α-Titanium phosphate intercalated with propylamine: An alternative pathway for efficient europium(III) uptake into layered tetravalent metal phosphates

1 Introduction

Recently, Ortíz-Oliveros et al. (2014) have published the synthesis, physical-chemical characterization and preliminary evaluation of α-Ti(HPO₄)₂·H₂O (α-TiP) sorption of Eu³⁺, which is often considered as a chemical analogue for trivalent heavy metals, lanthanides and also actinides (such as Np, Am, Cm and Pu, radionuclides found in high-level radioactive waste produced during the nuclear fuel cycle). The authors concluded that the retention process for Eu³⁺ can take place on the surface and in the cavities of the solid in two stages: (a) diffusion into the cavities of the material and (b) chemisorption. In this work, we will show that the uptake process is, in fact, limited to the surface of the α-titanium phosphate (the interlayer space is inaccessible to the europium species when the α-TiP hydrogen form is used as sorbent), while the diffusion into internal cavities is only possible when basal spacing is previously expanded (e.g., intercalated by propylamine).

Ortíz-Oliveros et al. (2014) addressed, among other things, the evaluation of Eu³⁺ sorption capacity by crystalline α-TiP, based on the fact that the titanium and zirconium phosphates, and other isomorphic tetravalent metal phosphates α-M(HPO₄)₂·H₂O (M = Zr, Ti, Sn, etc.) have been successfully used in the recovery of uranium, plutonium and fission products (Zhuravlev et al., 2002; Mrad et al., 2011). However, Zhuravlev et al. (2002) relied on the use of amorphous samples of titanium and zirconium phosphates (modified with Al³⁺ or Fe³⁺ ions during the gelation process), whereas Mrad et al. (2011) carried out their work using semi-crystalline Zr(HPO₄)₂·H₂O (where an adsorption occurs on the surface, and an ion-exchange mechanism is not fully confirmed).

Metal salts of phosphoric acid have been known for over a century. Because of their extraordinary properties, a great variety of potential and realized uses have been invoked. Among the many uses in addition to both ion-exchange and intercalation-chemistry are catalysis, polymer composites, proton conduction, drug delivery and many others, such as energy storage, sensors and biosensors, flame retardants, and antibiomaterials (Clearfield and Díaz, 2015). Titanium phosphate can be prepared either as gel or in intermediate stages of crystallinity, but also in several crystalline forms including γ-Ti(PO₄)(H₂PO₄)·2H₂O (γ-TiP) and α-TiP (Salvadó et al., 1996; García-Granda et al., 2010), being known that the main factors affecting the ion-exchange behavior of tetravalent metal phosphates are their structures and the degree of crystallinity (Llavona et al., 1989). α- and γ-TiP are layered compounds with flexible two-dimensional structures that, although they have ion-exchange properties, display low affinity toward both transition and inner transition metal ions (Álvarez et al., 1987; Trobajo et al., 1991). The synthesis of ion-exchange phases with polyvalent cations is possible by means of an indirect route. In the first stage, the reaction between the inorganic solid acid (α- or γ-TiP) and the vapor of an organic base (e.g., n-alkylamine), produced an essentially non-porous intercalation compound (Menéndez et al., 1993; Espina et al., 1998a, 1998b), usually with strong preferential orientation effects and thermal disorder of the alkyl chains (Mafra et al., 2005, 2008). In a second step, the intercalation compound reacts with an aqueous solution of the desired metal salt, and the substitution of intercalated organic cation by the inorganic metal cation takes place (Trobajo et al., 1991; Alfonso et al., 2005).

In this paper, α-TiP intercalated with propylamine (α-TiPPr) has been selected to investigate its sorption of Eu(III) by means of PXRD, SEM, TEM, TGA and PL, Moreover, the structural modeling of a hypothetical full ion-exchange phase, [Eu(H₂O)₆]_2/3Ti(PO₄)₂·[(H₂O)₆]_1/3, has been also proposed on the basis of DFT calculations.

2 Materials and methods

3 Results and discussion

PXRD analysis shows the both TiP_10⁻⁴ and TiP_0.1 samples are highly crystalline and their patterns are identical, as seen in Fig. 1, with the first characteristic peak at 2θ = 11.65° (d₀₀₂ = 7.58 Å) corresponding to the interlayer distance of α-TiP. TEM images in Figs. 2a and b and 3a and b show that both samples maintain the typical platelet-like pseudo-hexagonal morphologies (length: 0.2–0.8 μm; thickness: 10–30 nm) similar to other metal phosphates (Manickam et al., 2004). However, in the case of TiP_0.1, an additional phase with nanorod-like morphology (thickness of ca. 10 nm) has been formed (for instance Fig. 3b and c). This phase was undetectable by PXRD as a consequence of its low content, poor crystallinity and nano crystallite size. In addition, the selected area electron diffraction (SAED) pattern of these nanorods found in TiP_0.1 (Fig. 3d) has been indexed as standard hexagonal EuPO₄ H₂O phase (P3₁2₁, a = 6.91 Å, b = 6.34 Å; ICSD PDF4: 20-1044). In contrast, the only identified phase in the case of TiP_10⁻⁴ was α-TiP, as confirmed by SAED patterns in Fig. 2c and d. Furthermore, the chemical composition of EuPO₄ nanorods was confirmed by EDX analysis, revealing that both α-TiP and EuPO₄ phases are clearly distinguishable by EDX elemental mapping (see Fig. 4) in the case of TiP_0.1, while the only detectable element in the case of TiP_10⁻⁴ was Ti, P and O as seen in Fig. S2. In order to investigate the photoluminescence properties of TiP_10⁻⁴ and TiP_0.1, the emission spectra were obtained with λ_ex = 394 nm as shown in Fig. 5. The emission spectrum of TiP_10⁻⁴ does not show any transition line due to the sorption of Eu³⁺. In contrast, the TiP_0.1 spectrum shows the typical emission transitions of Eu³⁺, which are attributed to ⁵D₀ → ⁷F_J (J = 0–4) transitions at, i.e. 580 nm (⁵D₀ → ⁷F₀), 592 nm (⁵D₀ → ⁷F₁), 616 nm (⁵D₀ → ⁷F₂), 650 nm (⁵D₀ → ⁷F₃), and 689 and 697 nm (⁵D₀ → ⁷F₄), with the most intense peak corresponding to the ⁵D₀ → ⁷F₂ transition. These transition lines correspond basically to the EuPO₄·H₂O nanorod phase. The observed low-intensity ⁵D₀ → ⁷F₀ line is expected for Eu³⁺ in a site with C_n, C_nv and C_s symmetry. The high relative intensity of the hypersensitive ⁵D₀ → ⁷F₂ transition is in accord with the Eu³⁺ local site symmetry (C_s), which is determined based on the crystal structure of hexagonal EuPO₄·H2O phase. Moreover, the ⁵D₀ → ⁷F₄ transition is less intense than the ⁵D₀ → ⁷F₂ transition but much more intense than the ⁵D₀ → ⁷F₁ transition, and the same remarks have been made with Eu³⁺ in an asymmetric site with Cs symmetry in the case of LaBO₃:Eu³⁺ with an orthorhombic aragonite structure (Binnemans, 2015).

Close

Figure 1

PXRD patterns of the samples TiP_10⁻⁴ (red) and TiP_0.1 (blue), and calculated for α-TiP (black).

Export to PPT

Close

Figure 2

(a and b) TEM images, (c) selected area electron diffraction (SAED) patterns (left: simulated, right: experimental), and (d) nanobeam electron diffraction (NBD) pattern along [2–11] zone axis for a single particle of TiP_10⁻⁴ sample.

Export to PPT

Close

Figure 3

(a–c) TEM images of TiP_0.1 sample and (d) selected area electron diffraction (SAED) patterns (left: simulated, right: experimental) corresponding to the image (c).

Export to PPT

Close

Figure 4

Elemental mapping for TiP_0.1 sample: (a) Bright field-STEM images, (b) Ti-Kα₁ map, (c) Eu-Lα₁ map and (d) P-Kα₁ map.

Export to PPT

Close

Figure 5

Emission spectra of the samples TiP_0.1 (red) and TiP_10⁻⁴ (blue) obtained up excitation at 394 nm.

Export to PPT

From all the above data, we can conclude that the Eu³⁺ sorption process takes place only on the surface of α-TiP rather than in its interlayer space, because of the following summarized observations: (i) at high [Eu³⁺]_i no change in the interlayer distance of α-TiP has been observed by PXRD, and (ii) Eu³⁺ was undetectable by EDX analysis or by PL emission spectra at low [Eu³⁺]_i.

In this context, α-TiP intercalated with propylamine (TiPPr) was selected as a potential material for efficient europium(III) uptake into layered tetravalent metal phosphates. For this purpose, TiPPr was treated with europium nitrate solutions at different concentrations until the equilibrium is reached. PXRD patterns of TiPPr_10⁻⁴ to TiPPr_0.1 (Fig. 6) show a structural order in the direction perpendicular to the plane of the sheet. In the range 0.0001–0.0625 M, the interlayer distance of the sheets remains substantially constant (16.0 Å), which is slightly lower than that of the starting material (16.9 Å, α-TiPPr) (Menéndez et al., 1990) as a consequence of a partial evacuation of the amine occupying the interlayered space. However, a new diffraction peak at 2θ = 7.3° corresponding to d-spacing of 12.0 Å appears when the concentration is increased to 0.0750 M, which became the only observed peak when the [Eu³⁺]_i is raised to 0.1000 M. Moreover, the carbon content of the resulting solids (Table 1) continuously decreases with increasing concentration of europium nitrate in the starting solutions, in agreement with the results of STEM-EDS elemental mapping (Fig. S3), which obviously indicates the removal of the amine increases as the ionic strength of the solution increases. As a result, the behavior TiPPr_x can be divided into four different zones (red-blue-green-black), as illustrated in Table 1 and Fig. 6, depending on the concentration of Eu³⁺ in the initial solutions.

Close

Figure 6

PXRD patterns of the TiPPr_x samples.

Export to PPT

In the zone of low concentrations (red zone), the material consists of slightly wrinkled thin films (Figs. 7a–c and 8a–c) with lateral dimensions varying from about 0.1 μm to 1 μm and thicknesses of ca. 6.0 nm measured by atomic force microscopy. These thin films maintain a long-range order in the direction of the crystallographic c-axis as revealed by PXRD studies (Fig. 6), while a certain order is still maintained in ab plane as illustrated by the SAED pattern (see inset in Fig. 8a–c). Moreover, the propylamine content decreases in this red zone (see Table 1 and Fig. 6) compared to the starting material TiPPr with interlayer distance 16.9 Å (Menéndez et al., 1990). However, the presence of Eu³⁺ in these films was undetectable by STEM-EDS elemental mapping as shown in Fig. 9a and b, in agreement with the emission spectrum (for instance TiPPr_10⁻⁴ spectrum in Fig. 10) which does not show any transition line due to the sorption and/or intercalation of Eu³⁺.

Close

Figure 7

SEM images of TiPPr_x samples (x = 10⁻⁴ (a), 10⁻³ (b), 0.01 (c), 0.025 (d), 0.0375 (e), 0.0625 (f), 0.075 (g) and 0.1 (h)).

Export to PPT

Close

Figure 8

TEM images of TiPPr_x samples (x = 10⁻⁴ (a), 10⁻³ (b), 0.01 (c), 0.025 (d), 0.05 (e), 0.075 (f) and 0.1 (g and h)).

Export to PPT

Close

Figure 9

STEM-EDS elemental mapping of TiPPr_x samples (x = 10⁻⁴ (a), 10⁻³ (b), 0.05 (c), 0.075 (d and e) and 0.1 (f)); Ti-Kα₁ (red), P-Kα₁ (blue) and Eu-Lα₁ (green) elemental maps.

Export to PPT

Close

Figure 10

Emission spectra of TiPPr_10⁻⁴ (red), TiPPr_0.05 (blue) and TiPPr_0.1 (black), obtained up excitation at 394 nm.

Export to PPT

In the region of intermediate concentrations (blue zone), the layers are stacked in the direction perpendicular to the plane of the titanium phosphate sheet, where the interlayer distance is slightly lower than in the starting material TiPPr, but it is essentially constant (16.0 Å) as shown in Fig. 6, while the pseudo-hexagonal morphologies are preserved (Figs. 7d–f and 8d and e). The amine content gradually decreases (reflected in the decrease in carbon content, see Table 1); thus, the Eu-content increases in the solid phase as confirmed by EDX mapping (Fig. 9c) and by the presence of ⁵D₀ → ⁷F_J (J = 0–4) transition peaks in the emission spectrum of TiPPr_0.05 (Fig. 10). Assuming that an ion-exchange process is taking place between the propylammonium cations, C₃H₇NH₃⁺, and the hexahydrate species of europium, [Eu(H₂O)₆]³⁺, the following tentative formula, (C₃H₇NH₃)_2–3y[Eu(H₂O)₆]_yTi(PO₄)₂·[(H₂O)₆]_y/2, has been proposed for the resulting materials. This process proceeds via the formation of solid solutions, where C₃H₇NH₃⁺ species act as pillars adopting a bilayer structure with the tilt angle of the alkyl chain of ca. 50° relative to the plane of the titanium phosphate sheet, and occupying a fraction of the pseudo-zeolitic cavities of the titanium phosphate, while the rest of the cavities would be occupied by [Eu(H₂O)₆]³⁺ species. The calculated total mass loss decreases from 36.65% (y = 0.15) to 26.69% (y = 0.55), in a good agreement with experimental data: 36.1–26.6% obtained from TGA data (see Table 1).

At higher concentration (zones shown in green and black), the particles maintain their pseudo-hexagonal morphologies (Fig. 7g and h and 8f and g), and besides the presence of an interlayer distance of 16.0 Å in the case of TiPPr_0.075, the appearance of a new phase with d-spacing of 12.0 Å is also observed (Fig. 6), which is the only one that remains in the black zone (TiPPr_0.1). HRTEM (Fig. 8h) shows a d-spacing of ca. 11.0 Å, close to the value obtained by PXRD, and this difference is due to the instantaneous dehydration followed by a fast amorphization of the particles under the electron beam irradiation. In this latter zone, the low carbon content suggests that almost all the alkylammonium cations have been involved in the ion-exchange process, thus leading to the formation of a material with the idealized formula [Eu(H₂O)₆]_2/3Ti(PO₄)₂·[(H₂O)₆]_1/3, where [Eu(H₂O)₆]³⁺ occupying 2/3 of the pseudo-zeolitic cavities so as to counteract the negatively charged titanium phosphate layer. The rest of the pseudo-zeolitic cavities (1/3 of the total) will be occupied by water molecules. Based on this expected formula, the calculated mass loss of 24.15% is in good accordance with the experimental mass loss of 25.0%. Moreover, STEM-EDS elemental mapping of TiPPr_0.075 (Fig. 9d and e) and TiPPr_0.1 (Fig. 9f), corresponding to the green and black zones, respectively, shows a significant increase in Eu-content compared to the blue zone, and it is noteworthy that in the case of green zone there are particles with low and high Eu-content (Fig. 9d and e), which could be related to the presence of two phases with different basal spacing (16.0 Å and 12.0 Å), as revealed by PXRD analysis (Fig. 6). On the other hand, the emission spectrum of TiPPr_0.1 (Fig. 10) still shows the expected ⁵D₀ → ⁷F_J (J = 0–4) transition peaks but with a decrease in their relative intensities compared to the spectrum of the blue zone, which is probably associated with the absence of propylamine and high degree of hydration leading to emission quenching. The emission spectra of samples TiPPr_0.05 and TiPPr_0.1 show the five possible luminescence bands expected in the registered spectral region. The number of lines observed for the ⁵D₀ → ⁷F_J (J = 0–4) transitions in the luminescence spectrum allows determining the site symmetry of the Eu³⁺ ion (Binnemans, 2015). The highest energy ⁵D₀ → ⁷F₀ line is very weak, in agreement with Judd–Ofelt theory (the transition should only be expected to be observed for Eu³⁺ in a site with C_n, C_nv or C_s symmetry). The very intense hypersensitive transition ⁵D₀ → ⁷F₂ indicates that the Eu³⁺ is not at a site with a center of symmetry. In addition, the transition ⁵D₀ → ⁷F₃ has very weak intensity since they are forbidden by the selection rule of ΔJ for electric dipole transitions.

The colloidal suspensions (non-centrifugal phases) of TiPPr_10⁻⁴ to TiPPr_0.1 have been analyzed by ICP-MS. The results (Table 1) show that the concentrations of both P and Ti decrease with increasing [Eu³⁺]_i. In the red zone (low concentrations), about three quarters of the starting material TiPPr are found in the liquid phase, mainly in the form of nanometric colloidal TiP monolayers, although a portion of this ¾ is dissolved due to the partial hydrolysis of the starting material. While in the blue zone (intermediate concentrations), the amount of the nanometric colloidal TiP suspension decreases drastically with increasing [Eu³⁺]_i, which is reflected in the decrease in P and Ti content in the liquid phase. The fact that the percentage of P in solution is higher than Ti, in both the red and blue zones, indicates the existence of hydrolysis processes of the titanium phosphate, which may give rise to the formation of aqueous phosphate species and precipitation of titanium dioxide gels. Finally, in the green zone, and especially in the black one, high [Eu³⁺]_i causes the liquid phase to be substantially free of P and Ti. However, in the solid phase, and beside the hypothetical ion-exchange phase [Eu(H₂O)₆]_2/3Ti(PO₄)₂·[(H₂O)₆]_1/3, it shall be present variable amounts of EuPO₄·H₂O (for instance, see nanorods in Fig. 8g) and TiO₂·nH₂O (not observed by TEM), coexisting with a residual phase containing propylamine (Fig. S3d).

The final structural model proposed by using DFT calculations for the material with the idealized formula [Eu(H₂O)₆]_2/3Ti(PO₄)₂·[(H₂O)₆]_1/3 is shown in Fig. 11. Its structure consists of PO₄ tetrahedral (Fig. 11a) and TiO₆ octahedral (Fig. 11b) units connected through their vertices, giving rise to 2D-dimensional inorganic sheets, which are still closely related to those of the α-TiP. The hexahydrate species of europium, [Eu(H₂O)₆]³⁺ (Fig. 11c)_, and non-coordinated H₂O molecules occupy the interlamellar space, with the average Eu–O bond distance and Eu–Eu distance between adjacent [Eu(H₂O)₆]³⁺ cations of 2.58 Å and 4.91 Å, respectively.

Close

Figure 11

Perspective view of the coordination environment of P (a), Ti (b), Eu (c) and the projection of the structural model of [Eu(H₂O)₆]_2/3Ti(PO₄)₂·[(H₂O)₆]_1/3 along the a-axis (d).

Export to PPT

4 Conclusions

In summary, we have shown that when α-TiP is used as sorbent the Eu³⁺ uptake process is limited to its surface, being the interlayer space inaccessible to the hexahydrate [Eu(H₂O)₆]³⁺ cations. However, the Eu³⁺ uptake is considerably improved with α-TiPPr as a consequence of an ion-exchange process into the expanded interlayer space, involving C₃H₇NH₃⁺ cations and [Eu(H₂O)₆]³⁺ species, with the formation of α-[Eu(H₂O)₆]_2/3Ti(PO₄)₂·[(H₂O)₆]_1/3 as final product.