1 Introduction

Thorium phosphate–diphosphate (TPD), Th₄(PO₄)₄P₂O₇ in ceramic form, appears as a very nice host matrix for the immobilisation of long-lived radionuclides needed in new ways of waste management 〚1–3〛. This ceramics is obtained from powder, synthesised either by evaporation 〚4〛 of a mixture of solutions containing the components (thorium nitrate or chloride and phosphoric acid) or from a precursor compound, the thorium phosphate–hydrogenphosphate (amorphous Th₂(PO₄)₂HPO₄·(3–7) H₂O 〚5〛 or crystallised Th₂(PO₄)₂HPO₄·H₂O) 〚6〛). Both methods lead to the TPD ceramics by pressing the powder at room temperature then heating at 1150–1250 °C 〚7〛. The main qualities of this material are: high resistance to water corrosion 〚3, 8, 9〛, thermal and chemical stabilities, high density 〚7〛.

As a host matrix for radwaste, the TPD can be used to:

• • incorporate tetravalent actinides; this leads to solid solutions with a maximum value in mol% of 75 of uranium, 52 of neptunium, 41 of plutonium; the system remains as a single phase and most of the pure TPD properties are held for these solid solutions 〚3, 10–12〛;
• • incorporate trivalent cations such lanthanides and some actinides (Am, Cm); this leads to TPD ‘doped’ crystal structure that can only accommodate small amounts of these ions, less than 0.5 mol% 〚11〛.

Larger amounts form in the matrix a second phosphate phase MPO₄ (monazite or xenotime structure, depending on the cation). Zirconium, and probably other tetravalent cations of small radii, form also separated phases of Zr₂O(PO₄)₂ and/or ZrP₂O₇ 〚11〛.

The TPD can be also ‘loaded’ with other phosphates of mono- and divalent cations, whose solubility is as low as that of the TPD. The solids obtained are polyphase systems. Such solids were already prepared, using 23 cations simulating fission products: caesium, calcium, cadmium, strontium, barium, lanthanides, and zirconium (6–7 wt% of total amount of all cations) 〚13〛. In these very complex systems, it is difficult to determine the chemical composition of each phase; therefore, we chose to study some of them containing monovalent and divalent cations (as the behaviour of the trivalent and tetravalent ones was already studied) frequently present, in liquid radioactive waste.

Since the 1960s', the literature gives several series of metal–thorium phosphates of general formulae M^ITh₂(PO₄)₃ (M^I = alkali metals, Ag, Cu, Tl), M^IITh(PO₄)₂ 〚14–21〛, (M^II = Ca, Cd, Sr, Pb) 〚22〛, ${\mathrm{M}}_{0.5}^{\mathrm{II}}{\mathrm{Th}}_2{\left({\mathrm{PO}}_4\right)}_3$ (M^II = Ca, Sr, Ba, Cd, Pb) 〚23〛 and ${\mathrm{M}}_7^{\mathrm{II}}\mathrm{Th}{\left({\mathrm{PO}}_4\right)}_6$ (M^II = Ba, Pb, Sr; JCPDS files: 33-180, 33-775, 33-1354) 〚24〛. All these phosphates were obtained from mixture of thorium dioxide ThO₂, metal carbonate M₂CO₃ or metal oxide MO and ammonium dihydrogenphosphate NH₄H₂PO₄ using dry chemistry methods.

These compounds can be also obtained by soft chemistry by a co-precipitation process at pH > 5 〚13〛 or by evaporation of acidic solutions of the constituents. The products obtained by these ways have the same characteristics, after heat treatment, as those synthesised using the dry chemistry process.

In order to obtain the metal–thorium phosphates, we applied the evaporation method of acidic solutions since the radioactive wastes are usually stored in 1–2 M nitric acid. Synthesis of such compounds as M^ITh₂(PO₄)₃ (M^I = Na, Cs) and M^IITh(PO₄)₂ (M^II = Ba, Ca, Cd, Sr) were attempted. Furthermore, in the liquid waste, several anions are often present. In order to check whether some anions form phosphate compounds of the Th_mA(PO₄)_m type (m = 1, 2, 4), syntheses with anions A = F^–, SO₄^2–, SiO₄^4–were performed by soft chemistry methods.

2 Experimental section

Thorium chloride 2 M solution was from Rhône-Poulenc (France). Other chemical reagents were supplied by Aldrich-Fluka. For hydrothermal syntheses, Parr Instrument Company autoclaves supported with poly-tetrafluoroethylene (PTFE) containers (maximum volume 23 ml) were used. Samples were heated at 150 °C in order to increase the kinetics of the reaction of synthesis. In these conditions, the vapour pressure cannot be controlled but, depending on the temperature, may rise up to 300 bar.

The X-ray powder diffraction (XRD) data were collected with a Philips PW 1050/70 diffractometer using the monochromatic Cu Kα radiation and a nickel filter. Patterns were recorded form 10 to 60° (2 θ), with a step of 0.01°, using silicon powder as internal reference.

The infrared (IR) spectra were recorded from 4000 cm^–1 to 400 cm^–1 with a Hitachi I-2001 spectrophotometer. Samples of 1–1.5 wt% of the solids were prepared in KBr pellets.

The heating treatment of the samples (with a rate of 5 °C min^–1) was performed in alumina boats as well as in a Pyrox MDB15 furnace or in an ADAMEL FR20 one.

Electron probe microanalyses (EPMA) were carried out by means of a CAMECA SX 50 apparatus usually working with a 15 kV voltage and a 10 nA current beam. ThO₂ (Mα ray of thorium) and KTi₂(PO₄)₃ (Kα ray of phosphorous) were used as calibration references.

The elementary analysis was performed using the Particle Induced X-ray Emission (PIXE) method. The element amounts were calculated from the surface of the bands.

3 Results and discussion

3.1 Thorium fluoride phosphate

From the mixture containing Th⁴⁺–H₃PO₄–HF components, a white crystallised powder was obtained. The PIXE and EPMA results and the molar ratios of this product are given in Table 1. As it can be seen from this table, the experimental results for thorium and phosphorous amounts are in good agreement with the expected ones. The fluoride content is lower that the calculated value. This is due to the EPMA method used, which never gives results with good accuracy for atoms of atomic number below 10. However, the global formula of the thorium fluoride phosphate may be proposed as ThFPO₄·H₂O. Its XRD diagram and powder diffraction data are presented, respectively, in Fig. 1a and Table 2. This diagram is similar to that of uranium (IV) fluoride phosphate UFPO₄·H₂O (JCDPS file 13-0047) 〚25〛. Beside the thorium fluoride phosphate formed, a minor phase with a molar ratio Th:F:O = 1:2:1 in the solid appears from the EPMA results. This ratio could correspond to the thorium oxide fluoride ThOF₂. Nevertheless, in the XRD diagram of this phase, no lines could be assigned, neither to the orthorhombic form (JCPDS 12-0096) nor to the hexagonal one (JCPDS 39-0620) of the ThOF₂.

Table 1

Results of elementary analysis of the synthesised compounds.

	Elements	Weight %	Molar ratios
	experimental	EPMA	PIXE
		EPMA	PIXE	calculated
a	Th	65.9 ± 3.0	63.6 ± 0.9	63.74	—	—
P	8.68 ± 0.35	8.4 ± 0.5	8.51	$\frac{\mathrm{P}}{\mathrm{Th}}=0.99$	$\frac{\mathrm{P}}{\mathrm{Th}}=1.01$
	F	4.26 ± 0.71	—	5.25	$\frac{\mathrm{F}}{\mathrm{Th}}=0.79$	—
b	Th	—	59.3 ± 0.5	58.89	—	—
P	—	7.7 ± 0.3	7.86	—	$\frac{\mathrm{P}}{\mathrm{Th}}=0.97$
	S	—	4.2 ± 0.3	4.07	—	$\frac{\mathrm{P}}{\mathrm{Th}}=0.51$
c	Th	65.7 ± 2.16	65.5 ± 0.5	66.29	—	—
P	9.24 ± 0.84	9.0 ± 0.2	8.85	$\frac{\mathrm{P}}{\mathrm{Th}}=1.05$	$\frac{\mathrm{P}}{\mathrm{Th}}=1.03$
	Si	1.87 ± 0.59	2.0 ± 0.9	2.01	$\frac{\mathrm{P}}{\mathrm{Si}}=0.24$	$\frac{\mathrm{P}}{\mathrm{Si}}=0.26$
d	Th	52.0 ± 2.5	—	52.62	—	—
P	10.18 ± 0.61	—	10.54	$\frac{\mathrm{P}}{\mathrm{Th}}=1.47$	—
	Cs	15.21 ± 0.42	—	15.07	$\frac{\mathrm{Cs}}{\mathrm{Th}}=0.51$	—
e	Th	40.2 ± 0.8	—	41.50	—	—
P	10.77 ± 0.56	—	11.08	$\frac{\mathrm{P}}{\mathrm{Th}}=2.0$	—
	Ba	26.52 ± 0.52	—	24.52	$\frac{\mathrm{Ba}}{\mathrm{Th}}=1.12$	—

Table 2

Main lines of XRD powder data of the synthesised thorium phosphates.

ThFPO4·H2O	Th4(PO4)4SiO4	CsTh2(PO4)3	BaTh(PO4)2
2 θexp (°)	I/I0	2 θexp (°)	I/I0	2 θexp (°)	I/I0	2 θexp (°)	I/I0
12.85	9	10.76	3	13.63	20	14.17	100
14.01	51	11.26	10	16.15	4	15.97	6
15.12	7	14.31	3	16.90	22	17.06	4
17.59	49	16.09	19	18.29	39	17.80	11
18.05	49	16.89	100	19.86	79	19.46	6
19.20	42	18.20	4	20.43	13	20.93	51
21.08	22	20.42	45	22.30	8	20.33	24
22.39	100	22.35	6	25.23	30	23.89	7
24.01	20	25.10	15	25.52	18	25.15	6
24.10	30	25.65	7	26.67	78	26.28	48
24.56	15	26.44	25	27.28	25	27.05	84
26.40	19	27.44	9	28.87	100	28.54	51
26.98	22	29.08	17	30.72	50	30.52	9
28.84	69	30.03	9	31.84	6	32.85	48
30.20	28	30.47	41	32.55	21	34.34	4
31.88	57	31.78	41	33.63	7	36.56	4
35.36	37	32.41	4	35.41	5	38.15	5
36.86	9	34.28	16	37.27	8	39.09	7
37.83	9	35.05	16	38.30	6	39.64	14
40.20	8	36.81	2	40.28	27	42.73	29
41.48	37	37.70	2	41.11	26	43.45	20
41.95	28	38.97	7	42.27	17	43.70	21
42.93	29	41.71	8	42.68	10	44.80	17
43.39	12	43.04	16	44.24	8	47.81	3
44.13	16	43.93	3	45.67	6	49.28	3
45.31	14	45.67	5	45.95	8	49.77	4
45.83	18	46.52	12	46.60	18	51.09	3
46.63	19	47.56	20	46.93	18	51.90	8
46.96	12	48.78	8	47.55	4	52.81	8
47.02	10	49.90	5	48.52	28	52.88	8
47.62	11	50.41	5	49.86	10	54.26	10
49.92	10	50.47	5	51.27	5	55.60	8
50.63	19	51.09	8	52.31	20	55.71	8
51.60	23	51.75	4	53.50	12	57.25	5
52.42	10	52.78	7	53.68	9	58.66	4
54.04	7	54.82	6	54.41	9	59.21	7
54.41	10	56.66	6	55.07	6
56.99	14	58.96	3	57.12	10
57.92	10			59.77	8

In the IR spectrum in Fig. 2a, the bands can be assigned to the stretching and deformation modes of the P–O bond (1140–400 cm^–1), to the stretching of O–H bond (3600–3400 cm^–1) and the bending mode (1610 cm^–1) of the water molecule 〚26–28〛. These results are consistent with the hydrated formula of the synthesised compound. On the other hand, it is impossible to define the frequencies of Th–F bond, since the space group of the thorium fluoride phosphate is unknown. Furthermore, it cannot be compared to that of UFPO₄·H₂O, because its structure is also unknown.

When heated at 650–700 °C, the ThFPO₄·H₂O is transformed into dithorium oxide phosphate Th₂O(PO₄)₂ 〚6〛. This conversion let us presume that the first stage is a hydrolysis reaction (equation (1)) leading to the formation of thorium hydroxide phosphate, followed by a condensation process (equation (2)):

$${\mathrm{ThFPO}}_4.{\mathrm{H}}_2\mathrm{O}\to \mathrm{Th}\left(\mathrm{OH}\right){\mathrm{PO}}_4+\mathrm{HF}\uparrow$$

1

$$2\mathrm{Th}\left(\mathrm{OH}\right){\mathrm{PO}}_4\to {\mathrm{Th}}_2\mathrm{O}{\left({\mathrm{PO}}_4\right)}_2+{\mathrm{H}}_2\mathrm{O}\uparrow$$

2

3.4 Caesium dithorium phosphate and barium thorium phosphate

In the system Cs⁺–Th⁴⁺–H₃PO₄, after evaporation of the volatile products and then after heating at 1100 °C, a well-crystallised residue was obtained. As it was shown from the EPMA results, the solid appears as single-phased. The elementary weight percent and molar ratio of the elements are presented in Table 1. This compound, caesium–dithorium phosphate, CsTh₂(PO₄)₃ was obtained by Matković et al. 〚19〛 using the dry chemistry method, and crystallises in the monoclinic system (symmetry group C⁶_2h, Z = 4) with the following unit cell parameters: a = 1.773 nm, b = 0.688 nm, c = 0.817 nm, β = 102°06’ and the density d = 5.91. Neither XRD diagram nor powder diffraction data were published by the above-mentioned authors. They are shown in Fig. 3a and listed in Table 2. The same kind of diagram was obtained by heating caesium carbonate with thorium dioxide and ammonium dihydrogenphosphate 〚30〛.

The product obtained from the mixture Ba²⁺–Th⁴⁺–H₃PO₄ is also single-phased. The molar ratio obtained for the analysed samples is in good agreement with the expected one (Table 1). Thus, the proposed formula of this barium–thorium phosphate can be written as BaTh(PO₄)₂. The XRD diagram, presented in Fig. 3b, is different from those of the monoclinic compounds M^IITh(PO₄)₂ (M^II = Ca, Cd, Pb, Sr) brabantite type and of the cubic Ba₇Th(PO₄)₆. The systems M²⁺–Th⁴⁺–H₃PO₄ (M²⁺ = Ca, Cd, Sr) were also tested in the same conditions as the barium–thorium phosphate system, i.e. with the soft chemistry method. Their XRD powder diagrams are consistent with those of Schwarz 〚22〛.

4 Conclusions

Using hydrothermal conditions, a partially crystallised compound, thorium fluoride phosphate ThFPO₄·H₂O was synthesised. This compound is unstable at 650–700 °C and is transformed into the dithorium oxide phosphate, the volatile products being hydrogen fluoride and water.

The amorphous product supposed to be thorium sulphate phosphate Th₂(PO₄)₂SO₄·2 H₂O is also unstable, even at 450–500 °C.

Thorium silicate phosphate, Th₄(PO₄)₄SiO₄, caesium–dithorium phosphate CsTh₂(PO₄)₃ and barium-thorium phosphate, BaTh(PO₄)₂ were obtained as well crystallised single phases, after evaporation of the initial mixture of the appropriate solutions, followed by heating at high temperatures (1100–1150 °C). From the XRD diagram, the first one resembles the TPD, and the second is consistent with that synthesised by other authors. The third compound seems to be different from the monoclinic M^II–thorium phosphates (M^II = Ca, Cd, Sr, Pb) as well as from the cubic Ba₇Th(PO₄)₆.

It is interesting to compare the crystallographic systems of metal(II)–thorium phosphates in terms of ionic radii of the mentioned divalent cations (Table 3). Cations whose radii, in eight-fold coordination 〚31〛 (this coordination number was chosen because it is common to all cations below), spread from 0.110 to 0.129 nm (Cd²⁺, Ca²⁺, Sr²⁺, Pb²⁺) have M^IITh(PO₄)₂ as chemical formula and form a monoclinic lattice. The lead–thorium and strontium–thorium phosphates are polymorphous. The Sr²⁺ and Pb²⁺ cations with the radii 0.126 and 0.129 nm respectively form also, like Ba²⁺ (0.142 nm), a cubic lattice of the ${\mathrm{M}}_7^{\mathrm{II}}\mathrm{Th}{\left({\mathrm{PO}}_4\right)}_6$ type. The following question is still open: which is the crystallographic system of the barium–thorium phosphate BaTh(PO₄)₂?

In the series of M^II–thorium phosphates, two of them are missing, the mercury–thorium and the radium–thorium phosphates. No information about these compounds was found in the literature. Taking into account the mercury ionic radius ^〚8〛r = 0.114 nm 〚31〛 the supposed ‘HgTh(PO₄)₂’ system should have a monoclinic lattice, while the ‘RaTh(PO₄)₂’ system (^〚8〛r = 0.148 nm 〚31〛) should be similar to the BaTh(PO₄)₂ one.

With regard to thorium phosphate–diphosphate, it appears from the present results that monovalent and divalent cations such as Cs⁺, Ba²⁺, Ca²⁺, Cd²⁺, Sr²⁺ loaded into the TDP will form a polyphased system: TPD–CsTh₂(PO₄)₃–M^IITh(PO₄)₂. In the presence of minute amounts of anions such as silicate (or fluoride and sulphate), separate phases will also be present in the resulting solid during its synthesis. These experiments, as well as previous ones 〚11〛, show that the TPD can be a host matrix not only for tetravalent actinides 〚10–12〛 but also for other radionuclides, like fission products 〚13〛. The resistance to water corrosion of these systems is under investigation.

Acknowledgements

The authors would like to thank R. Podor from the University of Nancy (France) for the EPMA analyses and to G. Lagarde from the ‘Institut de Physique Nucléaire’ (Orsay, France) for the PIXE analyses.