Selective Crystallization of Phosphoester Coordination Polymer for the

Subscriber access provided by MONASH UNIVERSITY

Article

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 J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.6b09450 • Publication Date (Web): 14 Nov 2016 Downloaded from http://pubs.acs.org on November 19, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

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, Ibaraki, Japan. E-mail: [email protected]

Abstract Thermodynamics of the formation of coordination polymers (CPs) or metal–organic frameworks (MOFs) have not been focused on, while many CPs or MOFs have been synthesized in a solution. With a view to 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,

∗

To whom correspondence should be addressed Environmental Management Research Institute, National Institute of Advanced Science and Technology (AIST). 16-1 Onogawa, Tsukuba, Ibaraki, Japan. ‡ Current address: Graduate School of Science and Engineering, Saitama University. 255 Shimo-okubo, Sakura-ku, Saitama-shi, Saitama, Japan. †

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

which takes the apparent solubility products (K asp ) for [Ln(dbp)3 ] and the acid dissociation constant of Hdbp into account. The K asp values for [Nd(dbp)3 ] and [Dy(dbp)3 ] are experimentally determined to be (1.3 ± 0.1) × 10−14 M4 and (2.9 ± 0.4) × 10−18 M4 at 20 ◦ C. The ratio of these K asp values, i.e., ca. 4,500, 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 while 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 over 0.9. Using acid such as HCl or HNO3 , have 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 ratio 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 coexistence metal ions. The selective precipitation or crystallization of metal ions, for which inorganic anions 2 ACS Paragon Plus Environment

Page 2 of 20

Page 3 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


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 a high selectivity against the target 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 were feasible, it would mean that i) a wide variety of organic ligands could be applied to form CPs or MOFs, and ii) that organic solvents would not necessarily be required for this process, i.e., 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 i) a ligand that highly selectively coordinates with target metal ions, ii) environmentally friendly solvent, e.g., water, and iii) optimization of the solution composition. Point (i) is basically in agreement with the preference for mononuclear complexing. Point (ii) may be difficult to achieve, because CPs or MOFs are generally synthesized in organic solvents such as N, N-dimethylformamide and dimethylacetamide. However, any solvent, including water, can technically be employed if the target ligand is soluble in it. Point (iii) concerns the analysis of the reaction based on the solution equilibria. Such an analysis would be easier if water were 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(C2 H5 )(CH2 )3 CH3 )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



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

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 dibuthyl phosphoric acid (PO(OH)(CH2 )3 CH3 ; Hdbp) in anticipation of its higher solubility in water because Hdbp has shorter alkyl chains than Hdehp. Hdbp is a phosphoesters and should have a high selectivity for intra-lanthanide ions. The solution equilibria were analyzed by characterizing the crystalline and solution phases, which enables to optimize the separation condition. Furthermore, the effect of HCl and HNO3 were compared because HNO3 offered better separation in the previous study. 11

Experimental Formation of both the Nd– and Dy– coordination polymers 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– coordination polymer 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)), LnX3 (ci (Ln3+ )) were 0.06 ± 0.002 M 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 PTFE 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.


Page 4 of 20

Page 5 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Selective crystallization of Dy from Nd–Dy mixture The procedure was basically the same as described above. The ci (Nd3+ ) and ci (Dy3+ ) were 0.021 ± 0.003 M, while the 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 coordination polymers and solution phases The filtrates (3–5 mL depending on the concentration of the species) were decomposed in a mixture of HNO3 (60%, 3 mL)–and–H2 O2 (30%, 0.3 mL) under microwave irradiation. The concentrations of P, Nd, and Dy were then determined by inductively coupled plasma-optical emission spectroscopy (ICP-OES; Shimadzu, ICPE-9000). The precipitates were characterized by a scanning electron microscopy (SEM; JEOL, JSM-6010LA), and powder X-ray diffraction (PXRD; Bruker, D2PHASER). The contents of C, H, and N in the precipitates were obtained using an organic elemental analyzer (PerkinElmer, 2400 II). The final c(H+ ) values of the filtrates were determined from the electric potential using pH meter (Metrohm, pH meter 744) when ci (H+ ) was smaller 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+ ).

Results and discussion Determination of reaction equilibria HCl media.

Figure 1 shows the crystallization yields of Nd and Dy in HCl media as a

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



100

Crystallization yield / %

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

95

100

Dy

Nd

95

90

90

85

85

80 0.04

Page 6 of 20

0.06

0.08

0.10

80 0.0

0.2

0.4

0.6

ci(H+) / M

Figure 1: Change in crystallization yields of Nd and Dy with ci (H+ ). dissociation of Hdbp is involved in the crystallization. Hdbp ⇀ ↽ H+ + dbp−

(1)

The UV–VIS spectra revealed that Ln3+ in the solution phases is present in hydrated form (Figure S2). Figure 2 shows the SEM micrographs of the Nd– and Dy– coordination polymers. Regardless of c(H+ ), the Nd–precipitates are rod-like, 3–6 µm in thickness and 10–70 µm in length; the Dy–precipitates are fibrous, ∼1 µm in thickness and over 30 µm in length. The contents of C and H in [Ln(dbp)3 ] shown in Table 1 reveals that the chemical formula of the crystals is [Ln(dbp)3 ] (Ln = Nd or Dy). Attenuated total reflectance Fourier transform infrared (ATR-FTIR) 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


Page 7 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(A) [Nd(dbp)3] c(H+) = 0.057 M

c(H+) = 0.089 M

(B) [Dy(dbp)3] c(H+) = 0.21 M

c(H+) = 0.50 M

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 media [Nd(dbp)3 ] c(H+ ) C /M /% 0.057 37.2 0.089 37.3

[Dy(dbp)3 ] c(H+ ) C /M /% 0.21 37.2 0.50 37.2

H /% 7.20 7.18

H /% 7.20 7.20

Calc. for [Nd(dbp)3 ]: C, 37.35; H, 7.052. Calc. for [Dy(dbp)3 ]: C, 36.48; H, 6.889.

Figure 3 shows the PXRD patterns, revealing that [Nd(dbp)3 ] and [Dy(dbp)3 ] are crystalline. Dibuthyl 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 patterns the strongest peak is located at 2θ ∼ 7◦ and four or five moderate peaks are found at 7 ACS Paragon Plus Environment


2θ ∼ 12◦ , 14◦ , 18◦ , and 20◦ (Table S1). Additionally, the PXRD patterns are both very similar to that of [Ln(dehp)3 ] reported in the previous study. Based on that, we predict that the structure consists of hexagonally packed linear chains in which 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 rod-like morphologies of [Ln(dbp)3 ] (Figure 2) may be explained by growth in the main-chain 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. 400

(A) [Nd(dbp)3]

300

200

intensity / counts

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 20

c(H+) = 0.057 M

100

c(H+) = 0.089 M 0

(B) [Dy(dbp)3] 300

200

c(H+) = 0.21 M

100

c(H+) = 0.50 M 0

10

15

20

25

2θ(Cu-Kα) / degree

Figure 3: PXRD patterns of [Nd(dbp)3 ] (A) and [Dy(dbp)3 ] (B).

From these results, the following reaction should be taken into account when analyzing of crystallization: − ⇀ Ln3+ hyd + 3dbp ↽ [Ln(dbp)3 ]

HNO3 media.

(2)

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


Page 9 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Ln P O C

Figure 4: Schematics of the structure of [Ln(dbp)3 ]. 18 3+ caused by a large difference in the association of NO− and Dy3+ ; theoretical and 3 with Nd 3+ experimental studies have revealed that NO− than Dy3+ . 21 3 shows a stronger affinity for Nd

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, while the replacement of dehp− with NO− 3 in crystals may enhance the solubility of Nd crystals, resulting in a better separation of Nd3+ than Dy3+ in HNO3 than in HCl media. 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 for the crystals are almost the same as those for [Ln(dbp)3 ] obtained in HCl media, i.e., nitrogen was not detected. Therefore, the chemical composition of the crystals is [Nd(dbp)3 ] and [Dy(dbp)3 ] in common in HCl systems. This is supported by the fact that there were no differences in SEM micrographs of the crystals from HCl and HNO3 Table 2: C, H, and N contents in [Ln(dbp)3 ] obtained in HNO3 media. [Nd(dbp)3 ]

[Dy(dbp)3 ]

c(H+ ) /M

C /%

H /%

N /%

c(H+ ) /M

C /%

H /%

N /%

0.034

37.5

7.32

0.00

0.21

36.3

7.05

0.00

0.062

37.3

7.24

0.00

0.51

36.5

7.06

0.00



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

systems (Figure S4). This finding is not consistent with the previous study using Hdehp. A difference in the alkyl chain length, solvent (i.e., ethanol-water or water), or crystallinity 3+ plausibly affect the association of NO− 3 with Nd .

Figure 5 compares the absorption spectra of the filtrates for either HCl or HNO3 media. The band around 575 nm correspond to the 4 I9/2 →4 G5/2 , and 2 G7/2 transition 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 the band 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 enhances the association between Nd3+ and NO− 3 ; the 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 3+ significantly affects the association of NO− 3 with Nd , which enhances separation. However,

we have not investigate 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 that should be taken into account in aqueous media are given by Eqs. (1) and (2) irrespective of the kind of acid.


Page 10 of 20

Page 11 of 20

40

Molar absorption coefficient / M-1 cm-1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Ndhyd Nd(NO3)2+ 30

ci(H+)=0.05 M

20

10

HNO3 HCl

ci(H+)=0.08 M

0 550

560

570

580

590

600

Wavelength / nm

Figure 5: Absorption spectra of Nd3+ in HCl and HNO3 media. The broken blue lines are the differential spectra.

Estimation of selectivity A solubility product, Ksp = a(Ln3+ ) · a(dbp− )3 (a: activity), is a key parameter in separaa tion by crystallization. In this study, the apparent solubility product (Ksp ) when the ionic

strength is 0.6 is defined as follows:

a Ksp (Ln) =

Ksp (Ln3+ ) y(Ln3+ ) · y(dbp− )3

= c(Ln3+ ) · 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 (c.f. Eq. (1)) is similarly defined :

Kaa

c(H+ ) · c(dbp− ) = c(Hdbp)


(4)


1.5

c(Nd3+), c(Dy 3+) / 10-3 M

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 20

1.5

3+

c(Nd )

1.0

1.0

3+

0.5

0.0

c(Dy )

0.5

0

5

10

×10 9

15

0.0

0

{(1+α) / c (P)}3 t

2

4

6

8

×1013

/ M3

Figure 6: Change in c(Ln3+ ) with {(1 + α)/ct (P)}3 . Eqs. (3) and (4) yield ( 3+

c(Ln ) =

a Ksp (Ln)

·

1+α ct (P)

)3 (5)

where α = c(H+ )/Kaa and ct (P) = c(Hdbp) + c(dbp− ). The Kaa at 20 ◦ C was estimated to be 4.0 × 10−3 (±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+ ) vs a a {(1 + α)/ct (P)}3 (Figure 6) give Ksp (Nd)=1.3 × 10−14 (±0.1 × 10−14 ) M4 and Ksp (Dy)=2.9 × a a 10−18 (±0.4 × 10−18 ) M4 at 20 ◦ C. The reason that Ksp (Dy) is smaller than Ksp (Nd) is likely

due to the stronger electrostatic interaction of Dy3+ with dbp− caused by its smaller ionic a a radius compared with Nd3+ . It should be noted that the Ksp (Nd)/Ksp (Dy) ratio of ca. 4,500 a (3,600–5,600 if the deviation of Ksp is taken into account) is remarkably large as far as Nd3+ a and Dy3+ are concerned. For example, the ratio of the Ksp values for phosphorous salt, 23

oxalate, 24 and hydroxide 25 are estimated to be 0.1, 1.5, and 500, respectively. The high selectivity of the phosphoester, Hdbp, against intra-lanthanide ions is likely to be reflected in the result, although stability coefficients of Hdbp-complex with Nd3+ and Dy3+ has not been reported. Formation of [Ln(dbp)3 ] has the potential to crystallize Dy3+ selectively over Nd3+ in an aqueous solution.


Page 13 of 20

Fractional crystallization a The c(Nd3+ ) and c(Dy3+ ) in their mixture systems are calculated using the values of Ksp (Nd), a Ksp (Dy), and Kaa . Considering Eq. (3) for both Nd3+ and Dy3+ and mass balance of

phosphor, the following equation related to X = c(Nd3+ ) + c(Dy3+ ) can be derived: {

X a Ksp (Nd)

} 13 =

1 + ci (H+ )/Kaa ci (P) − 3{ci (Nd3+ ) + ci (Dy3+ ) − X}

(6)

a a a where the approximations Ksp (Nd) + Ksp (Dy) ≈ Ksp (Nd), and c(H+ ) ≈ ci (H+ ) were ema a ployed. Using Ksp (Nd)/Ksp (Dy) = c(Nd3+ )/c(Dy3+ ) , both c(Nd3+ ) and c(Dy3+ ) can be

calculated when ci (H+ ), ci (P), ci (Nd3+ ), and ci (Dy3+ ) are given. (A)

c(Nd3+), c(Dy3+) / M

0.020

c(Nd3+) 0.015

0.010

calc. HCl 0.005

HNO3

c(Dy3+) 0.000 0.058

0.060

0.062

0.064

ci(P) / M

(B) 0.020

c(Nd3+), c(Dy3+) / M

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


c(Nd3+) 0.015

0.010

0.005

c(Dy3+) 0.000 0.1

0.2

0.3

0.4

0.5

ci(H+) / M

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). 13 ACS Paragon Plus Environment


0.020

c(Nd3+) , c(Dy 3+) / M

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 20

c(Nd3+) 0.015

HNO HCl

0.010

0.005

0.000

c(Dy3+)

0

2

4

6

8

Time / day

Figure 8: Time course of c(Nd3+ ) and c(Dy3+ ) in HCl and HNO3 . In the present study, 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 Figures 7(A) and (B) represent the calculated c(Nd3+ ) and c(Dy3+ ) values as a function of ci (P) and ci (H+ ), a respectively, in consideration of the errors of the Ksp and Kaa values. The deviation of the a calculated results seems small compared with the relatively large deviations of the Ksp 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 in a range of ci (P) from 0.058 to 0.065 M and ci (H+ ) from 0.1 to 0.5 M, while 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 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 with triangles and circle, respectively, in Figures 7(A) and (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.


Page 15 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


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; in addition, the actual c(Dy3+ ) values are larger than the calculated ones. An inclusion of Nd3+ into [Dy(dbp)3 ] may be involved in since 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 the reaction time. Figure 8 shows the change in c(Nd3+ ) and c(Dy3+ ) with time in HCl and HNO3 media. The c(Nd3+ ) 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 a calculation using Ksp 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+ is significantly separated especially at ci (H+ ) = 0.18 M and ci (P) = 0.0635 M, where the separation factor is over 300 and the molar ratio of Dy to the total metal content in the crystal was over 0.9. For example, the separation factor for the Dy/Nd fractional precipitation system using oxalate 24 can be calculated to be about two. In addition, selective crystallization of camphorate metal–organic framework 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.

Conclusions Nd3+ and Dy3+ were successfully separated in an acidic aqueous solution via the fractional



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 20

precipitation of Hdbp coordination polymers. 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 suggested that selective crystallization of CPs or MOFs can enhance the separation performance of metal ions in a wide variety of metal ion separation.

Associated Content Supporting Information Time course of precipitation; UV-VIS spectra of the filtrates; ATR FT-IR spectra of [Ln(dbp)3 ]; SEM images of [Ln(dbp)3 ]; titration of Hdbp; peak list of PXRD

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. MetalOrganic 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 MetalOrganic Frameworks. Chem. Soc. Rev. 2009, 38, 1330–1352. (4) Batten, S. R.; Neville, S. M.; Turner, D. R. Coordination Polymers; 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, Phase-Selectivity, and Size Reduction. Coord. Chem. Rev. 2015, 285, 11 – 23.


Page 17 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(6) Feng, D.; Liu, T.-F.; Su, J.; Bosch, M.; Wei, Z.; Wan, W.; Yuan, D.; Chen, Y.-P.; Wang, X.; Wang, K. l. Stable Metal–Organic Frameworks Containing Single Molecule Traps for Enzyme Encapsulation. Nat. Commun. 2015, 6, 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, J. L.; 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 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 , H2 SO4 and 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-FeB Magnet Wastes with TBP in Tricaprylmethylammonium Nitrate. Solv. Extr. Res. Dev. Jpn. 2014, 21, 137–145.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(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 Metallic-TFSA 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) Stary, 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. Japan 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.


Page 18 of 20

Page 19 of 20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(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 Formtion of Rare Earth Inorganic Complexes as a Function of Ionic Strength. Geochim. Cosmochim. Acta. 1992, 56, 3123–3132.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Graphical TOC Entry Nd3+ + 3dbpDy3+ + 3dbp-

Ksp(Nd) Ksp(Dy)

[Nd(dbp)3] [Dy(dbp)3]

Ksp(Nd) Ksp(Dy)

~ 4,500 !!

Nadbp aq. Nd aq. [Dy(dbp)3 ]

Nd+Dy aq.


Page 20 of 20

Selective Crystallization of Phosphoester Coordination Polymer for the

Recommend Documents