One of the key strategies for improving the performance of thermoelectric devices is lowering the thermal conductivity. The thermal conductivity of a material at the nanoscale is very sensitive to the concentration and type of microstructure. Using an integrated suite of atomic-level simulation methods, the thermal conductivity of materials structures involving interfaces, multilayers, polycrystalline microstructures and point defects are analyzed. In particular, by directly analyzing the interactions of phonon wave packets with interfaces, it is possible to begin to understand of phonon-defect interactions and their role on the thermal-transport properties of nanostructured materials. This works was supported by the Center for the Materials Science of Nuclear Fuel, a DOE-BES Energy Frontiers Research Center.

The magnetocaloric effect is the change in temperature of a material caused by an applied magnetic field, and is largest when a phase transition is induced. It has been highlighted as a potential ingredient for efficient solid state cooling [1] and there have been significant advances over the last 10 years that together bring such technology closer to realisation [2,3,4]. The present study is different from the majority of entropy change and magnetometry-based studies already published about such magnetocaloric materials [5,6]. Indeed, the implementation of such materials, particularly in active magnetic regenerators (AMRs) requires knowledge of other physical parameters such as the thermal conductivity, κ, that plays a crucial role in the control of heat flow, itself dominating the efficiency of the cooling cycle [7], making it as important as the thermomagnetic properties of the refrigerant. The present work investigates the tuning of κ, in La0.67Ca0.33MnO3±δ, a material belonging to the manganite family, one of a small number of material families currently being trialled as room temperature magnetic refrigerants [8,9]. The main focus of the work is controlling κ, and thereby managing heat flow to/from/within the material in application. To do so, we examine two routes 1) silver impregnation [10] and 2) copper electrodeposition. We use an effective medium model to extract relevant physical trends in the data [11]. Such thermal management methods represent novel optimization tools for magnetocaloric oxides, or other functional materials that might have a non ideal thermal conductivity. As such this study should be of interest to a broad audience including those interested in the physics of magnetocaloric materials as well as those interested in thermal management by microstructural control and tuning.
References:
[1] G. V. Brown, Journal of Applied Physics 47 (8), 3673 (1976).
[2] K. A. Gschneidner Jr and V. K. Pecharsky, International Journal of Refrigeration 31 (6), 945 (2008).
[3] L. Theil Kuhn, N. Pryds, C. R. H. Bahl, and A. Smith, Journal of Physics: Conference Series 303 (1), 012082 (2011).
[4] Bingfeng Yu, Min Liu, Peter W. Egolf, and Andrej Kitano, International Journal of Refrigeration 33 (6), 1029 (2010).
[5] Brück Ekkes, Journal of Physics D: Applied Physics 38 (23), R381 (2005).
[6] Y. I. Spichkin A. M. Tishin, Materials Today 6 (11), 51 (2003).
[7] K. K. Nielsen and K. Engelbrecht, Journal of Physics D: Applied Physics 45 (2012).
[8] K. A. Gschneidner Jr, V. K. Pecharsky, and A. O. Tsokol, Reports on Progress in Physics 65, 1479 (2005).
[9] A. R. Dinesen, S. Linderoth, and S. Morup, Journal of Physics: Condensed Matter 17 (39), 6257 (2005).
[10] Turcaud J.A., Morrison K., Berenov A., Alford N. McN., Sandeman K. G., and Cohen L. F., arXiv:1210.0410v3 [cond-mat.str-el] (2012).
[11] David S. McLachlan, Michael Blaszkiewicz, and Robert E. Newnham, Journal of the American Ceramic Society 73 (8), 2187 (1990).

Heat and heat flow control are essential for widespread applications of heating, cooling, energy generation and consumption. Here we demonstrate the first solid state thermal thyristor using vanadium dioxide (VO2) beams, in which a thermal terminal actively modulates asymmetric heat flow. In this three terminal device, there are two switchable states, “Rectified” (also “Diode”) state and “Resistor” state, which can be accessed by global heating. In the “Rectification” (“Diode”) state or “on” state, up to 22% thermal rectification is observed. In the “Resistor” state or “off” state, the thermal rectification is suppressed (below 4%). This type of thermal thyristor can have substantial implications ranging from thermal management of micro-/nanoscale devices to thermal energy conversion and storage.

We investigated the electrocaloric (EC) effects in epitaxial Ba0.8Sr0.2TiO3 (BSTO) thin films grown on LAO and MgO substrates as well as polycrystalline BSTO sol gel thin films on Pt/TiO/SiO2/Si substrates. Polycrystalline BSTO sol gel thin films showed a giant EC temperature change with a peak near TC. However, the epitaxial BSTO thin films showed moderate EC temperatures changes with a broader peak extended to higher temperature than TC. The broad ferroelectric phase transition and the resultant wide range of the EC working temperature were explained by the epitaxial strain and grain size effect.

Eu11Cd6Sb12 Zintl phase has attracted our attention for its extremely low thermal conductivity, ranging from 0.8-0.6 W/mK from room temperature to 700K. The exceptionally low thermal conductivity makes this compound a good candidate for thermoelectric applications, both for power generation and refrigeration. An efficient thermoelectric material needs to facilitate an unusual combination of electrical, thermal, and structural properties. We have done some studies on electronic properties of Eu11Cd6Sb12. Seebeck coefficient measurements reveal that Eu11Cd6Sb12 is a p type material with absolute values of 105E-12 V/K at room temperature, suggesting that this compound is a good candidate for thermoelectric applications. Hall effect and electrical resistivity measurement studies elucidate that charge carriers in Eu11Cd6Sb12 are highly mobile. The measured mobility is about 6 times higher than Yb14MnSb11 at room temperature. However, Eu11Cd6Sb12 suffers from low charge carrier concentrations. Since the dominant charge carriers are holes in this system, we are investigating different crystallographic sites for injecting additional holes to optimize the charge carrier concentration. We will report the synthesis, characterization and thermoelectric properties of these new compounds and we will also discuss how different pathways of adding holes to this structure result in different thermoelectric properties.

Much of modern science is devoted to discovering new, better sources of renewable energy. Thermoelectric materials are one such source. Through the Seebeck effect, thermoelectrics allow waste heat to be harvested. Heavy element Zintl phases, like Yb₁₄MnSb₁₁, are promising thermoelectric materials with high figure-of-Merit.¹ Herein, we report synthesis and characterization of new zinc antimonide materials, A₄Zn₉Sb₉ (A = Na, Na/K). The crystal structure of A₄Zn₉Sb₉ consists of a Zn-Sb framework with large open channels hosting sodium and potassium cations. The Zn-Sb framework consist of two main components: slightly puckered square Zn₂As₂ layers similar to the layers present in NaZnSb and BaZn₂Sb₂ compounds,^2,3 and layers containing Zn₄Sb₆ pentagonal prisms, similar to those present in Eu₁₁Zn₆Sb₁₂.⁴ Sodium and potassium jointly occupy several crystallographic positions. To accommodate different sized cations, the Zn-Sb framework is perturbed in such a way that several Zn crystallographic positions are split. Thermoelectric properties of the title compound will be discussed.
1. Brown, S. R.; Kauzlarich, S. M.; Gascoin, F. Yb₁₄MnSb₁₁: New High Efficiency Thermoelectric Material for Power Generation. Chem. Mater.2006, 18, 1873-1877.
2. Jaiganesh, G.; Brito, T. M.; Eithraj, R. D. Electronic and structural properties of NaZnX (X = P, As, Sb): an ab initio study. J. Phys.: Condens. Matter.2008, 20, 085220.
3. Brechtel, E.; Cordier, G.; Schaefer, H. Preparation and crystal structure of barium manganese antimonide (BaMn₂Sb₂), barium zinc antimonide (BaZn₂Sb₂) and barium cadmium antimonide (BaCd₂Sb₂). Z. Naturfor.1979, 34, 921-925.
4. Saparov, B. Bobev, S. Ozbay, A.; Synthesis, structure and physical properties of the new Zintl phases Eu₁₁Zn₆Sb₁₂ and Eu₁₁Cd₆Sb₁₂. J. Solid State Chem. 2008, 181, 2690.

The Specific heat (Cp), thermal expansion (α) and bulk modulus of sodium doped Rare Earth manganites R1-xNaxMnO3 (R3+= La, Pr, Tb) has been studied by means of a Modified Rigid Ion Model (MRIM) and AIM theory. The partial replacement of rare earth cation by sodium introduces large size and charge mismatch at A-site affecting the bulk modulus and thermal properties. Lattice specific heat (Cp) of Pr0.8Na0.2MnO3 as a function of temperature (1K>T> 350K) is found to be in agreement with the published data. The trend of variation of Debye temperature with A-site cationic radius is predicted probably for the first time for the doped rare earth manganites

Thermoelectric devices can recycle waste heat and therefore enhance energy use ratio. They can also generate cool air without discharge of green house air. Type I clathrates have potential to serve as high efficiency thermoelectric materials in thermoelectric devices as its complex structure is considered to be an optimal phonon glass electron crystal (PGEC) and is beneficial for the thermoelectric property. A series of alkali metal containing compound with type I clathrate structure, K8Al8Si38, K8Ga8Si38, Rb8Ga8Si38 and Cs8Ga8Si38, were synthesized. Single crystal diffraction was performed on the series A8Ga8Si38 (A = K, Rb, Cs) and the structure of K8Al8Si38 was confirmed by synchrotron powder diffraction. The samples were consolidated by Spark Plasma Sintering (SPS) and the results of the study on their electrical resistivity and Seebeck coefficient will be presented and discussed.

Solid-state refrigeration based on thermoelectric phenomena is promising, due to its quiet, environmentally friendly, and size independent features. While amount of heat carried by Peltier effect is proportional to the amount of current, current causes internal heat generation by Joule heating, which is proportional to the square of the current. Hence, the optimum current should be determined for either maximum refrigeration effect or minimum temperature of refrigeration chamber. However, most of traditional optimization strategies are based on the assumption that temperature difference across thermoelectric materials is known and kept constant. The temperatures at the both sides of the thermoelectric materials should be determined from the energy conservation equations; amount of heat transferred by Peltier effect, conduction, and Joule heating should be matched with the heat dissipated to surroundings through a heat sink. Therefore, the temperatures depend on not only the amount of current but also thermal resistances at both sides. In this study, the new optimum current as well as optimum geometry of Peltier devices is suggested through rigorous evaluation of temperature difference across thermoelectric materials. Experimental evidence is also provided to verify the proposed model.

Rising demand for air conditioning and refrigeration, together with concern regarding the greenhouse gas potential of conventional HFC refrigerants have combined in recent years to drive research into cooling mechanisms that harness the entropy change of a solid state phase phase transition. The main examples are so-called "magneto-", "electro"-, "baro-" and "elasto-" caloric effects that utilise changes in entropy induced by changes in applied magnetic field, electric field, pressure and strain respectively. Each of these different effects, while perhaps similar at the level of a free energy description, is at a quite different stage of exploration in terms of applied physics, materials science and prototype engineering.
This presentation will review and compare approaches to material and device design used in the four aforementioned solid state cooling areas, in order to examine what lessons may be learned from one area and used to benefit another. Limits on entropy change per unit mass [1] and volume will be used to compare the power density of each technology and suggest routes to improving the properties of the materials in each phase transition class. Such an analysis also allows a comparison of the prospects of solid-state phase transition-based cooling with those of Peltier cooling based on high-efficiency thermoelectrics [2].
The research leading to these results has received funding from the European Community&’s 7th Framework Programme under grant agreement No. 214864 ("SSEEC") [3]. KGS acknowledges financial support from The Royal Society.
References:
[1] K.G. Sandeman, Scripta Materialia 67 (2012) 566-571.
[2] K.G. Sandeman, J. Phys. D, to appear
[3] http://www.sseec.eu

Recently thermoelectrics, which convert thermal energy to electrical energy without producing environmentally detrimental byproducts, do not have moving parts, and are scalable, have gained much attention in energy related applications. In this work we develop an analytical model for thermoelectric devices to explore the influence of the thermal and electrical interface resistances on device performance.
In order for thermoelectrics to be efficient, they need to meet the conditions of high ZT thermoelectric n and p type legs and have a low thermal and electrical resistance at the interfaces between the thermoelectric leg and electrode. Parasite effects at the interfaces can cause energy loss by joule heating and an undesired temperature drop across the interface. Small interface thermal resistance can be obtained by decreasing the lattice mismatch, and by decreasing the acoustic phonon mismatch, etc.
In this work we used analytical modeling to understand and minimize the parasite effects at the interface. An analytical model for thermoelectric devices was developed to explore the influence of interface thermal and electrical resistance on the device performance. The model consisted of oxide films, p type LaNi0.1Co0.9O3 and n type La0.1Ca0.9MnO3, with moderate thickness for high temperature application to maximize the device efficiency and cooling power. For the contact between electrode and thermoelectric legs we used high conductive LaNiO3 and heavily doped CaMnO3 to compensate for the lattice mismatch. To extract both thermal and electrical interface resistances, we used transmission line model. Thermal conductivity was measured by laser flash techniques and electrical resistance was measured by two probe cross plane resistance measurements. The aforementioned model served as a valuable tool to compare the performance of thermoelectric devices with and with out the interface layer.

The discovery of the giant magnetocaloric effect in Gd5Si2Ge2 and other R5T4 compounds (R = rare earth metal and T is a Group 14 element) generated a broad interest in the magnetocaloric effect and magnetic refrigeration near room temperature in particular, and in magnetostructural transitions in general. Reports on the giant magnetocaloric effect in other systems soon followed. These include MnFePxAs1-x and related compounds, La(Fe1-xSix)13 and their hydrides, Mn(AsxSb1-x), CoMnSixGe1-x and related compounds, Ni2MnGa and some closely related Heusler phases, and a few non-metallic systems. A common feature observed in all giant magnetocaloric effect materials is the enhancement of the magnetic entropy effect by the overlapping contribution from the lattice. In addition to the interplay between magnetic and lattice entropies, both of which are intrinsic materials&’ parameters that in principle can be modeled theoretically from first principles, extrinsic parameters such as microstructure and nanostructure, have been found recently to play a role in controlling magnetostructural transition(s) and magnetocaloric effect. Both the intrinsic and extrinsic parameters are, therefore, important in order to have the optimum magnetocaloric effect. The role of different control parameters and the potential pathways towards materials exhibiting large magnetocaloric effect will be discussed.
This work is supported by the U.S. Department of Energy, Office of Science, Materials Science and Engineering Division under contract No. DE-AC02-07CH11358 with Iowa State University.

The Heusler alloys Ni-Mn-X (X = Ga, In, Sn, Sb) have been identified as versatile functional materials. Due to the diffusionless phase transformation from austenite to martensite phase, which can be induced by magnetic field, and magnetic field induced reorientation of martensitic variants, the materials show the magnetic shape memory and the (inverse) magnetocaloric effect (MCE). Hence they are promising materials for magnetocaloric cooling devices. The transformation temperatures sensitively depend on the composition so applications at room temperature are possible. The used elements are cheaper and more environmentally friendly compared to rare-earth or As-based materials.
Due to the high surface-to-volume fraction of thin films a fast heat exchange and a higher frequency of field cycles is possible. Therefore higher cooling power can be achieved using less material compared to bulk cooling elements. The absence of grain boundaries in epitaxial films is exhibited to promote a faster heat transport within the cooling element.
Recently, we prepared epitaxial Ni-Co-Mn-In films by magnetron sputter deposition on single crystalline MgO(100) substrates using a Cr buffer [1,2]. Chromium improves the wetting during preparation and can be used as sacrificial layer for freestanding films [3]. We proved epitaxial growth and the reversible field induced transformation from austenite to modulated martensite. The magnetocaloric effect of Ni Co Mn In films proceeds with an entropy change of 8.8 J/(kgK) at an external magnetic field change of 9 T near room temperature in good agreement with bulk values.
Due to strong Indium evaporation during sputter deposition at higher temperatures [2] and the interdiffusion of Cr into the functional layer we additionally prepared and analyzed epitaxial Co-doped Ni Mn Ga films. For Ni-Mn-Ga-Co bulk an inverse MCE is known in comparison to the direct MCE of Ni-Mn-Ga [4]. We deposited epitaxial Ni-Mn-Ga-Co films by co-sputtering from a Ni-Mn-Ga alloy and a Co-target under similar sputter conditions like Ni-Co-Mn-In films. By adding a few percent of Co the transformation temperatures can be shifted to achieve a maximal MCE at room temperature. The results show that epitaxial Ni-Mn-based Heusler films are promising materials for efficient magnetocaloric cooling devices.
[1] R. Niemann, O. Heczko, L. Schultz and S. Fähler, Appl. Phys. Lett. 97, 2010, 222507
[2] R. Niemann, L. Schultz and S. Fähler, J. Appl. Phys. 111, 2012, 093909
[3] A. Backen, S.R. Yeduru, M. Kohl, S. Baunack, A. Diestel, B. Holzapfel, L. Schultz and S. Fähler, Acta Mater. 58, 2010, 3415
[4] S. Fabbrici, J. Kamarad, Z. Arnold, F. Casoli, A. Paoluzi, F. Bolzoni, R. Cabassi, M. Solzi,
G. Porcari, C. Pernechele and F. Albertini, Acta. Mater. 59, 2011, 412

Recently, MM'X (M, M'=3d transition elements such as Mn, Ni, Fe, Co and X=Al, Si, Ge, P) alloys have attracted much attention due to the unique magnetocaloric effect (MCE) which makes them ideal candidates as magnetic refrigeration materials like Gd₅Si₂Ge₂, MnFeP_0.45As_0.55 and NiCoMnIn. Nowadays, it has been accepted by researchers that anomaly large MCE usually stems from the joint efforts of first-order magnetic transitions (FOMT) coupled with a structure change as the response to the external magnetic field. Mn-Co-Ge is one important member of MM'X family, but FOMT is absent in stoichiometric MnCoGe. Here, substitution of Fe for Mn atoms in Mn-Co-Ge alloys was employed to tailor the magnetic and structural transitions to coincide, leading the FOMT from paramagnetic parent phase to ferromagnetic martensitic phase, which should enable the magnetic field-induced martensitic transformation in this system. The in-situ synchrotron high-energy X-ray diffraction directly evidenced the martensitic transformation when subjected to the external magnetic field, accompanying with a large lattice strain (~12%). The alloy performs giant MCE of ΔS_M=23 J kg^-1 K^-1 for ΔH=5 T in maximum, which is comparable to the well-known giant MCE materials like Gd₅Si₂Ge₂ (-18 J kg^-1 K^-1 for ΔH=5 T) and MnFeP_0.45As_0.55 (-18 J kg^-1 K^-1 for ΔH=5 T). Thus, Fe doped Mn-Co-Ge alloys could be the potential candidates as giant MCE materials used in magnetic refrigeration technique.

We report on a study of the magnetocaloric effect (MCE) in cold-rolled Gd ribbons. The starting material was a 20 mm-thick ingot of 99.9%-pure gadolinium metal. Two representative samples of the rolled material with thickness of 0.215 mm (Sample #1) and 0.036 mm (Sample #2) obtained at different stages of the cold rolling process were taken for the study. Room-temperature X-ray diffraction patterns of the as-prepared ribbons revealed no significant change of structure, lattice parameters, or widths of the diffraction lines in the cold-rolled samples.
Magnetocaloric effect was studied by means of direct measurements of the adiabatic temperature change ΔTad. Heat capacity in zero magnetic field was measured by a commercially available Nertzsch differential scanning calorimeter equipped with specialized software.
The results obtained revealed that the cold rolling depresses magnetocaloric effect. Specifically, in the thinner ribbon (Sample #2) the adiabatic temperature change ΔTad turned out to be twice smaller than in bulk Gd. Moreover, the peak of heat capacity in the Sample #2 was smeared out in the vicinity of Curie temperature. It is suggested that the depression of MCE is connected with a rather high reduction of the magnetization and, as a consequence, the reduction the magnetic part of heat capacity which occurs during severe plastic deformation process. Despite low magnetocaloric effect in the high deformed case it is possible to recover it up to MCE in polycrystalline Gd by sa pecial heat treatment procedure presented in this work.
Authors appreciate RFBR for financial support (grant 12-07-00676-a).

Recent development of many kinds of magnetocaloric materials and prototype equipment for magnetic refrigeration technology has attracted much attention as new avenue for the cooling technology of preventing the global warming. However, in order to put the magnetic refrigeration technology into practical use such as household refrigerator, air-conditioner, industrial freezer and cold storage, many challenges still exist in both refrigeration cycle equipment and magnetocaloric materials. We had started from the basic research on the active magnetic regenerative (AMR) cycle using a primitive apparatus, and had focused on illustrating the potential of generating temperature span and outline of the cooling performance for varying operation conditions. More than 40 degrees C of temperature spans were obtained by operating the AMR cycle using only one kind of ferromagnetic refrigerant of Gd-alloy with spherical shape. Moreover, we adopted a NaZn₁₃-type La(Fe,Si)₁₃ compound formed in spherical shape into the AMR cycle operation using the same apparatus, and obtained the temperature spans of about 20 degrees C [1]. While the entropy change of the La(Fe,Si)₁₃ compound is larger than that of Gd-alloy, the temperature span generated by the La(Fe,Si)₁₃ compound is around half that generated by the Gd-alloy. This is mainly attributable to the large volumetric specific heat of the La(Fe,Si)₁₃ compound. On the other hand, large volumetric specific heat is better for heat regeneration in the AMR cycle, that is, it has an advantage in cooling power of refrigeration. Therefore, La(Fe,Si)₁₃ system is considered as a potential candidate for magnetic refrigerants in practical aspects due to its large entropy change triggered by small magnetic field with little thermal hysteresis, and its large heat regeneration capability, and ease of availability of raw materials such as iron, silicon and lanthanum. Nowadays higher frequency has been progressing in the operation of AMR cycle toward a higher cooling power and smaller size of refrigeration system. Plate shape is preferable to spherical shape of magnetic refrigerants for higher frequency operation of the AMR cycle. In this paper, a processing of La(Fe,Si)₁₃ compound which formed into plate shape will be also presented.
This work is partially supported financially by the Next Program (JSPS).
[1] Fujita, A., Koiwai, S., Fujieda, S., Fukamichi, K., Kobayashi, T., Tsuji, H., Kaji, S.,
Takahashi Saito, A., Jpn. J. Appl. Phys., 2, Let. 46, 8-11, (2007) p. L154.

Large magnetocaloric effect (MCE) owing to occurrence of the first-order phase transition is indispensable in development of the magnetic refrigeration at room temperature, and various materials with the first-order phase transition have been extensively studied. The most popular benchmarks for the MCE are an isothermal entropy change ΔS_M and an adiabatic temperature change ΔT_ad. However, in a general case, they are estimated as static (averaged) values, and influence of the kinetic behavior is not included in the evaluation.
An itinerant electron metamagnetic (IEM) transition causes large MCE in La(Fe,Si)₁₃ and their hydrides¹⁾. Unique feature of this transition is an isotropic change of lattice at the transition, which is not realized in other systems with magnetostructural transition such as Gd₅(GeSi)₄ or Ni-based ferromagnetic shape memory alloys. Especially, shear strain at the phase boundary is involved in temporal evolution of the magnetostructural transition, and the transition velocity strongly depends on the boundary mobility as a function of shear strain. On the other hand, the boundary mobility is mainly determined by release of latent heat generated at the boundary and the elastic term is not significant in the IEM transition.
From measurements of differential scanning calorimetry (DSC) and magnetic susceptibility chi;, signals of La(Fe_0.88Si_0.12)₁₃ under a continuous change of temperature across transition show relatively smooth variation, and Barkhausen-like noises of chi; with small frequency are appeared in the progress of transition. On the other hand, both the DSC and the chi; signals of Ni-Mn-Ga alloy show significant jerky behaviors as reported by the other group²⁾. For Gd₅Ge₂Si₂, the jerky behavior appears in the DSC signal while it is weak in the chi; signal. These differences indicate that impingement on nucleation-growth is mainly magnetic origin in the IEM transition, while that in the magnetostrctural transitions depends on coupling manner of magnetic polarization and structural displacement.
Since the magnetic influence appears in transition progress, the transitions from the paramagnetic to the ferromagnetic (PM-FM) and the reverse transition (FM-PM) show an asymmetric supercooling behavior in thermally induced transition of La(Fe_0.88Si_0.12)₁₃ at the Curie temperature T_C. Small external magnetic fields gradually diminish the asymmetry; therefore, the origin is attributed to magnetic domains in the FM state. This situation is comparable to progress of the electric-field induced transition in Pb(Zn, Ti)O₃ electrocaloric materials³⁾, however, the difference in magnitude of dipole coupling in magnetic and electric cases characterizes their influence on the kinetics.
1) Fujita et al, Phys. Rev. B67, 104416 (2003).
2) Perez-Reche et al, Mater. Sci. Eng. A378, 353 (2004)
3) Karthik and Martin, Appl. Phys. Lett. 99, 032904 (2011)

Fe2P is the parent compound of the giant magnetocaloric MnFe(P,Z) series, and is a weakly itinerant magnetic compound with a peculiar magnetic structure: there are very large magnetic moments of about FeII=2.4mu;B on the 3g site and smaller FeI=0.8mu;B moments on the 3f site. The alloy exhibits a first order ferro- to para-magnetic (FM-PM, Curie) transition at around 212 K [1] without a change in the crystal symmetry [2]. The physical origin of this unusual arrangement of the magnetic moments lies in the hexagonal symmetry (189) of the lattice.
Partial replacement of phosphorus with other p-block elements (B, Si or As) increases the Curie (Tc) temperature sharply while simultaneously changing the Curie transition from first order to second order [3]. The Curie temperature increases drastically with even a small amount of doping: 10% replacement of P by B leads to ~120% change in Tc, while the same amount of Si and As substitution also results in a ~70% or ~60% increase in Tc, respectively.
Furthermore, partial replacement of the 3d element (Fe) by Mn results in a significant increase of the saturation magnetisation value of the alloys. Interestingly, the metamagnetic transition is preserved and it yields a large magnetocaloric effect around room temperature (Dung et al., 2011).
In this work, we use high resolution neutron diffraction (HRPD) to investigate the effect of doping (with boron or carbon) on Fe2P and to identify and explain the exchange striction mechanism in these magnetocaloric alloys. We found a significant contraction of the basal plane on heating through the magnetic transition temperature, with a simultaneous increase of the c-axis resulting in a small overall volume change of ~ 0.1%. The magnetic transition is also apparent in the change of the two shortest Fe-Fe separations. On heating, the FeI-FeI distance, dominated by the a-lattice parameter, drops at Tc and it becomes shorter than the FeI-FeII separation. The shortest metal-metalloid (FeI-PII) distance also shrinks sharply. Detailed density functional theory (DFT) reveals the importance of the latter as the FeI moment, which is strongly de-localized along the 001 (Fe-P chain) direction above Tc becomes localized below Tc. Furthermore, a finite magnetic moment at the strongly magnetic FeII site is retained above Tc as reported experimentally and explained by theoretical calculations [5,6]. In this talk, the effect of doping on these characteristic metal-metal and metal-metalloid distances will be critically discussed.

REFERENCES
[1] Yamada H and Terao K Phase Transitions, 75: 231. (2002)
[2] Fujii H, Hokabe T, Kamigaichi T and Okamoto T J. Phys. Soc. Jpn. 43: 41 (1977).
[3] Chandra R et al. J. Sol. State Chem. 34: 389 (1980) and references therein.
[4] Dung, NH et al. Advanced Energy Materials 1: 1215. (2011)
[5] E. K. Delczeg-Czirjak et al Phys. Rev. B 86, 045126 (2012)
[6] E. K. Delczeg-Czirjak et al. Phys. Rev. B 85, 224435 (2012)

The magnetocaloric effect, i.e., heating or cooling of magnetic material upon magnetic field variation, has been intensively studied in literature[1,2] due to its potential application in magnetic refrigeration. Recent theoretical and experimental works point out that the barocaloric effect (heating or cooling of magnetic material upon pressure variation) is worth of further investigation. [3-5]. It has been experimentally shown that at ambient pressure, the compound Gd₅Si₂Ge₂ undergoes first order transition with giant magnetocaloric effect around room temperature. Moreover, experimental data also show that an applied pressure as large as 5 kbar increases the critical temperature of this compound and keeps the first order phase transition. Experimental data show that the compound Tb₅Si₂Ge₂ undergoes a second order phase transition with a normal magnetocaloric effect around 100 K. In addition, it has been experimentally shown that an applied pressure as large as 10.2 kbar increases the critical temperature of the compound Tb₅Si₂Ge₂ and changes the order of the phase transition from second to first order. In this work, we calculate the magnetocaloric and barocaloric effects in and Gd₅Si₂Ge₂ and Tb₅Si₂Ge₂ . For this purpose, we use a model of localized magnetic moments including the magnetoelastic interaction. In the model, the order of the phase transition is controlled by the ratio between the exchange interaction integral and the magnetoelastic coupling parameter. Our calculations show that these compounds exhibit large values of the entropy changes upon pressure variation in good agreement with the available experimental data. Our theoretical prediction together with the experimental data, open a brand new horizon to build high performance magnetic refrigerators with the barocaloric effect as the working principle.
[1] K. A. Gschneidner Jr, V. K. Percharsky, A. O Tsokol Rep. Prog. Phys. 68, 1479 (2005).
[2] N. A . de Oliveira, P. J. von Ranke, Physics Reports, 489, 89 (2010).
[3] N. A . de Oliveira, Appl. Phys. Lett. 90, 052501 (2007).
[4] L. Manosa et al, Nature Materials, 9, 478 (2010).
[5] S. Yuce et al, Appl. Phys. Lett.101, 071906 (2012).

The magnetocaloric effect (or the magnetic refrigeration) of
typical ferromagnetic materials is simulated by a
molecular-dynamics technique with classical spins.
A spin represents a magnetic atom, and spins are interacting
each other through short-range interactions and long-range
dipole-dipole interactions. Each spin has effective mass
to reproduce relaxation (or damping). Such technique has
been commonly and successfully used in simulations of
ferroelectric materials, but not in those of ferromagnetic ones.

The electrocaloric effect (ECE) is referring to a temperature and/or entropy change of an insulating polar material under application and removing of an electric field. The recent findings of large ECEs in ferroelectric polymers and in ferroelectric ceramic thin films have attracted great interest for developing new cooling devices that are environmental friendly and have the potential to reach better efficiency than the existing vapor-compression approach which employs strong greenhouse gases as the refrigerant. Compared with these solid state ECE materials, a dielectric fluid with a large ECE can be more interesting since it could lead to new cooling cycles with simpler structures and even better performance than these based on solid state ECE materials; for example, they can be utilized as both the refrigerant and heat exchange fluid. Here we show that a large ECE can be realized in the liquid crystal (LC) 5CB near its nematic-isotropic (N-I) phase transition. The LC 5CB possesses a large dielectric anisotropy which facilitates the electric field induced large polarization change from the isotropic phase to the nematic phase near the N-I transition. As a result, a large ECE, i.e., an isothermal entropy change of more than 24 Jkg^-1K^-1 was observed near 39 °C, at temperatures near the N-I transition.

A pyroelectric crystal develops a spontaneous electrical polarization when its temperature changes. It is possible to cycle the temperature and electric field to drive the crystal through an order/disorder phase transition to convert between heat and electric energies. This thermodynamic cycle relies upon the phenomena known as the electrocaloric effect (ECE), in which the temperature changes due to an applied electric field. The ECE is important for future technologies such as solid-state refrigeration.
In this presentation we focus on lead-free perovskite compounds, because lead is an environmental contaminant that is known to strongly effect human development and health. Whereas previous theoretical studies have largely relied on thermodynamic models, for example the Ginzburg-Landau-Devonshire model, in this presentation we calculate the pyroelectric response using a molecular dynamics approach that is parameterized exclusively from ab initio, first-principles methods. In addition to the properties of the bulk compound, we report the impact of epitaxial strain on the ECE.

Materials with large electrocaloric effect (ECE) have the promise of realizing dielectric refrigeration which is more efficient and environmentally friendly compared to other techniques [1,2]. The electrocaloric effect (ECE) in a given material is related to the conversion of the electrical energy into heat and vice versa. It was shown recently by direct measurements that the large ECE is common in polymer and ceramic relaxor ferroelectrics [3,4] and that the electrocaloric responsivity is significantly enhanced in the proximity of the critical point similarly to the enhancement of the giant electromechanical response [5]. A review of recent direct measurements of the large ECE in polymeric and perovskite ceramic relaxor materials including thick ceramic multilayers, substrate-free thick films and thin films will be given. It was found that the giant ECE is also common in these systems. [1] A. S. Mischenko et al., Science 311, 1270 (2006). [2] B. Neese et al., Science 321, 821 (2008). [3] R. Pirc, Z. Kutnjak, R. Blinc, Q.M. Zhang, Appl. Phys. Lett. 98, 021909 (2011). [4] S.-G. Lu, B. Rozic, Q. M. Zhang, Z. Kutnjak, B. Malic, M. Kosec, R. Blinc, R. Pirc, Appl. Phys. Lett. 97, 162904 (2010). [5] Z. Kutnjak et al, Nature 441, 956 (2006).

Recently, solid state refrigeration based on the electrocaloric effect, EC, has attracted a great deal of attention due to the easy generation of large electric fields, high efficiency and relatively low cost of working media. It has been showed theoretically that the electrocaloric effect originates from the entropy changes in the material upon the application of electric field [1] and thus can be related to the degree of structural disorder in the material. In this work the effect of introduced A-site cation disorder (quantified as cation variance [2]) in the perovskite structure on the EC behaviour was evaluated.
A series of perovskites with the general formula Ba_1-x-ySr_xCa_yTiO₃ (0le;xle;0.35, 0le;yle;0.22) were prepared by the solid state reaction. All studied specimens crystallizised in the tetragonal perovskite structure. The dependence of T_C on both the average ionic radii and the cation variance was established. The studied specimens showed typical ferroelectric behaviour with low leakage current.
Molar entropy change, ΔS_mol, during the EC response was measured both directly by the modified DSC technique and indirectly from the P-E loops. A very good agreement between the values of ΔS_mol measured by both techniques was obtained. The EC performance of the studied ceramics was evaluated as a function of temperature, T, and applied electric field, E. The maximum ΔS_mol was observed close to the T_C and increased with the T_C. The EC response depended on the applied electric field as ΔS_mol~Eⁿ. The temperature dependence of n was similar to the one observed in magnetocaloric materials. Universal curves of the temperature dependence of ΔS_mol and n were constructed by analogy with magnetocaloric materials [3]. The parameters of the universal curves depended on the values of T_C and allowed complete description of the EC effect in (Ba,Sr,Ca)TiO3 perovskites. The temperature changes observed in the studied alkaline earth titanates (e.g. 0.92 K at 20 °C and 44 kV/cm for Ba_0.65Sr_0.35TiO₃) suggested that they are very promising EC materials which are environmentally friendly alternatives to current lead containing EC compounds (e.g. PMN-PT [4]).
[1] L. J. Dunne et al., Appl. Phys Lett., 93 (2008) 122906
[2] L.M. Rodriguez-Martinez et al., Phys. Rev. B, 54 (1996) R15622.
[3] V. Franco et al., Int. J. Refrig., 33 (2010) 465
[4] G. Sebald et al., J. Appl. Phys., 12 (2006) 124112.

We will describe the recent advances in thin-film nanoscale and superlattice materials for cooling devices. In particular, we will descibe progress in the growth of thicker (~20 microns) epitaxial Bi2Te3-based superlattice structures, to reduce the electrical and thermal parasitics in a device, that has led us to achieve >60K cooling while maintianing x25 times cooling power density compared to commercial bulk thermoelectric devices. We will also describe the progress in characterization of thermal properties of thicker superlattice films by a variety of techniques. More recently, we have also been exploring the potential poperties and advantages of ultra-thin Bi2Te3 materials for both toplogical insulator behavior and thermoelectrics.

There is no known thermodynamic or other fundamental physical principle that limits the possible values of the thermoelectric figure of merit, ZT. However, thermoelectric performance requires combinations of materials properties that do not normally occur together, for example, high thermopower combined with high conductivity and high carrier mobility with low lattice thermal conductivity. As a result, high thermoelectric performance is typically found not in materials that follow standard text-book behavior, but in materials with unusual features such as highly non-parabolic or other complex band structures, proximity to lattice instabilities, unusual bonding, etc. In this presentation we discuss different ways of obtaining high ZT along with materials examples and also suggest new directions and possible realizations of them.
This work was supported by the Department of Energy through the S3TEC Energy Frontier Research Center.

Traditional thermoelectric Peltier coolers exhibit a cooling limit which is primarily determined by the figure of merit, zT. Rather than a fundamental thermodynamic limit, this bound can be traced to the difficulty of maintaining thermoelectric compatibility. Self-compatibility locally maximizes the cooler&’s coefficient of performance for a given zT and can be achieved by adjusting the relative ratio of the thermoelectric transport properties that make up zT . In this study, we investigate the theoretical performance of thermoelectric coolers that maintain self-compatibility across the device. We find that such a device behaves very differently from a Peltier cooler, and we term self-compatible coolers “Thomson coolers” when the Fourier heat divergence is dominated by the Thomson, as opposed to the Joule, term. A Thomson cooler requires an exponentially rising Seebeck coefficient with increasing temperature, while traditional Peltier coolers, such as those used commercially, have comparatively minimal change in Seebeck coefficient with temperature. When reasonable material property bounds are placed on the thermoelectric leg, the Thomson cooler is predicted to achieve approximately twice the maximum temperature drop of a traditional Peltier cooler with equivalent figure of merit (zT ).We anticipate that the development of Thomson coolers will ultimately lead to solid-state cooling to cryogenic temperatures.

We seek to develop a universal and correlative understanding of effects of point defects on charge transport and heat transfer in crystalline materials.
Native point defects such as vacancies and self-interstitials exist ubiquitously in crystalline materials. They can significantly affect thermal, electrical and thermoelectric properties of the materials, and become critical for improving the performance and thermal management of thin-film devices including thermoelectrics. Experimentally, using high-energy particle irradiation or implantation, point defects can be generated uniformly throughout the film while their concentration depends on the dose of irradiation. It is known, for example, that native defects in semiconductors can be electronically active, acting as donors or acceptors depending on species of the defect as well as conduction/valence band offset of the semiconductor; consequently, irradiation damage makes some semiconductors more conductive, and the others more resistive. In contrast, it is less known how thermal conductivity and thermopower behave in point defects - controlled semiconductors, and whether these defects can be engineered for enhanced device performance. Although these defects are expected to scatter acoustic phonons and thus possibly reduces the thermal conductivity, a predictive model is lacking for a quantitative description. Our work presents a generally applicable defect model by combining experimental exploration and theoretical understanding of all electrical, thermal and thermoelectric aspects in defects controlled semiconductors.

For over 60 years thermoelectric materials have been investigated for use in solid-state refrigeration devices where the only moving parts are electrons and holes. This presentation will review the basics of thermoelectric refrigeration and discuss the current status of materials used for refrigeration near room temperature including various approaches to improving device efficiency. I will then discuss some of our and other groups more recent efforts to push thermoelectric refrigeration to liquid nitrogen temperatures, including some interesting work on doped FeSi and the discovery and explanation of a large (-4500 mu;V/K at 18 K) Seebeck coefficient in pure CrSb2 single crystals. Research was supported by the U. S. Department of Energy, Basic Energy Sciences, Materials Sciences and Engineering Division.

Single crystalline n-type Bi1-xSbx alloys have shown high thermoelectric performance at cryogenic temperatures, especially below 200K, reaching maximum zT of 0.5 when heat and current fluxes are along the trigonal axis. These materials not only are the best cryogenic thermoelectrics available, but they also are tellurium-free, which makes them potentially economically attractive even for room-temperature Peltier cooling. However, single crystalline Bi1-xSbx has mechanical stability issues due to the presence of a cleavage plane that limits its widespread use. In contrast, polycrystalline Bi1-xSbx alloys are easy to synthesize and robust against cleaving. This talk presents investigations made to improve thermoelectric performance of polycrystalline n-type Bi1-xSbx alloys and develop reasonable zT values even at room temperature, although their zT remains lower than that of single crystals. Three techniques are presented. The first is an optimization of the carrier concentration by slight p-type compensation doping. Secondly, Goldsmid [1] recently published under what conditions increasing ionized impurity scattering can improve zT, and this technique is shown appropriate to Bi1-xSbx alloys. Finally, the use of composites is explored to manipulate thermoelectric properties of this system: it is known that a composite of non-interacting systems cannot have zT higher than that of its constituents [2], but here we use inclusions into Bi1-xSbx that affect the properties of host Bi1-xSbx material. Experimental results as well as comparison with theoretical calculations will be presented.

Work supported by AFOSR MURI “Cryogenic Peltier Cooling” Contract #FA9550-10-1-0533

[1] H. J. Goldsmid, “Improving the Power Factor and the Role of Impurity Bands”, presented at the 2012 International Conference on Thermoelectricity, Aalborg, Denmark, proceedings to be published in the Journal of Electronic Materials.
[2] D. J. Bergman and L. G. Fel, J. Appl. Phys. 85 8205 (1999)

We present thermopower and thermal conductivity data taken on a thermocouple in which both arms are made of the same material (n-type InSb) with the same electron concentration (n ~ 4x1014 cm3) , but the phonons have different mean free paths at cryogenic temperatures. This experiment isolates the phonon-drag contribution to the thermopower from the diffusion thermopower. With this, we can experimentally decouple the behavior of the subthermal phonons that drag the electrons, and the thermal phonons that carry most heat. The experiment sheds new light on the details of the physical mechanism behind the giant spin-Seebeck signal recently observed [1] in single crystalline InSb. That signal was attributed to a combination of electron-phonon drag that pushes the electrons, which are spin-polarized by Zeeman splitting, far from thermal equilibrium, and strong spin-orbit interactions that make the Zeeman splitting sensitive to the electron momentum. Furthermore, we may have found experimental clues about the nature of the phonon force. Phonon-drag is classically understood as a momentum transfer between electrons and phonons; the coupling mechanism between electrons and acoustic phonons is also understood to be mediated by the deformation potentials; but not all features of the experiment can be understood with these simple models. Finally, phonon-drag is potentially useful for cryogenic Peltier cooling.
The experiment uses the geometry suggested by Geballe & Hull [2] in 1955; the work is suggested theoretically by Stewart Barnes (University of Miami), and supported by AFOSR MURI “Cryogenic Peltier Cooling” Contract #FA9550-10-1-0533
1. C.M. Jaworski, R.C. Myers, E. Johnston-Halperin and J.P. Heremans, “Giant spin Seebeck effect in a non-magnetic material”, Nature 487, 210-213 (2012)
2. T. H. Geballe and G. W. Hull, Conference de physique des basses temperatures, p460, Paris, 1955

We have demonstrated high coefficient of performance (COP) cooling using superelastic transitions of shape memory alloys. The cooling is based on latent heat of a reversible martensitic transformation. Upon applying a critical stress in the austenite state, the material undergoes a transition to the martensitic state releasing latent heat. Upon unloading, the material undergoes a reverse transition absorbing the latent heat. This cycle can be used to run an efficient heating/cooling cycle. With the latent heat of 12 J/g, NiTi wires display delta T as large as 21 K in the adiabatic limit [1]. The measured material COP of thermoelastic cooling can be as high as 3.7 and 11.8 for tension and compression modes, respectively, provided we can utilize the unloading energy in the cycle. In the compression case, this COP corresponds to up to 83.7% of the theoretical Carnot cycle COP. We will discuss prototype units of thermoelastic coolers we have constructed: a 30W refrigerator and a 1kW air-conditioner based on NiTi wires. [1] Cui et al., Applied Physics Letters 101, 073904 (2012). This project is funded by ARPA-E, ARO, and NSF.