Neutron diffraction structure study of borosilicate based matrix glasses

 

M. Fábián1, E. Sváb1, Gy. Mészáros1, L. Kőszegi1, L. Temleitner1, E. Veress 2

 

1 Research Institute for Solid State Physics and Optics, H-1525, Budapest POB 49. Hungary

2 Babeş-Bolyai University, Faculty of Chemistry, 11 Arany János St., RO-3400 Cluj, Romania  

 

 

Borosilicate glasses assure the safe immobilization of most radionuclides even in large quantities, their properties being highly adaptive as concerns the nature, quantity and activity level of the radioactive species present in the waste [1-5]. Therefore the study of borosilicate glasses is of significant current interest as suitable materials for isolating host media for radioactive waste materials.  

We have performed structure investigation by means of neutron diffraction on a newly synthesized borosilicate glass system. The matrix glass with general composition of 65SiO2*25Na2O*5BaO*5B2O3 was doped with ZrO2 to increase the stability, and with CeO2 to simulate  radioactive PuO2. As Ce and Pu coordination is similar in complex oxide environments, it can be expected that Pu coordination will be properly simulated by Ce addition in the host glasses. The samples were prepared by melting in platinum crucible at 1300-1450 ºC, working in atmospheric conditions. The melt was quenched by pouring on an inox plate.  We have investigated two series of samples with the general composition of  (65-x)SiO2.25Na2O.5BaO.5B2O3.xZrO2 and
{(65-x)SiO2.25Na2O.5BaO. 5B2O3.xZrO2 + 10CeO2}  0 ≤ x ≤ 5 (in mole%).

 

Neutron diffraction measurements were carried out at the 10 MW Budapest research reactor using the ‘PSD’ [6] and ‘MTEST’ [7] neutron diffractometers. Powder samples were prepared by milling in an agate mill of the poured and quenched glasses. Despite of the great hydrolytic stability of the samples, the first few experiments revealed their tendency to superficially adsorb H2O. Atmospheric humidity caused a surface swelling (hydrolysis) of the air-kept samples, and the hydrogen contained by the hydrolysed layer produced great incoherent scattering, causing difficulties in the data treatment. The samples were dried at 120 ºC for 4 hours under vacuum conditions, which proved to be completely sufficient to obtain neutron diffraction pattern adequate for data treatment.

All samples from the (65-x)SiO2.25Na2O.5BaO.5B2O3.xZrO2 series were found to be amorphous, while  addition of Ce to the matrix composition lead to partial crystallization of the glass. Amorphous phase could be stabilized by increasing the Zr-content.  Specimens of the (65-x)SiO2.25Na2O.5BaO.5B2O3.xZrO2 + 10CeO2 series were found to be partly crystalline for x ≤ 3, while they were amorphous for x ≥ 4.

 

 

Text Box:  
a/	 
b/	 
c/
 
d/	 
e/	 
f/
Figure 1. Neutron diffraction spectrum and RMC simulation of 62SiO2*25Na2O*5BaO*5B2O3*3ZrO2 glass: 
a/ measured pattern and RMC simulation b/ Si-O c/ B-O d/ O-O e/ O-Na and  f/ O-Zr partial pair correlation functions

The structure factor, S(Q) of the amorphous specimens was determined from the measured pattern using correction and normalization procedure. For data treatment both the traditional direct Fourier-transformation, and the reverse Monte Carlo (RMC) simulation [8] method were applied. For the RMC starting model a disordered atomic configuration was build up. The convergence of the RMC calculation was good inspite of the extremely high number of simulated parameters (about 600), and the final fit matched very well the experimental structure factors as it shown in Figs. 1/a and 2/a, as representatives for a Zr containing sample and for a Zr and Ce containing specimen, respectively. The fit consists of minimizing the squared difference between the experimental and calculated structural factors by moving atoms randomly.

Text Box:  a/         b/     c/

          
                                 d/                                       e/                                       f/                                  g/

Figure 2. Neutron diffraction spectrum and partial pair correlation functions from RMC simulation of 60SiO2*25Na2O*5BaO*5B2O3*5ZrO2+10CeO2  (mol%) glass:
a/ measured S(Q) and RMC simulation b/ O-O c/ O-Na d/ O-Zr e/O-Ce f/Zr-Na and  g/ Zr-Ba

We could successfully calculate the most important partial atomic pair correlation functions, g(r) for all specimens. The different atomic distances of the glass network were calculated, i.e. the first neighbour Si-O and B-O distances at 1.4 and 1.7 Å, respectively The Si-O and B-O pair correlation functions contributing to the 1st coordination shell are shown in Fig. 1/b,c and the O-O, O-Na and O-Zr g(r)'s contributing to the 2nd coordination sphere are collected in Fig. 1/d-f for the 62SiO2*25Na2O*5BaO*5B2O3*3ZrO2 glass. For the 60SiO2*25Na2O*5BaO*5B2O3*5ZrO2+10CeO2  (mol%) sample the results of the RMC simulation are collected in Fig. 2 showing the contributing g(r)'s to the 2nd atomic sphere.

 

From the RMC simulation of the neutron diffraction data we have obtained the partial atomic pair correlation functions for these multi-component glasses, making possible to determine first neighbour atomic distances and coordination numbers. A slight dependence of the parameters on the Zr and Ce concentration was analysed and will be presented. Addition of Zr proved to stabilize the amorphous structure, even Zr can compensate the crystalizing effect of Ce.

It was established that the basic network configuration is the same of the investigated samples, making them suitable for radioactive waste material storage.

 

Acknowledgement

This study was supported by the Hungarian Research Grant OTKA T-42495.

 

1.        Inman, J.M. , Houde-Walter, S.N, McIntyre, B.L., Liao, Z. M., Parker, R.S. and Simmons, V. (1996) Journal of Non-Crystalline  Solids 194, 85

2.     Donald, I.W., Metcalf, B.L. and Taylor, R.N.j.(1997) Journal of Material and Science 32, 5851

3.        Li, H., Vienna, J.D., Hrma, P., Scchweiger, M.J., Smith, D., Gong, M., (1996) Ceramic Transactions, 72, 399-408

4.       Macfarlane, A. (1998) Science & Global Security7, 271-309

5.        Orlhac, X., Fillet.C, and Phalippon , J. (1998) Scientific Basis for Nuclear Waste Manag., 556, 263-270

6.     Sváb, E., Deák, F. and Mészáros, Gy.(1996) Mater. Sci. Forum. 228-231, 247- 252

7.     Kőszegi, L. Journal on Neutron Research (2001) 9, 381-384

8.     McGreevy, R.L. and Pusztai, L. (1988) Molec. Simul. 1, 359-367