy3+) are almost the same for either HCl or HNO3 media. This indicates that the presence of NO−3 does not affect the separation performance under the conditions in the present study, possibly due to the small Kc value of ca. 0.64. The experimental results are in substantial agreement with the calculated ones. However, the actual c(Nd3+) values tend to be slightly smaller than the calculated values, whereas the actual c(Dy3+) values are larger than the calculated ones. Inclusion of Nd3+ into [Dy(dbp)3] may be involved, as the structures of [Nd(dbp)3] and [Dy(dbp)3] are very similar (Figure 3). Because the incorporation of Nd3+ into [Dy(dbp)3] would be a kinetic phenomenon, we studied the change in c(Nd3+) and c(Dy3+) as a function of reaction time. Figure 8 shows the change in c(Nd3+)
1/3 ⎫ ⎧ ⎪ ⎪ 1 + c i(H+)/K aa X ⎬ ⎨ = a ⎪ ⎪ c i(P) − 3{c i(Nd3 +) + c i(Dy 3 +) − X } ⎩ K sp(Nd) ⎭
(6)
Kasp(Nd)
+ Kasp(Dy) ≈ Kasp(Nd) and Using Kasp(Nd)/Kasp(Dy) = 3+
where the approximations c(H+) ≈ ci(H+) were employed. c(Nd3+)/c(Dy3+), both c(Nd3+) and c(Dy ) can be calculated when ci(H+), ci(P), ci(Nd3+), and ci(Dy3+) are given. In the present study, both ci(Nd3+) and ci(Dy3+) are set to 0.02 M, and the effects of ci(H+) and ci(P) on the separation were studied. The colored regions in Figure 7A,B represent the
Figure 8. Time course of c(Nd3+) and c(Dy3+) in HCl and HNO3.
and c(Dy3+) with time in HCl and HNO3 media. The c(Nd3+) value increases from 0.018 to 0.020 M, and c(Dy3+) decreases from 0.025 to 0.020 M. At first, the crystallization would be kinetically governed, before the crystals are purified by dissolution−recrystallization to achieve equilibrium. Despite the slight disagreement between the calculation and the experiment, the simple calculation using Kasp and Kaa values allows us to predict the crystallization reaction. To the best of our knowledge, this is the first study to analyze the solution equilibria of CP formation. Nd3+ and Dy3+ are significantly separated, especially at ci(H+) = 0.18 M and ci(P) = 0.0635 M, where the separation factor is above 300 and the molar ratio of Dy to the total metal content in the crystal was higher than 0.9. For example, the separation factor for the Dy/Nd fractional precipitation system using oxalate24 can be calculated to be about 2. In addition, selective crystallization of camphorate MOF offered separation factors smaller than 10, even for Yb/Gd separation, for which their ratio of ionic radii is greater than that of Nd and Dy.9 Although the separation efficiency may vary with the experimental conditions, we believe that the separation efficiency in this study is superior to that of traditional fractional crystallization systems and those reported in recent studies.
Figure 7. Change in c(Nd3+) and c(Dy3+) with ci(P) (A, when ci(H+) = 0.2 M) and ci(H+) (B, when ci(P) = 0.0635 M).
calculated c(Nd3+) and c(Dy3+) values as a function of ci(P) and ci(H+), respectively, in consideration of the errors of the Kasp and Kaa values. The deviation of the calculated results seems small compared with the relatively large deviations of the Kasp and Kaa values. To crystallize Dy3+ selectively, c(Dy3+) ≈ 0 and c(Nd3+) ≈ ci(Nd3+) = 0.02 M are desired. According to the calculation, Nd3+ tends to remain in the solution phase with ci(P) in the range from 0.058 to 0.065 M and ci(H+) from 0.1 to 0.5 M, whereas c(Dy3+) is much smaller than c(Nd3+). The optimal condition is found to be ci(H+) ≈ 0.18 M and ci(P) ≈ 0.0635 M. When either ci(H+) or 12734
DOI: 10.1021/acs.jpcb.6b09450 J. Phys. Chem. B 2016, 120, 12730−12735
Article
The Journal of Physical Chemistry B
■
Camphorate Metal-Organic Frameworks for Lanthanide Separation. J. Am. Chem. Soc. 2014, 136, 12572−12575. (10) Tasaki-Handa, Y.; Abe, Y.; Ooi, K.; Narita, H.; Tanaka, M.; Wakisaka, A. Environmentally Friendly Separation of Dysprosium and Neodymium by Fractional Precipitation of Coordination Polymers. RSC Adv. 2014, 4, 20496−20498. (11) Tasaki-Handa, Y.; Abe, Y.; Ooi, K.; Narita, H.; Tanaka, M.; Wakisaka, A. Separation of Neodymium and Dysprosium by Forming Coordination Polymers. Sep. Purif. Technol. 2016, 157, 162−168. (12) Narita, H.; Tanaka, M. Separation of Rare Earth Elements from Base Metals in Concentrated HNO3, H2SO4and HCl Solutions with Diglycolamide. Solvent Extr. Res. Dev., Jpn. 2013, 20, 115−121. (13) Kikuchi, Y.; Matsumiya, M.; Kawakami, S. Extraction of Rare Earth Ions from Nd-Fe-B Magnet Wastes with TBP in Tricaprylmethylammonium Nitrate. Solvent Extr. Res. Dev., Jpn. 2014, 21, 137−145. (14) Ishioka, K.; Matsumiya, M.; Ishii, M.; Kawakami, S. Development of Energy-Saving Recycling Process for Rare Earth Metals from Voice Coil Motor by Wet Separation and Electrodeposition using MetallicTFSA Melts. Hydrometallurgy 2014, 144−145, 186−194. (15) Uda, T. Recovery of Rare Earths from Magnet Sludge by FeCl2. Mater. Trans. 2002, 43, 55−62. (16) Starý, J. Separation of Transplutonium Elements. Talanta 1966, 13, 421−437. (17) Yoshizuka, K.; Kosaka, H.; Shinohara, T.; Ohto, K.; Inoue, K. Structural Effect of Phosphoric Esters Having Bulky Substituents on the Extraction of Rare Earth Elements. Bull. Chem. Soc. Jpn. 1996, 69, 589− 596. (18) Ellis, R. J.; Demars, T.; Liu, G.; Niklas, J.; Poluektov, O. G.; Shkrob, I. A. In the Bottlebrush Garden: The Structural Aspects of Coordination Polymer Phases Formed in Lanthanide Extraction with Alkyl Phosphoric Acids. J. Phys. Chem. B 2015, 119, 11910−11927. (19) Peppard, D.; Mason, G.; Maier, J.; Driscoll, W. Fractional Extraction of the Lanthanides as Their Di-alkyl Orthophosphates. J. Inorg. Nucl. Chem. 1957, 4, 334−343. (20) Saleh, M. I.; Bari, M.; Saad, B. Solvent Extraction of Lanthanum(III) from Acidic Nitrate-Acetato Medium by Cyanex 272 in Toluene. Hydrometallurgy 2002, 63, 75−84. (21) Bonal, C.; Morel, J.-P.; Morel-Desrosiers, N. Interactions between Lanthanide Cations and Nitrate Anions in Water Part 2 Microcalorimetric Determination of the Gibbs Energies, Enthalpies and Entropies of Complexation of Y3+ and Trivalent Lanthanide Cations. J. Chem. Soc., Faraday Trans. 1998, 94, 1431−1436. (22) Rao, L.; Tian, G. Complexation of Lanthanides with Nitrate at Variable Temperatures: Thermodynamics and Coordination Modes. Inorg. Chem. 2009, 48, 964−970. (23) Liu, X.; Byrne, R. H. Rare Earth and Yttrium Phosphate Solubilities in Aqueous Solution. Geochim. Cosmochim. Acta 1997, 61, 1625−1633. (24) Weaver, B. Fractional Separation of Rare Earths by Oxalate Precipitation from Homogeneous Solution. Anal. Chem. 1954, 26, 479− 480. (25) Millero, F. J. Stability Constants for the Formation of Rare Earth Inorganic Complexes as a Function of Ionic Strength. Geochim. Cosmochim. Acta 1992, 56, 3123−3132.
CONCLUSIONS Nd and Dy3+ were successfully separated in an acidic aqueous solution via the fractional precipitation of Hdbp CPs. The calculation based on the solution equilibrium predicted the experimental results well. We believe that this method has a potential to contribute to the development of separation and correction of Nd and Dy from manufacturing scraps and used magnets. Furthermore, this study suggests that selective crystallization of CPs or MOFs can enhance the separation performance of metal ions in a wide variety of such processes.
■
3+
ASSOCIATED CONTENT
S Supporting Information *
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jpcb.6b09450. Time course of precipitation; UV−vis spectra of the filtrates; ATR-FTIR spectra of [Ln(dbp)3]; SEM images of [Ln(dbp)3]; titration of Hdbp; peak list of PXRD (PDF)
■
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Tel: +81-(0)48 858 3524. ORCID
Yuiko Tasaki-Handa: 0000-0002-4777-7541 Present Address †
Graduate School of Science and Engineering, Saitama University, 255 Shimo-okubo, Sakura-ku, Saitama-shi 3388570, Saitama, Japan (Y.T-H.).
Notes
The authors declare no competing financial interest.
■
REFERENCES
(1) Li, J.-R.; Kuppler, R. J.; Zhou, H.-C. Selective Gas Adsorption and Separation in Metal-Organic Frameworks. Chem. Soc. Rev. 2009, 38, 1477−1504. (2) Lee, J.; Farha, O. K.; Roberts, J.; Scheidt, K. A.; Nguyen, S. T.; Hupp, J. T. Metal-Organic Framework Materials as Catalysts. Chem. Soc. Rev. 2009, 38, 1450−1459. (3) Allendorf, M. D.; Bauer, C. A.; Bhakta, R. K.; Houk, R. J. T. Luminescent Metal-Organic Frameworks. Chem. Soc. Rev. 2009, 38, 1330−1352. (4) Batten, S. R.; Neville, S. M.; Turner, D. R. Coordination Polymers; Royal Society of Chemistry: Cambridge, UK, 2009; Chapter 10, pp 313−344. (5) Khan, N. A.; Jhung, S. H. Synthesis of Metal-Organic Frameworks (MOFs) with Microwave or Ultrasound: Rapid Reaction, PhaseSelectivity, and Size Reduction. Coord. Chem. Rev. 2015, 285, 11−23. (6) Feng, D.; Liu, T.-F.; Su, J.; Bosch, M.; Wei, Z.; Wan, W.; Yuan, D.; Chen, Y.-P.; Wang, X.; Wang, K. l.; et al. Stable Metal-Organic Frameworks Containing Single Molecule Traps for Enzyme Encapsulation. Nat. Commun. 2015, 6, No. 5979. (7) Feng, D.; Wang, K.; Su, J.; Liu, T.-F.; Park, J.; Wei, Z.; Bosch, M.; Yakovenko, A.; Zou, X.; Zhou, H.-C. A Highly Stable Zeotype Mesoporous Zirconium Metal Organic Framework with Ultralarge Pores. Angew. Chem., Int. Ed. 2015, 54, 149−154. (8) Wang, L. J.; Deng, H.; Furukawa, H.; Gandara, F.; Cordova, E. K.; Peri, D.; Yaghi, M. O. Synthesis and Characterization of Metal-Organic Framework-74 Containing 2, 4, 6, 8, and 10 Different Metals. Inorg. Chem. 2014, 53, 5881−5883. (9) Xiang, Z.; Matthew, W.; Chengyu, M.; Thuong, X.-T.; Jian, Z.; Pingyun, F.; Xianhui, B. Size-Selective Crystallization of Homochiral 12735
DOI: 10.1021/acs.jpcb.6b09450 J. Phys. Chem. B 2016, 120, 12730−12735
