Structure–property relations in hexagonal and monoclinic BiPO₄:Eu³⁺ nanoparticles synthesized by polyol-mediated method

Abstract

Hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O nanoparticles were synthesized by a polyol-mediated method employing diethylene glycol. The powder X-ray diffraction revealed the phase purity and isostructural nature of both undoped and Eu³⁺-doped BiPO₄. The monoclinic Bi_0.95Eu_0.05PO₄ was obtained by heating the hexagonal Bi_0.95Eu_0.05PO₄·xH₂O at 600 °C. The microscopical characterization revealed the formation of nanocrystalline materials. Water molecules present in the bismuth precursor favoured the formation of hexagonal phase at low temperature. The role of DEG molecules in arresting the particle growth during the phase transformation of Bi_0.95Eu_0.05PO₄ from hexagonal to monoclinic was observed. The synthesized materials were characterized using different spectroscopic techniques such as FT-IR, Raman, ³¹P MAS-NMR, DRUV-Vis and PL. The difference in the crystal structures and symmetries is clearly reflected in the spectral results of hexagonal and monoclinic Bi_0.95Eu_0.05PO₄. The structure-property relations were studied to derive its importance from both fundamental and technological aspects.

---

1 Introduction

BiPO₄ is known for its multifarious applications such as catalyst,¹orthophosphate ion sensor,² microwave dielectric,³ and in separation of radioactive elements.⁴ Recent focus on BiPO₄ is due to its application as a host lattice for rare earth luminescent centers and as a photocatalyst.^5–8BiPO₄ crystallizes in three different structures with hexagonal and monoclinic (low and high temperature modifications) symmetries.^9,10 All the three phases of BiPO₄ consist of three dimensional network structure made up of PO₄ tetrahedra and BiO₈ polyhedra. The most stable form is the low temperature monoclinic monazite-type (LTBP) that can be converted into high temperature monoclinic structure (HTBP) when heated at 700 °C and above. Both LTBP and HTBP have closely related crystal structures. The third form is the hexagonal phase of hydrated BiPO₄·xH₂O which can be synthesized at room temperature through a wet chemical method.⁹ Synthesis of BiPO₄ with different morphologies has received considerable research attention in recent times. An urchin-like morphology of monoclinic BiPO₄ nanoparticles with rare earth doping was obtained by hydrothermal method for its application in high performance luminescent devices.⁶ Monoclinic BiPO₄ nanowires were obtained by the decomposition of single source precursor tris-chelated diselenophosphato complex.¹¹ Monoclinic BiPO₄ and BiPO₄:Tb³⁺ were synthesized by polyol-mediated method and the Tb³⁺ photoluminescence (PL) emission intensity was found to increase with increasing post annealing temperatures.⁵ Hexagonal BiPO₄ nanorods were obtained by a sonochemical method without the assistance of surfactant or ligands.¹² Recently, the hexagonal BiPO₄ nanorods have been synthesized by electrochemical anodization of Bi metal substrate and its PL property has been reported.¹³

Recent studies revealing the potential of BiPO₄ as a host lattice for rare earth luminescent ions motivated us to study the PL properties of Eu³⁺ in different polymorphs of BiPO₄. We used the polyol mediated synthesis which is a simple and versatile technique involving the precipitation of nanocrystalline materials in the presence of a multidendate and high-boiling alcohol such as diethylene glycol (bp 246 °C).^14,15 The alcohols, in addition to their role as solvent, act as stabilizers of nanoparticles by preventing the agglomeration and inhibiting the particle growth. The objectives of this work are: primarily the spectroscopic characterization of the host BiPO₄ in its two different polymorphs, namely hexagonal and monoclinic, in the nanoscale and to elucidate their structure–spectral properties correlation. Although these correlations has been carried out in the bulk, a comparative study of the spectral properties of BiPO₄ nanoparticles, either undoped or doped, is necessary considering the changes in properties at the nanoscale.⁹ Secondly, we chose Eu³⁺ since its emissions are hypersensitive i.e., they strongly depend on the local structural symmetry of the host lattice.^16,17 Understanding the PL properties of Eu³⁺ in hexagonal and monoclinic BiPO₄ nanoparticles is of fundamental interest and this may pave way for the improvement of properties by controlling the surface-bound species in the nanoscale phosphors, especially for their application in white LED technology. In the present study, we report the synthesis of nanocrystalline hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O by polyol-mediated method and the conversion of hexagonal Bi_0.95Eu_0.05PO₄·xH₂O into monoclinic monazite phase by thermal treatment without affecting the particle size. The obtained hexagonal and monoclinic Bi_0.95Eu_0.05PO₄nanoparticles were characterized by various spectroscopic techniques and the results are compared.

2 Experimental

2.1 Synthesis

The polyol-mediated synthesis procedure adopted in the present study is similar to the one reported by Roming and Feldmann except for the starting materials.⁵ In a typical synthesis, 1.6009 g (3.3 mmoles) of Bi(NO₃)₃·5H₂O (Merck 98%) was dissolved in 100 ml of diethylene glycol (DEG) (SISCO Research Lab. Pvt. Ltd. 99%) at 60 °C. A solution of phosphate precursor consists of 0.4358 g (3.3 mmoles) of (NH₄)₂HPO₄ (Lobachemie 98–100%) in 4 ml dist. H₂O was added into DEG solution containing bismuth nitrate under vigorous stirring. The particles were formed immediately and the solution was heated rapidly to 160 °C and kept at this temperature for 1 h followed by cooling to room temperature under constant stirring. After cooling, 100 ml ethanol was added and stirred for 30 min and centrifuged. The obtained particles were washed several times with ethanol and dried at 100 °C in an air oven for 8–10 h. In the case of Bi_0.95Eu_0.05PO₄, a stoichiometric amount of Eu₂O₃ was initially converted into nitrate by dissolving in conc. HNO₃ and the above synthesis procedure was followed. We obtained about 0.98 g of final product (yield 98%).

2.2 Characterization

Thermogravimetric analysis (TGA) of synthesized materials was carried out in air from 30 to 1000 °C with a heating rate of 20 °C min⁻¹ using TA (SDT Q600) instrument. The phase formation and purity were monitored by powder X-ray diffraction (XRD) technique using Cu-Kα radiation (D8 Advance, Bruker). The FT-IR spectra were recorded in the mid-IR region using KBr pellet technique (Tensor 27, Bruker). Laser Raman spectra were recorded using Reninshaw Invia Laser Raman microscope equipment with 633 nm He-Ne Laser source through CCD detector. Solid state ³¹P magic angle spinning (MAS) NMR spectra were recorded in Bruker Advance 400 spectrometer with a magnetic field of 9.4 T operating at a frequency of 161.98 MHz for ³¹P nuclei. The ³¹P MAS-NMR spectra were recorded at different spinning rate of the sample tube ranging between 5000–8000 rpm. The scanning electron micrographs (SEM) were obtained using Hitachi (S-3000H) microscope. The transmission electron microscopic (TEM) images were obtained using FEI Technai 20 G2 microscope. The diffuse reflectance UV-Vis (DRUV-Vis) absorption spectra were recorded using a spectrometer (Cary 500, Varian) and BaSO₄ was used as a standard to correct the base line. Photoluminescence (PL) emission spectra of powder samples were recorded using JASCO (FP-6500) spectrophotometer.

3 Results and discussion

3.1 Phase formation

The thermogravimetric analysis (TGA) of the undoped and Eu³⁺-doped BiPO₄ obtained by polyol-mediated method was carried out in order to find the presence of lattice water molecules and the results are shown in Fig. 1a. In the case of BiPO₄·xH₂O, a weight loss (∼2%) from room temperature to 150 °C was observed and this could be due to the removal of adsorbed water and ethanol over the particle surface.¹⁰ Another major weight loss was observed between 200 and 350 °C and this could be due to the removal of zeolitic lattice water molecules present in the open channels of hexagonal BiPO₄·xH₂O structure and some residual DEG which was used as the solvent and capping agent in the synthesis.¹⁸ The TGA result of Bi_0.95Eu_0.05PO₄·xH₂O shows almost a similar behaviour (Fig. 1b). Thus, the thermogravimetric analysis revealed the presence of zeolitic lattice water molecules and the residual organics in the sample as obtained by polyol-mediated method. There is no weight change after ∼550 °C and hence the calcination of hexagonal Bi_0.95Eu_0.05PO₄·xH₂O was carried out at 600 °C above which the conversion into high temperature monoclinic phase may occur.³ The TGA result of Bi_0.95Eu_0.05PO₄ obtained by heat treatment of hexagonal phase at 600 °C/6 h (Fig. 1c) shows a negligible weight loss (<1%) which could be due to adsorbed moisture. This result clearly reveals the complete removal of zeolitic water molecules and residual organics during the phase transformation of hexagonal Bi_0.95Eu_0.05PO₄·xH₂O into monoclinic Bi_0.95Eu_0.05PO₄ by thermal treatment.

Fig. 1 TGA curves of (a) BiPO₄·xH₂O (—) and (b) Bi_0.95Eu_0.05PO₄·xH₂O (- - - -) as obtained by polyol-mediated synthesis and (c) Bi_0.95Eu_0.05PO₄ (- · · -) obtained after calcinations of (b) at 600 °C/6 h.

The powder XRD patterns of BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O samples as obtained by polyol-mediated method are shown in Fig. 2a and 2b, respectively. All the reflections were indexed based on hexagonal symmetry with reference to the standard pattern available in the JCPDS data base (No. 15-0766). The single phase formation of both BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O is clearly evident from the fact that there are no unidentified reflections in the XRD patterns. The similarity in the powder XRD patterns of BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O reveals their isostructural nature. The calculated hexagonal lattice parameters are listed in Table 1 and values of BiPO₄·xH₂O are in good agreement with the values available in JCPDS file. There is a marginal decrease in both ‘a’ and ‘c’ lattice parameters in the case of Bi_0.95Eu_0.05PO₄·xH₂O when compared to BiPO₄ which is due to the smaller ionic radius of Eu³⁺ (1.07 Å) than that of Bi³⁺ (1.11 Å) in 8-fold coordination.¹⁹ The intense and broader reflections in the powder XRD patterns reveal the formation of the nanocrystallites. The crystallite sizes were calculated from the full width at half maximum (FWHM) of selected reflections using Scherrer’s formula.²⁰ In the case of BiPO₄·xH₂O, the calculated crystallite sizes corresponding to the (100) and (110) reflections are 51 and 46 nm, respectively. Similar crystallite sizes were obtained for Bi_0.95Eu_0.05PO₄·xH₂O. Thus, the polyol-mediated synthesis yielded the nanocrystalline hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O since a low temperature was involved in the synthesis.

Fig. 2 Powder XRD patterns of (a) BiPO₄·xH₂O, (b) Bi_0.95Eu_0.05PO₄·xH₂O as obtained by polyol-mediated method and (c) Bi_0.95Eu_0.05PO₄ obtained by calcination of (b) at 600 °C.

Table 1 Calculated lattice parameters of BiPO₄ and Bi_0.95Eu_0.05PO₄

Lattice parameters	BiPO4^a	Bi0.95Eu0.05PO4
		Hexagonal^a	Monoclinic^b
a α = β = 90.00°, γ = 120.00°. b α = 90.00°, β = 103.71°, γ = 90.00°
‘a’ (Å)	6.9841(2)	6.9790(1)	6.7521(2)
‘b’ (Å)	6.9841(2)	6.9790(1)	6.9385(1)
‘c’ (Å)	6.4755(2)	6.4702(2)	6.4690(2)

---

The hexagonal BiPO₄·xH₂O is known to form at room temperature with associated lattice water molecules (x = 0.5–0.67).^9,10 Hexagonal BiPO₄ consists of an open framework structure due to the symmetrical arrangement of chain of alternating BiO₈ polyhedra and PO₄ tetrahedra with zeolitic water molecules located in the channels parallel to ‘c’ axis as shown in Fig. 3a.⁹ In the case of monoclinic BiPO₄, the chains are less symmetrically arranged and hence the structure consists of a compact space-filling network (Fig. 3c). Most of the reports on BiPO₄ deals with the monoclinic phase since the syntheses were performed at relatively high temperatures. In the present study, BiPO₄ was synthesized following the procedure described by Roming and Feldmann.⁵ However, the hexagonal BiPO₄ has been obtained in the present study whereas the monoclinic phase was obtained by Roming and Feldmann. The reason for this could be the difference in the bismuth precursors used in both these studies. BiI₃ was used by Roming and Feldmann whereas Bi(NO₃)₃·5H₂O has been used in the present study. The presence of water molecules in the starting material could lead to the formation of hexagonal BiPO₄·xH₂O. The structurally analogous hexagonal LnPO₄·xH₂O [Ln = La, Ce, Nd, Sm, Eu, Gd and Tb] nanorods were synthesized starting from aqueous solutions of Ln(III) nitrates and NH₄H₂PO₄ under solvothermal conditions in a microwave oven in which the aqueous condition might favor the formation of hexagonal phase.²¹ To confirm the role of water molecules, we carried out the polyol-mediated synthesis in which the Bi(NO₃)₃·5H₂O in DEG was initially heated at 150 °C for 30 min in order to remove the water molecules present in the precursor. Then, the aqueous solution of (NH₄)₂HPO₄ (4 ml) was carefully added to get the final product by the method as described in the experimental section. The powder XRD pattern of this sample (Fig. 4) revealed the presence of monoclinic BiPO₄ as major phase and hexagonal BiPO₄ as minor phase. The presence of additional hexagonal BiPO₄ could be due to the trace amount of water remaining either in the bismuth precursor solution or from the aqueous solution of (NH₄)₂HPO₄. However, the water in the phosphate precursor may not influence the phase formation as Roming and Feldmann used aqueous phosphate precursor in their synthesis of monoclinic BiPO₄:Tb. This result substantiates that the water molecules associated with the bismuth precursor favours the formation of hexagonal BiPO₄ rather than the monoclinic BiPO₄ as obtained with non-aqueous precursor BiI₃.⁵ A similar strategy has been adopted in the direct precipitation of monoclinic LaPO₄.²² The water containing precursor La(NO₃)₃·6H₂O was added to hot H₃PO₄ solution kept at 150 °C in order to avoid the participation of water molecules which would otherwise lead to the formation of hexagonal LaPO₄·0.5H₂O. Another possible reason for the formation of hexagonal BiPO₄ could be the use of acidic Bi(NO₃)₃·5H₂O precursor as the acidic environment favours the formation of hexagonal BiPO₄·xH₂O.⁹

Fig. 3 Crystal structures of (a) hexagonal BiPO₄·xH₂O (with H₂O molecules shown inside the channels) and (c) low-temperature monoclinic BiPO₄ with BiO₈ polyhedra (dark blue) and PO₄ tetrahedra (majenta). The –PO₄ tetrahedron of (b) hexagonal and (d) monoclinic BiPO₄ are shown. Crystal structures are drawn using VESTA program⁴⁰ using the atomic coordinates given in ref. 9.

Fig. 4 Powder XRD pattern of BiPO₄ obtained by polyol-mediated method with heating the precursor Bi(NO₃)₃·5H₂O in DEG at 150 °C to remove H₂O. The pattern contains both monoclinic BiPO₄ (+, major phase) and hexagonal BiPO₄ (* minor phase).

In order to convert the hexagonal Bi_0.95Eu_0.05PO₄·xH₂O into low temperature monoclinic phase and also to remove the zeolitic lattice water molecules and residual DEG molecules, the sample was calcined at 600 °C for 6 h. This temperature was chosen based on the thermogravimetric result that is shown in Fig. 1. In analogues LnPO₄·xH₂O [Ln = La and Ce], the hexagonal to low temperature monoclinic phase transition has been found by in situhigh temperature XRD to occur at about 600 °C.²³Calcination of Bi_0.95Eu_0.05PO₄·xH₂O at 600 °C resulted in the phase transformation from hexagonal to low temperature monoclinic phase of Bi_0.95Eu_0.05PO₄ as evident from the powder XRD (Fig. 2c) and all the reflections were indexed based on the standard pattern (JCPDS 015-0767). The zeolitic water molecules present in the open channels stabilize the hexagonal structure and the removal of which results in the irreversible transformation of the phase to low temperature monoclinic BiPO₄ structure.¹⁰ The calculated lattice parameters (Table 1) of Bi_0.95Eu_0.05PO₄ revealed a marginal decrease when compared to the lattice parameters available with standard pattern of undoped monoclinic BiPO₄ and this is due to the substitution of Bi³⁺ by smaller Eu³⁺ ion. The calculated crystallite sizes of monoclinic Bi_0.95Eu_0.05PO₄ from the FWHM of the reflections (200) and (012) using Scherrer's formula are 48 and 51 nm, respectively. This result shows that the calcination process had no influence on the particle growth which will be discussed in a later section.

3.2 Morphological analysis

The morphologies of undoped and Eu³⁺ doped BiPO₄·xH₂O were analyzed by scanning and transmission electron microscopic techniques and the images are shown in Fig. 5 and 6, respectively. The SEM images of BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O (Fig. 5a and 5b) show that the particles are cocoon-like in shape and are uniformly distributed. The length of each cocoon is of 1 μm and the width is around 500 nm. A similar morphology of hexagonal BiPO₄ was obtained by hydrothermal synthesis using glycerol/water solvent and the size of the nano-cocoons was around 100–150 nm.²⁴ Based on the crystallite size calculated from the XRD results (∼50 nm), it is assumed that each micron sized cocoon is made up of compactly packed smaller primary nanoparticles of the size less than 100 nm. In order to observe this feature clearly, TEM analysis was carried out and from the images (Fig. 6a and 6b) it is clearly seen that the individual particles of sizes less than 50 nm are packed inside the cocoon-like secondary particles. An envelope of organic layer is seen over the particles and this could be the residual surface bound DEG. This result confirms that the primary particles are smaller in size of about 50 nm and correlates well with the crystallite sizes calculated from the powder XRD patterns of BiPO₄ and Bi_0.95Eu_0.05PO₄·xH₂O. The morphology of monoclinic Bi_0.95Eu_0.05PO₄ has changed from the cocoon-like larger particles into clusters of spherical nanoparticles as observed in SEM image (Fig. 5c). Further, the TEM image of monoclinic Bi_0.95Eu_0.05PO₄ (Fig. 6c) shows the random aggregates of nanoparticles with the size ranging from 20 to 100 nm. These results clearly reveal that the effect of calcination (600 °C/6 h) is only on the morphology (arrangement of nanoparticles) and not on the particle size. This could be attributed to the presence of DEG that prevents the particle growth.

Fig. 5 SEM images of hexagonal (a) BiPO₄·xH₂O, (b) Bi_0.95Eu_0.05PO₄·xH₂O as obtained by polyol-mediated method and (c) monoclinic Bi_0.95Eu_0.05PO₄.

Fig. 6 TEM images of hexagonal (a) BiPO₄·xH₂O, (b) Bi_0.95Eu_0.05PO₄·xH₂O as obtained by polyol-mediated method and (c) monoclinic Bi_0.95Eu_0.05PO₄.

3.3 Role of DEG

In the polyol-mediated synthesis of undoped and Eu³⁺-doped BiPO₄ nanoparticles, the DEG acts both as a solvent and a capping agent that prevents the primary nanoparticles to aggregate in a random fashion. DEG assisted in assembling the primary nanoparticles in a uniform manner to form a cocoon-like morphology. To confirm this, a simple precipitation of BiPO₄ was carried out by adding diammonium hydrogen phosphate solution to an aqueous solution of bismuth nitrate at 80 °C to obtain the BiPO₄ particles. The formation of hexagonal BiPO₄·xH₂O was confirmed from the powder XRD result (Fig. S1, ESI)†. The XRD pattern contained sharp and intense reflections revealing a highly crystalline material. The SEM image (Fig. S2, ESI)† showed the presence of large particles (0.2–1.0 μm) with random shapes varying from brick-like to irregular ones. This result confirms the role of DEG as a capping agent in preventing the random aggregation of primary BiPO₄ nanoparticles and limiting the particle growth during the polyol-mediated synthesis. Further, on heat treatment at 600 °C for 6 h, the hexagonal Bi_0.95Eu_0.05PO₄·xH₂O converted into monoclinic Bi_0.95Eu_0.05PO₄ by the removal of zeolitic water molecules as well as the surface bound DEG molecules. The thermal treatment also changed the morphology from the cocoon-like secondary particles into random aggregates of primary nanoparticles (Fig. 5c and more TEM images in Fig. S3 and S4, ESI† depicting the hexagonal Bi_0.95Eu_0.05PO₄·xH₂O and monoclinic Bi_0.95Eu_0.05PO₄ respectively). This result suggests that the removal of surface bound DEG molecules on heating leads to the rupture of cocoons and the spill out of the primary nanoparticles from the cocoons to form the random and non-uniform spherical aggregates as depicted schematically in Fig. 7.

Fig. 7 Schematic illustration showing the conversion of cocoon-like morphology of Bi_0.95Eu_0.05PO₄ stabilized by DEG into random spherical aggregates by heating at 600 °C by loss of DEG.

3.4 Spectral properties

The FT-IR spectra of hexagonal BiPO₄·xH₂O, Bi_0.95Eu_0.05PO₄·xH₂O and monoclinic Bi_0.95Eu_0.05PO₄ are shown in Fig. 8. The absorption band at 1022 cm⁻¹ corresponds to the asymmetric stretching (ν₃) vibration and the bands at 598 and 544 cm⁻¹ are due to the asymmetric bending vibrations (ν₄) of –PO₄group.²⁵ The observed ν₃ absorption bands of hexagonal BiPO₄·xH₂O, Bi_0.95Eu_0.05PO₄·xH₂O (Fig. 8a and 8b) are more symmetric without any splitting when compared to that of monoclinic Bi_0.95Eu_0.05PO₄ in which a splitting into four absorption bands is observed (Fig. 8c). Our result is similar to the one reported in the literature.¹⁰ The –PO₄group in hexagonal BiPO₄·xH₂O has C₂ symmetry with two different P–O bond lengths (1.5406 and 1.5341 Å) and it could be considered to have a pseudo-tetrahedral symmetry.¹⁰ The symmetry of –PO₄group in monoclinic BiPO₄ is C₁ since all the four P–O bonds are different with lengths varying between 1.4684 and 1.5368 Å. This difference in the symmetry of the –PO₄group resulted in the observed difference in the FT-IR spectra of hexagonal and monoclinic Bi_0.95Eu_0.05PO₄. The absorptions at 3450 and 1620 cm⁻¹ are due to the –OH stretching and HOH bending vibrations, respectively, of lattice water in undoped and Eu³⁺-doped BiPO₄·xH₂O.²⁴ The absorption bands observed between 2800 and 3000 cm⁻¹ correspond to the stretching vibrations of C–H bonds of DEG.²⁶ These results indicate that the DEG molecules are adsorbed over the surface of BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O samples and are in agreement with the TGA results. A weak absorption band observed at ∼3400 cm⁻¹ in monoclinic Bi_0.95Eu_0.05PO₄ could be due to adsorbed moisture. Thus, the calcination of hexagonal Bi_0.95Eu_0.05PO₄·xH₂O converted it into monoclinic Bi_0.95Eu_0.05PO₄ with the removal of zeolitic lattice water molecules.

Fig. 8 FT-IR spectra of hexagonal (a) BiPO₄·xH₂O and (b) Bi_0.95Eu_0.05PO₄·xH₂O as obtained by polyol-mediated method; (c) monoclinic Bi_0.95Eu_0.05PO₄ obtained after calcination of (b) at 600 °C.

The laser Raman spectra of hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O and monoclinic Bi_0.95Eu_0.05PO₄ are shown in Fig. 9a–c. Broad bands were observed with the hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O samples and this could be due to the presence of zeolitic water molecules and the residual organics. The Raman spectrum of monoclinic Bi_0.95Eu_0.05PO₄ contains relatively narrow bands which are clearly distinguishable from the hexagonal phase. In all the three spectra, the observed intense line at 170 cm⁻¹ and a shoulder at 231 cm⁻¹ are due to the stretching vibration modes of heavy metal ion, Bi³⁺ (Bi–O), in BiO₈ polyhedron.^27,28 The Peaks at 1035 and 966 cm⁻¹ are due to the asymmetric (ν₃) and symmetric (ν₁) stretching vibrations of the –PO₄group, respectively. The bands in the region between 550 and 600 cm⁻¹ correspond to the ν₄ bending vibration modes of –PO₄groups.¹² The weak bands at 403 and 458 cm⁻¹ can be assigned to the ν₂ bending vibration of –PO₄ units. Though the structural symmetry difference between hexagonal and monoclinic Bi_0.95Eu_0.05PO₄ has been reflected in the FT-IR spectral results, we could not find the difference in the Raman spectra due to the observed broadness in the spectra of both undoped and Eu³⁺-doped hexagonal BiPO₄.

Fig. 9 Laser Raman spectra of hexagonal (a) BiPO₄·xH₂O, (b) Bi_0.95Eu_0.05PO₄·xH₂O and (c) monoclinic Bi_0.95Eu_0.05PO₄.

We, for the first time, report here the ³¹P MAS-NMR spectral results of hexagonal as well as monoclinic BiPO₄ and attempted to reason out the difference observed between them. The ³¹P MAS-NMR spectra of hexagonal BiPO₄·xH₂O, Bi_0.95Eu_0.05PO₄·xH₂O, and monoclinic Bi_0.95Eu_0.05PO₄ are shown in Fig. 10a–c, respectively. Only one line is expected for ³¹P nucleus in a given site since the nuclear spin (I) is ½. The NMR spectra consisted of one major signal flanked by small bands. The small bands were identified as spinning side bands since their positions uniformly shifted when the spectra were recorded under different spinning rates (see Fig. S5 in ESI)†.²⁹ However, the position of the major signal remained unaltered and this result reveals that ³¹P exhibits only one signal in both undoped and Eu³⁺-doped BiPO₄ with hexagonal and monoclinic structures. This result is also in agreement with the crystal structure of BiPO₄ in which the P atom occupies only one crystallographic site.^9,10 The chemical shifts observed with the major line are −7.39 and −7.21 ppm for hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O, respectively. The chemical shift of the major line of monoclinic Bi_0.95Eu_0.05PO₄ is −6.88 ppm. The major difference between the hexagonal and monoclinic Bi_0.95Eu_0.05PO₄ is the broadness in the signal as well as the appearance and intensities of spinning side bands. The ³¹P MAS-NMR signal of monoclinic Bi_0.95Eu_0.05PO₄ is broader than that of hexagonal phase. The relative intensities of spinning side bands are higher with the monoclinic BiPO₄ when compared with the hexagonal phase. In general, the NMR signal in the solid state is broad due to the anisotropic nature in the spin–lattice and spin–spin relaxation processes and the relaxation is predominantly by spin–spin interaction.³⁰ The FWHM values of the main signal of hexagonal Bi_0.95Eu_0.05PO₄·xH₂O and monoclinic Bi_0.95Eu_0.05PO₄ are 754 and 2729 Hz, respectively. This shows that there is a difference in the spin–spin and/or spin–lattice relaxation. The presence of zeolitic water molecules and residual DEG may play a role in the case of hexagonal Bi_0.95Eu_0.05PO₄. The broadening of ³¹P MAS-NMR signal was observed in analogous hexagonal LaPO₄·0.5H₂O during its conversion into anhydrous hexagonal phase and finally to monoclinic LaPO₄ upon heating.²³ This broadening was attributed to the increase in the distribution of –PO₄ tetrahedron local environments in the monoclinic structure. In the present study, the difference in FWHM of NMR signal between hexagonal Bi_0.95Eu_0.05PO₄·xH₂O and monoclinic Bi_0.95Eu_0.05PO₄ could be due to the difference in the crystal structure and symmetry. The internuclear distance between ³¹P nuclei are different in hexagonal (4.1802 Å) and monoclinic (4.0294 and 5.3535 Å) BiPO₄.¹⁰ Hence, there could be anisotropic relaxation of spins in monoclinic BiPO₄ due to the two different P–P distances and this may result in the observed broadening of the NMR signal. The surface bound DEG molecules stabilize the hexagonal Bi_0.95Eu_0.05PO₄·xH₂O nanoparticles in the form of cocoons and due to this the spin–spin relaxation between ³¹P nuclei in nanoparticles of different cocoons becomes less efficient leading to narrow spectrum. On the other hand, in monoclinic Bi_0.95Eu_0.05PO₄, the relaxation between ³¹P nuclei of different nanoparticles among the random aggregates may be more efficient resulting in the broad spectrum. A detailed measurement of the relaxation times and theoretical NMR spectral simulations may provide a better understanding of the observed broadening of signal in the ³¹P MAS-NMR spectra.

Fig. 10 ³¹P MAS-NMR spectra of hexagonal (a) BiPO₄·xH₂O, (b) Bi_0.95Eu_0.05PO₄·xH₂O and (c) monoclinic Bi_0.95Eu_0.05PO₄.

3.5 Optical properties

The optical properties of undoped and Eu³⁺-doped BiPO₄ samples were studied by DRUV-Vis absorption and PL spectroscopic techniques. The DRUV-Vis absorption spectra of hexagonal BiPO₄·xH₂O, Bi_0.95Eu_0.05PO₄·xH₂O, and monoclinic Bi_0.95Eu_0.05PO₄ are shown in Fig. 11. In all the three spectra, the absorption band shows a splitting with the maxima around 225 and 255 nm. Such absorptions in the UV region originate mainly from the optical transitions ¹S₀→¹P₁ and ¹S₀→³P₁ of 6s² lone pair cation Bi³⁺.³¹ From Fig. 11, it is clear that the absorption onset lies in the UV region at ∼320 nm and there is no difference in the absorption onset between hexagonal and monoclinic Bi_0.95Eu_0.05PO₄. The estimated optical band gap [inset (i) in Fig. 11] for BiPO₄ from the Kubelka-Munk function is 3.90 eV which is in close agreement with the literature report (3.85 eV).⁷ A closer look at the absorption spectrum of monoclinic Bi_0.95Eu_0.05PO₄ reveals a shoulder at around 290 nm. This could be due to the Eu³⁺–O²⁻ charge transfer (c.t.) absorption band. In order to find this more clearly, the spectrum was deconvoluted into three bands with the absorption maxima at 222, 255 and 285 nm. This clearly reveals the presence of both Bi³⁺ as well as Eu³⁺ absorption bands which is in good agreement with the analysis by Blasse according to which the optical absorptions of Bi³⁺ (S–P transition) and Eu³⁺ (Eu³⁺–O²⁻ c.t.) occur closely in the UV region.³¹

Fig. 11 Diffuse reflectance UV-Vis absorption spectra of hexagonal (a) BiPO₄·xH₂O, (b) Bi_0.95Eu_0.05PO₄·xH₂O and (c) monoclinic Bi_0.95Eu_0.05PO₄. The insets (i) the Kubelka-Munk function revealing the optical band gap of hexagonal BiPO₄ and (ii) the deconvolution of absorption spectrum of monoclinig Bi_0.95Eu_0.05PO₄ revealing three absorption bands.

In the case of hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O, an enhanced absorption in the visible region from 400 to 700 nm is observed and such absorption is completely absent in monoclinic Bi_0.95Eu_0.05PO₄. This could be due to the presence of residual DEG molecular species over the surface of the phosphate nanoparticles which is also evident from the FT-IR spectroscopic results. There is a possibility of formation of composite of BiPO₄ and organic DEG molecular species through phosphate ester linkage. Recently in the literature, a similar additional broad absorption in the region between 290 and 550 nm has been observed in the DRUV-Vis spectrum of LaPO₄:Eu³⁺ nanorods capped with oleic acid (OA).³² This additional absorption has been attributed to the formation of midgap states due to the chemical bonding between OA and LaPO₄:Eu³⁺. A similar interaction between residual DEG and BiPO₄ may result in such enhanced absorption in the visible region for hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O in the present study. However, the exact reason for the origin of such absorption is not clear at this moment. The calcination step (600 °C/6 h.) involved in the synthesis of monoclinic Bi_0.95Eu_0.05PO₄ resulted in the complete removal of DEG molecules and hence the disappearance of visible light absorption in the DRUV-Vis spectrum of this sample. The presence of visible light absorption in hexagonal Bi_0.95Eu_0.05PO₄·xH₂O may either enhance the PL emission properties through sensitization or diminish the emission from Eu³⁺ by blocking the light absorption by the active center. The influence of such absorption on the PL properties is discussed below.

We recorded the PL excitation and emission spectra of monoclinic Bi_0.95Eu_0.05PO₄ obtained by conventional solid state reaction method as a reference (Fig. S6 in ESI)†. The excitation spectrum showed the high intense peak at 395 nm which is due to ⁷F₀-⁵L₆ absorption of Eu³⁺. We did not observe any emission from the host BiPO₄ and this result is in agreement with the literature report.⁵ The PL emission spectrum of hexagonal Bi_0.95Eu_0.05PO₄·xH₂O as obtained by polyol method is shown in Fig. 12a. The observed emissions of hexagonal Bi_0.95Eu_0.05PO₄·xH₂O were due to different emission transitions originating from ⁵D₀ state (⁵D₀→⁷F_j=0,1,2) of Eu³⁺. However, the emission intensities are very low. The possible reason stems from the DRUV-Vis spectral result in which hexagonal Bi_0.95Eu_0.05PO₄·xH₂O exhibited an enhanced absorption in the visible region when compared to that of monoclinic Bi_0.95Eu_0.05PO₄ . This absorption from the surface bound DEG molecules resulted in the pale yellow color of hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O samples possibly by the formation of a composite of DEG and BiPO₄. Thus, the effective light absorption by Eu³⁺ during excitation may be hindered due to the absorption by surface bound organics which thereby reduces the emission intensity. There are also chances of re-absorption of light emitted by Eu³⁺ in the visible region (550–650 nm) by surface bound DEG molecules resulting in the poor emission intensity. Another possible reason could be the presence of lattice H₂O molecules and the residual DEG in the sample which could quench the PL emission through a non-radiative emissive transition.^33,34 In the case of BiPO₄:Tb³⁺ synthesized by polyol-method using DEG, an enhancement in the emission intensity was observed with increase in the calcination temperature.⁵ We believe that this could be due to the complete removal of the surface bound DEG molecules by calcination. Our results also reveal that the surface bound DEG molecules have a deteriorating effect on the PL emission of Bi_0.95Eu_0.05PO₄·xH₂O despite its role as a stabilizer of the nanoparticles. The PL emission spectrum of monoclinic Bi_0.95Eu_0.05PO₄ obtained by calcination of hexagonal sample is shown in Fig. 12b. We could clearly observe the enhancement in the intensity of orange emission caused by the magnetic dipole transition ⁵D₀→⁷F₁ of Eu³⁺ ion. Due to the difference in the crystal structures and symmetry between hexagonal and monoclinic BiPO₄, a difference in the emission spectral feature for ⁵D₀→⁷F₁ transition of Eu³⁺ was observed.^35–37 A difference in the PL emission properties of the hexagonal and monoclinic EuPO₄ nanoparticles was previously reported in the literature.³⁸ The ratio of ⁵D₀→⁷F₂/⁵D₀→⁷F₁ transitions increased with the loss of water within the hexagonal phase while the emission decreased with the conversion of hexagonal EuPO₄ into monoclinic phase. In the hexagonal Bi_0.95Eu_0.05PO₄·xH₂O, a broad emission line centered at 594 nm with a shoulder at 589 nm is observed for the ⁵D₀→⁷F₁ transition whereas a resolved doublet (at 585 and 594 nm) with equal intensities is observed in monoclinic Bi_0.95Eu_0.05PO_4. The observed Eu³⁺ emission spectrum of monoclinic BiPO₄ is in agreement with the reported one.⁶ It is known that the nanophosphors are much less efficient than the corresponding micron-sized phosphors.³⁹ The organic capping agents stabilize the nanoparticles by preventing the random aggregation of particles and limiting the particle growth. At the same time, these capping agents could create surface trap states that quench the PL emission of the nanophosphors. Alternate organic stabilizers are necessary to realize higher efficiencies with the rare earth based nanophosphors and optimization of particle size is also important in bridging the gap in PL emission efficiency between the nano and bulk phosphors.

Fig. 12 PL emission spectra (λ_exc. = 395 nm) of (a) hexagonal Bi_0.95Eu_0.05PO₄·xH₂O as obtained by polyol-mediated method and (b) monoclinic Bi_0.95Eu_0.05PO₄ obtained by calcination of (a) at 600 °C.

4 Conclusions

Crystalline hexagonal BiPO₄·xH₂O and Bi_0.95Eu_0.05PO₄·xH₂O nanoparticles were synthesized by a polyol-mediated method. The water molecules present in the precursor, Bi(NO₃)₃·5H₂O, favoured the formation of hydrated hexagonal BiPO₄. The hexagonal Bi_0.95Eu_0.05PO₄·xH₂O was converted into monoclinic Bi_0.95Eu_0.05PO₄ with the loss of zeolitic lattice water molecules on heating at 600 °C. The morphological analysis by SEM and TEM showed the size of the primary nanoparticles is much less than 50 nm in the undoped and Eu³⁺-doped hexagonal as well as monoclinic BiPO₄. The primary nanoparticles were stabilized by DEG to arrange themselves into cocoon-like microcrystallites. The DEG molecules could play an important role in retaining the particle size even after annealing at 600 °C at which the phase transition from hexagonal to monoclinic occurs. The crystal structural symmetry difference between the hexagonal and monoclinic BiPO₄ phase is evident from the spectral results of FT-IR, ³¹P MAS-NMR and Eu³⁺ PL spectroscopies. The observed weak PL emission intensity of hexagonal Bi_0.95Eu_0.05PO₄·xH₂O could be due to the presence of zeolitic water as well as surface bound residual DEG molecules which on removal led to the enhancement in the emission intensity along with the phase transition from hexagonal to monoclinic structure. The hexagonal BiPO₄·xH₂O nanoparticles could find potential application in shape selective catalysis due to its open framework structure with hexagonal tunnels.

Acknowledgements

The authors acknowledge the Central Instrumentation Facility (CIF) of CECRI for the characterization facilities. Prof. U.V. Varadaraju and Dr Asiri Naidu of IIT Madras are acknowledged for their help in recording the PL spectra. The structure model was drawn using VESTA program.⁴⁰

References

1. T. S. Chang, G. J. Li, C. H. Shin, Y. K. Lee and S. S. Yun, Catal. Lett., 2000, 68, 229–234.
2. V. G. Alekseev, I. P. Gorelov and M. V. Kornilov, J. Anal. Chem., 2000, 55, 1055–1057.
3. I-S. Cho, J. R. Kim, D. W. Kim and K. S. Hong, J. Electroceram., 2006, 16, 379–383.
4. Z. Holgye and R. Poliak, J. Radioanal. Nucl. Chem., 1991, 153, 267–272.
5. M. Roming and C. Feldmann, J. Mater. Sci., 2009, 44, 1412–1415.
6. M. Guan, J. Sun, F. Tao and Z. Xu, Cryst. Growth Des., 2008, 8, 2694–2697.
7. C. Pan and Y. Zhu, Environ. Sci. Technol., 2010, 44, 5570–5574.
8. C. Pan and Y. Zhu, J. Mater. Chem., 2011, 21, 4235–4241.
9. R. C. L. Mooney-Slater, Z. Kristallogr., 1962, 117, 371–385.
10. B. Romero, S. Bruque, M. A. G. Aranda and J. E. Iglesias, Inorg. Chem., 1994, 33, 1869–1874.
11. Y. F. Lin, H. W. Chang, S. Y. Lu and C. W. Liu, J. Phys. Chem. C, 2007, 111, 18538–18544.
12. J. Geng, W. H. Hou, Y. N. Lv, J. J. Zhu and H. Y. Chen, Inorg. Chem., 2005, 44, 8503–8509.
13. M. Yang, N. K. Shrestha, R. Hahn and P. Schmuki, Electrochem. Solid-State Lett., 2010, 13, C5–C8.
14. C. Feldmann and H. O. Jungk, Angew. Chem., Int. Ed., 2001, 40, 359–362.
15. C. Feldmann, Adv. Funct. Mater., 2003, 13, 101–107.
16. G. Blasse and B. C. Grabmaier, Luminescent Materials, Springer-Verlag, Berlin, 1994.
17. G. Blasse and A. Bril, J. Inorg. Nucl. Chem., 1967, 29, 2231–2241.
18. B. Glorieux, M. Matecki, F. Fayon, J. P. Coutures, S. Palau, A. Douy and G. Peraudeau, J. Nucl. Mater., 2004, 326, 156–162.
19. R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1969, 25, 925–946.
20. A. R. West, Solid State Chemistry and its Applications, John Wiley and Sons, New York, 1984.
21. C. R. Patra, G. Alexandra, S. Patra, D. S. Jacob, A. Gedanken, A. Landau and Y. Gofer, New J. Chem., 2005, 29, 733–739.
22. M. T. Schatzmann, M. L. Mecartney and P. E. D. Morgan, J. Mater. Chem., 2009, 19, 5720–5722.
23. S. Lucas, E. Champion, D. Bernache-Assollant and G. Leroy, J. Solid State Chem., 2004, 177, 1312–1320.
24. F. Xue, H. B. Li, Y. C. Zhu, S. L. Xiong, X. W. Zhang, T. T. Wang, X. Liang and Y. Qian, J. Solid State Chem., 2009, 182, 1396–1400.
25. K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Compounds, Part A, John Wiley and Sons, New York, 1997.
26. S. Takeshita, J. Honda, T. Isobe, T. Sawayama and S. Nikura, Cryst. Growth Des., 2010, 10, 4494–4500.
27. A. Moguš-Milanković, A. Šantić, V. Ličina and D. E. Day, J. Non-Cryst. Solids, 2005, 351, 3235–3245.
28. N. Kitamura, K. Ohno, K. Fukumi, J. Nakamura, T. Hidaka, T. Ikeda, H. Hashima, H. Kozuka and J. Nishii, Jpn. J. Appl. Phys., 2010, 49, 082601–1-082601-4.
29. W. Kemp, Organic Spectroscopy, 3rd Ed., The Macmillan Press Ltd., Hong Kong, 1994.
30. R. S. Macomber, A. Complete Introduction to Modern NMR Spectroscopy, John Wiley & Sons, Inc., New York, 1998.
31. G. Blasse, J. Solid State Chem., 1972, 4, 52–54.
32. L. Li, Y. Su and G. Li, J. Mater. Chem., 2010, 20, 459–465.
33. S. W. Buckner, R. L. Konold and P. A. Jelliss, Chem. Phys. Lett., 2004, 394, 400–404.
34. A. K. Gulnar, V. Sudarsan, R. K. Vatsa, R. C. Hubli, U. K. Gautam, A. Vinu and A. K. Tyagi, Cryst. Growth Des., 2009, 9, 2451–2456.
35. G. Blasse, A. Bril and W. C. Nieuwpoort, J. Phys. Chem. Solids, 1966, 27, 1587–1592.
36. N. Lakshminarasimhan and U. V. Varadaraju, J. Solid State Chem., 2004, 177, 3536–3544.
37. N. Lakshminarasimhan and U. V. Varadaraju, J. Solid State Chem., 2005, 178, 3284–3292.
38. C. Zollfrank, H. Scheel, S. Brungs and P. Greil, Cryst. Growth Des., 2008, 8, 766–770.
39. J. M. Phillips, M. E. Coltrin, M. H. Crawford, A. J. Fischer, M. R. Krames, R. Mueller-Mach, G. O. Mueller, Y. Ohno, L. E. S. Rohwer, J. A. Simmons and J. Y. Tsao, Laser Photonics Rev., 2007, 1, 307–333.
40. K. Momma and F. Izumi, J. Appl. Crystallogr., 2008, 41, 653–658.

---

Footnote

† Electronic Supplementary Information (ESI) available. See DOI: 10.1039/c1ra00389e/

---