WorksPhysics and Astronomy
On correctly assessing the reversibility of the magnetocaloric effect from indirect measurements
Abstract
The adiabatic temperature change () of a magnetic refrigerant can be indirectly estimated through field () and temperature () dependent magnetization () and specific heat () measurements. A direct integration approach for this estimation is frequently reported, which is an approximation to a rigorous mathematical approach. In this work, we propose an iterative method in small steps, to estimate from indirect measurements. We show that this approach is able to reproduce the reversibility of the magnetocaloric effect, and provides a more accurate estimation of , up to 10% when considering a detailed and dataset that reproduces the magnetothermal properties of gadolinium, a benchmark room-temperature magnetic refrigerant.
Introduction
The magnetocaloric effect (MCE) gives rise to a temperature change in magnetocaloric materials when exposed to a magnetic field, serving as the fundamental principle behind magnetic refrigeration. The MCE is a reversible process, and so, under adiabatic conditions, applying and removing an external magnetic field will keep the system at its starting temperature. The thermodynamics describing the MCE is well established, with over a century of research, and its discovery credited to Weiss and Piccard[1, 2], in 1917.
Magnetic refrigeration benefits from a large MCE which will depend on the magnetocaloric material (refrigerant) used, and it largely dictates the performance of such devices[3, 4, 5, 6, 7]. The study of magnetic refrigeration hinges on accurately gauging the adiabatic temperature change (), either from indirect measurements (the refrigerant’s magnetization and specific heat at different temperatures and external magnetic fields) or by direct measurement of . The accurate direct measurement of is challenging, requiring dedicated equipment. As an alternative, can be estimated using a detailed magnetization and specific heat dataset, both as a function of temperature and magnetic field.
MCE - Thermodynamics
The total differential of the total entropy of a magnetocaloric material can be written as[8],
is the heat capacity under constant magnetic field and pressure, is the magnetization of the material, and is the bulk thermal expansion coefficient.
In an adiabatic-isobaric process ( and ), we can write the infinitesimal temperature change due to the MCE as
This mathematical description is standard, and eq. 2 is frequently seen throughout literature[6, 9, 10]. From here, obtaining diverges into two different branches[7, 10].
Yet, eq. 2 is a total differential equation, which cannot be rigorously solved with neither eq. 3 nor 4, as and are both functions of temperature and magnetic field. Reversing the limits of integration in equation 3 evidently only changes the sign of , as the integral path is the same. So, the direct use of equation 3 results in a non-reversible , an incorrect description of the MCE.
Smith et al. have reported this inaccuracy, proposing the use of eq. 3, but “numerically integrating in sufficiently small increments”, updating on each subsequent integral. Also, Pęcharsky and Gschneidner[10] have criticized the use of eq. 4, suggesting as an alternative calculating from the isentropic difference between the and : .
Correctly assessing the reversible MCE
Instead of successive numerical integration, as suggested by Smith et al., can be estimated by approximating the total differential equation (eq. (1)), by taking small steps in magnetic field (), to which a small temperature change is associated. This methodology is equivalent to a finite difference approach and is grounded in the accurate physical description of the MCE,
Then, letting the temperature evolve by iteration;
until , where is the final magnetic field intensity desired. The adiabatic temperature change from this method is simply the difference between the final and initial temperatures: .
To obtain from and , these thermophysical properties for a magnetocaloric material are necessary and, for the iterative method (eq. (5)), its calculation requires detailed information on and . In this work, we have considered detailed simulated and data that adequately replicate the thermophysical properties of gadolinium, the benchmark material for room-temperature magnetic refrigeration. These were calculated via a hexagonal close packed model lattice of spin 7/2 Ising spins, by Monte Carlo sampling of its Joint Energy and Magnetization dependent Density of States (JDOS)[11]. The nearest-neighbor magnetic exchange parameter was chosen to lead to the experimentally observed value of gadolinium, and the value used was meV. The total specific heat is then the sum of the magnetic specific heat and the lattice contribution described by the Debye model, with a Debye temperature K[12]. Further details on the model, Monte Carlo methodology and comparison with experimental data are available elsewhere[13]. Figure 1 shows and for 5 (out of 102) different external magnetic fields: [0, 0.5, 1, 1.5, 2] (T), for 79 temperature values between 80 and 440 K.

Figure 1. data (top) and data (bottom) for T, of an Ising spin 7/2 HCP lattice, where the lattice contribution to is obtained from the Debye model with K. The dashed line indicates the critical temperature for the phase transition.
Results
Using the iterative method of eq. (5), the adiabatic temperature change was calculated for a field change from 0 to 2 T, using the full dataset of Figure 1. To verify the convergence, a set of field steps were considered, : [0.01, 0.05, 0.025, 0.001] (T). By comparing each associated to a , with the obtained using the smallest field step considered ( T), a maximum relative difference was observed. Also, the from field application and removal describes a reversible process, within a maximum error of .
As shown in Figure 2, the magnitude of from field application, using the iterative method (eq. (5)) is closer to the field application estimate using eq. (3). Still, even assuming this best case scenario comparison, the difference is considerable, with a maximum relative difference larger than 10%.

Figure 2. Top - from field application using and data from simulations that replicate Gd (Figure 1), estimated using eq. (5) (green and blue dots) and eq. (3) (red dots). Bottom - Relative difference of the estimated from eq. (3) and 5, for field application.
Conclusion
The conventional procedure for calculating the adiabatic temperature change () from and data (eq. 3) is shown to not result in a reversible process, contrary to the thermodynamic description of the system. To correctly estimate the adiabatic temperature change, indirectly from and data, preserving the reversibility of the process, we propose approximating the total differential equation by an iterative method of using small steps in magnetic field (eq. 5).
From detailed and data, with comparable magnetothermal properties to gadolinium, we show that estimating from direct integration (eq. 3) leads to a result that is 10% deviated from the more accurate value obtained through the proposed iterative method (eq. 5). This deviation is maximized near the Curie temperature, which is the temperature of interest, as it typically established the working temperature range for a given magnetic refrigeration device.
While the proposed methodology allows for the accurate indirect estimate of the from magnetization and specific heat data, we also encourage the careful use of widely reported approximations.
Acknowledgments
This work was developed within the scope of the project CICECO-Aveiro Institute of Materials, UIDB/50011/2020, UIDP/50011/2020 & LA/P/0006/2020, financed by national funds through the FCT/MCTES (PIDDAC), and projects PTDC/EME-TED/3099/2020, UIDP/04968/2020-Programático, UIDB/04968/2020, NECL-NORTE-010145-FEDER-022096. J.H. Belo acknowledges FCT for contract DL57/2016 reference SFRH-BPD-87430/2012 and R. Almeida acknowledges FCT for PhD. grant reference 2022.13354.BD.
References
- [1]P. Weiss and A. Piccard, “Le phénomène magnétocalorique,” J. Phys. Theor. Appl. 7, 103–109 (1917).
- [2]A. Smith, “Who discovered the magnetocaloric effect?” The European Physical Journal H 38, 507–517 (2013).
- [3]J. Romero Gómez, R. Reguera García, A. De Miguel Catoira, and M. Romero Gómez, “Magnetocaloric effect: A review of the thermodynamic cycles in magnetic refrigeration,” Renewable and Sustainable Energy Reviews 17, 74–82 (2013).DOI
- [4]B. Yu, M. Liu, P. W. Egolf, and A. Kitanovski, “A review of magnetic refrigerator and heat pump prototypes built before the year 2010,” International Journal of Refrigeration 33, 1029–1060 (2010).DOI
- [5]K. A. Gschneidner, Jr. and V. K. Pecharsky, “Thirty years of near room temperature magnetic cooling: Where we are today and future prospects,” International journal of Refrigeration-revue internationale du froid 31, 945–961 (2008).DOI
- [6]V. Franco, J. Blázquez, J. Ipus, J. Law, L. Moreno-Ramírez, and A. Conde, “Magnetocaloric effect: From materials research to refrigeration devices,” Progress in Materials Science 93, 112–232 (2018).
- [7]A. Smith, C. R. Bahl, R. Bjørk, K. Engelbrecht, K. K. Nielsen, and N. Pryds, “Materials challenges for high performance magnetocaloric refrigeration devices,” Advanced Energy Materials 2, 1288–1318 (2012).DOI
- [8]A. Tishin and Y. Spichkin, The Magnetocaloric Effect and its Applications (CRC Press, 2003).
- [9]A. Tishin, “Magnetocaloric effect: Current situation and future trends,” Journal of Magnetism and Magnetic Materials 316, 351–357 (2007).
- [10]V. K. Pecharsky and J. Gschneidner, K. A., “Magnetocaloric effect from indirect measurements: Magnetization and heat capacity,” Journal of Applied Physics 86, 565–575 (1999).
- [11]J. C. Inácio, A. L. Ferreira, and J. S. Amaral, “Accurate estimate of the joint density of states via flat scan sampling,” (2022), arXiv:2203.02718 [cond-mat.stat-mech].arxiv.org/abs/2203.02718
- [12]R. W. Hill, S. J. Collocott, K. A. G. Jr, and F. A. Schmidt, “The heat capacity of high-purity gadolinium from 0.5 to 4 k and the effects of interstitial impurities,” Journal of Physics F: Metal Physics 17, 1867 (1987).DOI
- [13]R. Almeida, S. C. Freitas, C. R. Fernandes, R. Kiefe, J. P. Araújo, J. S. Amaral, J. O. Ventura, J. H. Belo, and D. J. Silva, “Rotating magnetocaloric effect in polycrystals—harnessing the demagnetizing effect,” Journal of Physics: Energy 6, 015020 (2024).DOI