Engineering Phase Transitions in Ni-Mn-Sn

Motivation

The central question investigated in my PhD dissertation thesis is whether, through appropriate materials design, a martensitic phase transition can improve the performance of one particular class of "smart materials", magnetocaloric effect materials.

"Smart materials" respond to external stimuli with one or more significant changes in their physical properties or state. While all materials will respond to stimuli to some degree, smart materials are distinguished from ordinary materials by their sizable and controllable responses. Because of these properties, they have found numerous applications as both sensors and actuators. In both cases, the history of smart material research has been defined by a continuous search to find the largest reversible response to a given stimulus. The principle motivation driving this search is the ability of smart materials to 1) replace existing electro-mechanical systems, with the potential for making more scalable, more efficient, and more robust devices, and 2) to develop entirely new functionalities, not previously possible. Successful examples of smart materials that have met these goals include piezoelectric, thermochromic, photochromic, magneto-rheological, and shape-memory materials, all of which have found commercial applications.

Phase transitions can lead to large discrete material responses induced by only small changes in the applied stimulus. Exploiting this property is especially useful when the desired outcome is binary - for example, the turning on and off of a switch, or the storing of a bit of information in the state of a material. Smart materials based on phase transitions essentially represent digital materials, and when properly tuned, can be easily switched back and forth between their two states. Shape memory alloys (SMAs) already employ this property by utilizing a martensitic transition to "reset" the alloy to its original, programmed shape. Potentially, a similar phase transition may result in materials with larger reversible magnetocaloric effects.

Magnetocaloric Effect Materials

Magnetocaloric effect (MCE) materials, first discovered in 1881, undergo a temperature and/or entropy change in response to a change in the external magnetic field. If this response is large enough, it can form the basis of a magnetic refrigeration device. Magnetic refrigeration is inherently more efficient than compressor-based refrigeration systems, as it avoids many of the irreversibilities associated with compressing and expanding a gas or fluid. However, magnetic refrigeration has not become commercially viable because inexpensive high-performance magnetocaloric effect materials have not yet been realized.

The discovery of "giant" MCE materials at the end of the 20th century based on first-order phase transitions led to the development of materials with very large magnetic entropy changes. Despite these advances, practical devices based on first-order phase transitions have thus far eluded researchers due to problems with 1) the reproducibility of the phase transition, 2) the kinetics of the phase transition, and 3) hysteresis losses at the phase transition. In an effort to overcome these challenges, this study investigates the "giant" MCE associated with a thermoelastic martensitic phase transition. This form of transition is generally observed to occur reproducibly, to proceed very rapidly, and to be accompanied by a small thermal hysteresis (< ~20K). However, the specific influence of non-idealities on the MCE in a thermoelastic martensitic system has not yet been investigated.

Objective

The goal of this study is to fully understand the properties of the martensitic phase transition in Ni-Mn-Sn Heusler alloys and its influence on the MCE in this system. This study is not confined to the discovery of materials with large magnetic entropy changes; such studies already exist in abundance. Rather, we question the practical implementation of first-order phase transition materials as magnetic refrigerants: What physical properties define a superior refrigerant? Do first-order phase transitions impose thermodynamic limits on material performance? How do irreversible thermodynamics affect the MCE near a first-order phase transition? The ultimate intent of this work is to suggest general methods of enhancing the MCE in first-order phase transition systems. This thesis will be divided into four principle sections:

I. Structure of non-stoichiometric Ni-Mn-Sn Heusler alloys
The functional properties of Ni-Mn-Sn alloys result directly from a magneto-structural martensitic phase transition. Here, we investigate the equilibrium structure of both the parent and daughter phases, as well as the region of co-stability. This portion of the study seeks to: 1) Determine phase-stability, lattice parameters, and atomic ordering in the ternary Ni-Mn-Sn system, 2) Investigate the phase transition for any evidence of intermediate phases, 3) accurately determine the final modulated martensite structure, and 4) investigate the correlation between changes in atomic structure and magnetic ordering associated with the phase transition.

II. Kinetics of non-stoichiometric Ni-Mn-Sn Heusler alloys
The overall transformation kinetics of martensitic transitions is the subject of numerous studies, but remains a complex and largely unsolved problem. Here, we investigate the kinetics of the forward and reverse martensitic transition in Ni-Mn-Sn and their effect on the MCE. We present isothermal and constant cooling rate transformation behavior consistent with a nucleation-limited martensitic reaction and demonstrate that nucleation is distributed over a range of temperatures due to heterogeneous elastic strain energy in the system. A distributed activation energy kinetic model is proposed which explains the observed quasi-logarithmic isothermal transformation in polycrystalline alloys.

III. Detailed study of hysteresis around the structural phase transition
As one of the major limitations of first-order phase transitions is hysteresis losses, it is important to question how the losses arise, and how they affect the MCE. Here, we investigate the hysteresis associated with the temperature- and magnetic field-induced phase transition. We quantify the hysteresis loss associated with the transformation, and demonstrate the limitations that hysteresis exerts on the extent of transformation (and thus, on the MCE).

IV. Deviation of Bulk Magnetic Property Changes at Low Fields
While the entropy change associated with the structural phase transition in Ni-Mn-Sn is responsible for the primary contribution to the "giant" MCE, other effects that are less susceptible to kinetic and hysteretic losses may also contribute to the total entropy change of the system. Here, we present anomalous behavior of bulk magnetization at low magnetic fields (< 0.2 T) in which the magnetization change is decoupled from the structural phase transition. We demonstrate the lack of hysteresis under these conditions and analyze the resulting MCE of the system at low fields.