Spectrally-selective solar absorbers are widely used in solar hot water and concentrating solar power (CSP) systems. However, the performance at high temperatures (>500 ^oC) can be further improved. Recent progress on cermet-based solar absorbers has shown promising high temperature thermal stability and wavelength selectivity. Here we explore W-Ni-Al₂O_3, W-Ni-YSZ (yttria-stabilized-zirconia), and W-Ni-SiO₂ cermet based spectrally selective surfaces for high-temperature solar absorber applications. The developed multilayer selective surfaces are deposited by magnetron sputtering on different substrates depending on applications. The absorber consists of two solar absorbing cermet layers with different W-Ni volume fraction inside the dielectric matrix, one or two anti-reflection coatings (ARCs), and one tungsten IR reflection layer for reduced IR emittance and improved thermal stability. All these absorbers show an absortance of > 90% for temperature up to 500 °C and emittance of ~5% at about room temperature and 10-15% at 500 °C.

Recently we are developing a new kind of absorber that reflects a certain range of wavelength and absorbs the rest of the whole solar spectra. The absorbed part is used for electrical power generation by steam engine and the reflected part is used for solar photovoltaic conversion. The thermal energy can be easily stored for later conversion to provide electrical power around the clock without worrying the Sun’s night time.

Implementing larger and faster recording capacities, like in Heat Assisted Magnetic Recording (HAMR) devices, requires investigating the magnetization dynamics of nanostructures at the sub-picosecond time scale. The case of L1₀ FePt “nanocrystals” is of particular interest as HAMR media can be designed with grain diameter below today’s D ~ 8 nm. The magnetic anisotropy is sufficiently high and results in coercive fields larger than 5 Tesla at room temperature [1, 2].

Femtosecond magneto-optics allows investigating the dynamical properties of such films [3] and nanoparticles [4] with a temporal resolution well adapted to the actual needs of performant materials that can be addressed in the time scale of a few picoseconds or faster. In the case of materials for HAMR, the pre-heating with femtosecond laser pulses allows reaching very high electron temperatures beyond the Curie point without over heating the lattice. It is therefore a relevant approach to use femtosecond pulses as it allows improving the conditions for obtaining an efficient switching due to the laser pre-heating. In that context, the variation of the coercive field H_c and magnetization at saturation M_S are important quantities to be characterized. We have investigated such dynamics in L1₀ FePt nanoparticles and accurately characterized the nonlinear variation of M_S and H_c upon varying the laser density of energy. We demonstrate that the Curie temperature can be reached during a few hundreds of femtoseconds, showing that the speed for addressing bits of information can be further improved in ultrafast HAMR applications.

[1] O. Mosendz, et al., J. Appl. Phys. 111, 07B729 (2012)

[2] D. Weller et al., Phys. Stat. Solidi A 210, 1245 (2013)

[3] S. Wicht et al., J. Appl. Phys. 114, 063906 (2013) & J. Appl. Phys. 117, 013907 (2015)

[4] E. Beaurepaire, J.-C. Merle, A. Daunois, J.-Y. Bigot, Phys. Rev. Lett. 76, 4250 (1996)

[5] J.-Y. Bigot, M. Vomir, Annalen der Physik 525, 2–30 (2013)

Because of the ability to exploit multiple absorption bands, Multi-junction Solar Cells (MJSCs) are the most efficient solar cells ever developed. As on single junction solar cells, Antireflection Coatings (ARCs) are utilized to achieve broadband absorption. Due to the fact that the total current of MJSCs is limited by the subcell that has the lowest generated current, ARCs on MJSCs must have the ability to minimize light loss at the range of limiting cell, but optimally all across the visible spectrum.

Unlike conventional Double Layer Antireflection Coatings (DLARCs) that can only reduce light reflection at certain wavelengths, moth eye structures are able to mitigate light loss over broadband wavelength due to their smooth change of refractive index. We report the fabrication of such a Subwavelength Structure (SWS) by using wet etching and dry etching of dielectric materials. Silicon wafers are used here as the substrate to test the quality of ARCs. ZnO has been chosen as one dielectric material because of its excellent transmittance and durability. It also has a close index match to the underlying GaInP window layer in this tandem cell. Wet etching using oxalic acid has been utilized here to texture the ZnO surface because it is a simple and cost-effective method. A low reflection (less than10%) over a broad range of wavelength (400-970nm) has been achieved. However, it turns out that wet etching is not very controllable and cannot fabricate the high aspect ratio periodic structure necessary for optimal absorption. A Zn(C₂O₄)Ÿ2H₂O bulk phase was found on the ZnO surface.

Thus, lithography and plasma etching have been employed because of their better process control capability and less dependence on the crystalline orientation of the material. We then switched to using Ta₂O₅ since it has similar optical properties to ZnO and also a broad bandgap to ensure its transparency. Dry etching of Ta₂O₅ gives us nanocones with aspect ratio (height over base diameter) of 1.26. At an angle of 8 degree from normal incidence, textured Ta₂O₅ achieved the averaged reflection as low as 6% over 320-900nm and itoutperforms DLARCs and textured ZnO over a wide range of wavelength. Future work will focus on fabricating this moth eye structure on III-V/SiGe tandem cells and simulating the reflectance spectra using Finite Difference Time Domain methods.

Different kind of thin films with remarkable magnetic and magnetoeletric properties require strong preferred orientation in order to utilize strong anisotropy of their properties.

The main aim of the work is to prepare the films of hexagonal ferrites showing magnetoelectric effects. There are several types of these materials marked as e.g. M-type [(Ba,Sr)Fe₁₂O₁₉, space group P6₃/mmc], Y-type [(Ba,Sr)₂Me₂Fe₁₂O₂₂, s. g. R-3m], Z-type [(Ba,Sr)₃Me₂Fe₂₄O₄₁, s. g. P6₃/mmc], and others. Our attention was focussed mainly to Y-type where the best properties are expected. All these lattices are long along c-axis and this should be oriented perpendicular to the surface. The films were prepared through the chemical solution method either on SrTiO₃ (111) or sapphire Al₂O₃ (0001) substrates, respectively. We are looking not only for suitable substrates but also we have succeeded in using seed template interlayers, for example M hexaferrite SrFe₁₂O_19. A detailed inspection revealed that growth of seed layers starts through the break-up of initially continuous film into isolated grains with expressive shape anisotropy and hexagonal habit. Promising type of such seed layers seem also to be SrAl₁₂O₁₉ films where two kinds of preparation were investigated – deposition of SrAl₁₂O₁₉ onto sapphire substrate and reaction SrO + Al₂O₃.

Other type of magnetic films studied were magnetic spinels Co₃O₄ prepared by decomposition of films of layered cobaltates Na_xCoO₂ deposited by chemical solution deposition method and grown on sapphire substrates.

Chemically ordering metal alloys such as FePt are hard magnets and good candidates for magnetic data storage in their ordered phase but not in the disordered state. The order-disorder phase transition temperature is impacted by the size of the particle with surface energies becoming significant for nanometer sized particles. Theoretical and computational approaches have predicted lowering of phase transition temperatures as the particle size decreases. Experimental evidence is more limited. Observation of ordering in nanoparticles is a complex interplay between thermodynamic and kinetic factors. Using Fe-Ni pseudo binary alloys with Pt allows isolation of thermodynamic variables. We see a size dependent reduction of order-disorder temperature in this system with particles in the 5 – 12 nm range. At 6 nm the reduction is ~15%