Magnetic investigation of rare earth phosphate glasses

Abstract

Phosphate glasses containing high concentrations of rare earth elements are interesting materials with potentially important applications in laser and optoelectronics technology. In earlier work [1, 2], we used ultrasonic techniques to study the acoustic behaviour of these glasses, which revealed striking variations in their elastic properties. This letter reports on the first magnetic susceptibility and magnetization measurements using portions of the same samples. The samples have the general composition (R203)x(P2Os)l_x, where R = La, Eu, Sm, Nd or Gd. Details of their preparation were given in [3]. We found that a wide range of starting compositions all give a final glass composition near x--0.25, corresponding to the crystalline metaphosphate composition R(PO3)3. The samples weighed between 65 and 290mg, with densities of 3.03.6 g cm -3. Two kinds of magnetic measurements were made using the facilities at the Interdisciplinary Research Centre for Superconductivity at Cambridge. Firstly, a Lake Shore Model 7000 susceptometer was used to measure the ac susceptibility at a field of 3 0 e (240 A m -1) and a frequency of 333 Hz in the temperature range from 2 to 100 K. Checks were carried out to ensure that there were no field or frequency dependencies of the data. On two samples only there were d.c. magnetization measurements at fields up to 10 Oe (800 KAm -1) in the temperature range from 2 to 300 K, which were made using a commercial Quantum SQUID magnetometer in a superconducting magnet. SQUID runs were carried out either at constant field over a range of temperatures, or at constant temperature over a range of fields. In all the measurements the magnetization of the sample container was subtracted from the experimental data. The Nd 3+ and Gd 3+ glasses showed large values of susceptibility with a strong temperature dependence. Several of the samples had an apparent zero offset; that is, the values of susceptibility did not approach zero at high temperatures. This appears to result from an uncorrected experimental artefact, so the curves were shifted by adding or subtracting a constant amount from each measured susceptibility value. Fig. 1 shows susceptibility plotted against the reciprocal temperature for two of the samples. Here