Critical examination of heat capacity measurements made on a Quantum Design physical property measurement system

We examine the operation and performance of an automated heat-capacity measurement system manufactured by Quantum Design (QD). QD's physical properties measurement system (PPMS) employs a thermal-relaxation calorimeter that operates in the temperature range of 1.8-395 K. The accuracy of the PPMS specific-heat data is determined here by comparing data measured on copper and synthetic sapphire samples with standard literature values. The system exhibits an overall accuracy of better than 1% for temperatures between 100 and 300 K, while the accuracy diminishes at lower temperatures. These data confirm that the system operates within the +/-5% accuracy specified by QD. Measurements on gold samples with masses of 4.5 and 88 mg indicate that accuracy of +/-3% or better can be achieved below 4 K by using samples with heat capacities that are half or greater than the calorimeter addenda heat capacity. The ability of a PPMS calorimeter to accurately measure sharp features in C-p(T) near phase transitions is determined by measuring the specific heat in the vicinity of the first-order antiferromagnetic transition in Sm2IrIn8 (T-0 = 14 K) and the second-order hidden order (HO) transition in URu2Si2 (T-N = 17 K). While the PPMS measures C-p(T) near the second-order transition accurately, it is unable to do so in the vicinity of the first-order transition. We show that the specific heat near a first-order transition can be determined from the PPMS-measured decay curves by using an alternate analytical approach. This correction is required because the latent heat liberated/absorbed at the transition results in temperature-decay curves that cannot be described by a single relaxation time constant. Lastly, we test the ability of the PPMS to measure the specific heat of (MgB2)-B-11, a superconductor of current interest to many research groups, that has an unusually strong field-dependent specific heat in the mixed state. At the critical temperature the discontinuity in the specific heat is nearly 15% lower than measurements made on the same sample using a semi-adiabatic calorimeter at Lawrence Berkeley National Laboratory. (C) 2003 Elsevier Science Ltd. All rights reserved.