Recently, several thermoelectric semiconductor systems have been developed with double the thermoelectric figure of merit, zT, of conventional materials. Almost all the progress comes from achieving reductions in thermal conductivity using nanotechnologies or structural disorder. Because the lattice thermal conductivity has a lower limit, the amorphous limit, further progress must come from an enhancement of the electrical properties, in particular the thermoelectric power or Seebeck coefficient. We present a new technique to achieve that, based on creating distortions of the density-of-states by doping with resonant impurities, resulting in a doubling of the zT of PbTe, a semiconductor used for power generation applications near 500 ^oC. We will also review current progress with this technique in Bi₂Te₃, the other classical thermoelectric, which we study because the commercial materials used for Peltier cooling are (Bi_1-xSb_x)₂(Te_1-ySe_y)₃ alloys. The theory behind this approach will be outlined, and its applicability to a wide variety of thermoelectric semiconductors discussed.

properties due to the modification of the carrier type and concentration.

Narrow band gap nanostructures such as cadmium sulfide quantum dots (QDs) are known to show size quantization effects as well as multiple exciton generation. They are therefore beneficial for absorption of light in the visible and near infrared region of the solar spectrum and can be used to fabricate photovoltaic devices with high theoretical efficiencies. In quantum dot sensitized solar cells (QDSSCs), these QDs can be engineered to transfer the electron to a wide band gap semiconductor such as titanium dioxide (TiO₂). However, performance in such devices is reduced by charge recombination at the TiO₂ surface and hence use of organic linkers and electron conductors such as self-assembled monolayers (SAMs) on these devices could provide a means of eliminating recombination sites and lead to increased efficiency. In this study, we investigated the effects of different aliphatic and aromatic SAMs with phosphonic acid headgroups and varied tailgroups on the bonding and performance of cadmium sulfide (CdS) QDSSCs. Our studies focus on bonding of the CdS QDs on both planar and nanoporous TiO₂ with or without the SAM linkers. To study the SAM/QD growth on planar surfaces, TiO₂ was deposited on Piranha-cleaned Si or microscope glass via atomic layer deposition (ALD) and the resulting surfaces were characterized by X-ray photoelectron spectroscopy (XPS), Auger electron spectroscopy (AES), scanning electron microscopy (SEM), and atomic force microscopy (AFM). Next, different SAMs were attached to the TiO₂ substrates from solution, where the effects of chain length, aromaticity, tailgroup, solvent, and dip time on the quality of the SAMs were investigated by the same techniques as well as infrared (IR) spectroscopy, water contact angle (WCA) measurements, and ellipsometry. Finally, CdS QDs were grown on the SAM-passivated TiO₂ surfaces by the successive ionic layer adsorption and reaction (SILAR) process, and the bonding and performance of the resulting materials were evaluated by UV-visible and other spectroscopic techniques. The results were compared to the case of QDs grown on the non-passivated TiO₂ surfaces. Our results show promising differences in the bonding of the CdS QDs at the TiO₂ surfaces with the SAM linkers. We will also present results on the dependence of solar cell performance on the properties of the SAM.

Multi-junction solar cells based on the GaInP₂/GaAs/Ge triple junction architecture have achieved the highest efficiency of any photovoltaic device, for either space or terrestrial applications. However, the cost of these cells is high compared to other photovoltaic devices. For space applications (i.e., satellite power) the higher efficiencies are more critical than cell cost, and on a system level, including launch and deployment costs, the multi-junction cells are actually lower on a $/watt basis than Si solar cells. The high efficiency multi-junction cell is now widely used on satellites. For terrestrial applications, the high efficiency cells must be used in high concentration systems where the high cell cost is offset by the lower costs of lenses, mirrors, and structure metal. The cell becomes such a small part of the system cost that a doubling of cell cost has a small effect on the cost of power generation. However, the efficiency of the cell may have a much larger effect on the cost of power generated. So for either space or terrestrial applications the efficiency of the photovoltaic device is very important.

Multi-junction solar cells offer a performance advantage over single junction solar cells because of the reduction in carrier thermallization losses. Different parts of the solar spectrum are absorbed by different band gap materials. III-V materials are able to achieve very high performance on a single junction basis, and when appropriately combined provide an even higher performance advantage. Fortunately, GaInP₂, GaAs, and Ge junctions can be combined in a lattice matched configuration, making a monolithic device with a good combination of junctions for converting the solar spectrum into power. Appropriate modifications can be made to the device to optimize it for either space or terrestrial applications.

To achieve even higher efficiencies, more junctions need to be added and/or the junctions need to be better matched to the solar spectrum. A number of approaches have been studied, including novel materials (e.g., InGaAsN, ZnGeAs₂, etc.), mechanically stacked devices, and metamorphic devices. With metamorphic devices junctions are grown lattice mismatched on one another to achieve the optimal set of band gaps in a complete device. One metamorphic approach, the inverted metamorphic multi-junction (IMM) solar cell, has demonstrated significant performance improvements over the lattice matched triple junction device.

We report on these improvements, describing how the IMM approach is enabling for both space and terrestrial power generation.

Nowadays the application of thin SiN_x layers as bulk passivating and antireflection coatings for Si-based solar cells applications is considered to be a successful solution for the increase in efficiency¹. A new concept to further increase the efficiency of the solar cell is based on the light conversion mechanism: according to this approach the solar spectrum can be efficiently modified by shifting the photons towards a wavelength range where the solar cell has a better or higher response².

Recently, a novel class of rare earth (RE)- doped SiN_x layers has been demonstrated to be a highly promising red-emitting conversion phosphor for white-LED applications. These materials have allowed the shifting of the emission wavelength by tuning the concentration of a specific RE element in a SiN_x based crystalline matrix³. The investigation of RE-doped amorphous silicon nitride (SiN_x) compounds, where the electronic properties of Si are combined with the optical properties of RE³⁺ ions, have been shown already potential in optoelectronics⁴. Therefore, parallel studies on the incorporation of a RE material in amorphous SiN_x host lattices, which could be implemented in solar cells to increase the efficiency, are considered to be presently a challenge.

In this contribution the properties of europium- and samarium- doped amorphous SiN_x layers are investigated. The RE-doped SiN_x layers are deposited using a remote PECVD expanding thermal plasma fed with Ar/SiH₄/NH₃ mixtures in combination with a RE magnetron sputtering source implemented in the proximity of the substrate holder. Growth rates of the RE doped layers obtained from Spectroscopic Ellipsometry (SE) measurements were in the range 0.6-2.2 nm/s. The successful incorporation of RE in the SiNx matrix has been demonstrated by means of Rutherford Back Scattering (RBS) and X-ray Photoelectron Spectroscopy (XPS) analysis, i.e. up to 2%. Preliminary photoluminescence results point out a broad band emission in the region of 500-800 nm when excitation wavelengths of 270 nm and 320 nm have been used. The emission band observed can be attributed to Sm²⁺.

[1] J. Hong et al., J. Vac. Sci. Technol. B. 21 (5).

[2] C. Strümpel et al., Sol. Energ. Mat. Sol. C. 91 (2007) 238 – 249.

[3] Y. Q. Li et al., J. Alloys. and Comp. 417, 273 – 279.

[4] A. R. Zanatta, et al., J. Phys.:Condens. Matter 19 (2007) 436230.