Search

The formation mechanisms that lead to the layered $M$-modulated $\mathrm{In}M{\mathrm{O}}_{3}(\mathrm{ZnO}{)}_{n}$ structures ($M=\mathrm{In}$, Ga, and Al; $n=\mathrm{\text{integer}}$) are revealed and confirmed by first-principles calculations based on density functional theory. We show that all ground state structures of $\mathrm{In}M{\mathrm{O}}_{3}(\mathrm{ZnO}{)}_{n}$ satisfy the octahedron rule for the ${\mathrm{InO}}_{2}$ layers; they contain an inversion domain boundary located at the $M$ and Zn fivefold trigonal bipyramid sites and maximize the hexagonality in the ($M{\mathrm{Zn}}_{n}){\mathrm{O}}_{n+1}$ layers. They also obey the electronic octet rule. This understanding provides a solid basis for studying and understanding the physical properties of this group of homologous materials.

The effects of impurities in room-temperature monoclinic ${\text{WO}}_{3}$ were studied by using the local density approximation to density-functional theory. Our main focus is on nitrogen impurity in ${\text{WO}}_{3}$, where both substitutional and interstitial cases were considered. We have also considered transition-metal atom impurities and some codoping approaches in ${\text{WO}}_{3}$. We find that, in general, band gap reduction was a common result due to the formation of impurity bands in the band gap. Also, the changes of band-edge positions, valence-band maxima and conduction-band minima, were found to depend on the electronic properties of the foreign atom and their concentration. Our results therefore provide guidance for making ${\text{WO}}_{3}$ a suitable candidate for photoelectrodes for hydrogen generation by water splitting.

We recently reported a new record total-area efficiency, 19.9%, for CuInGaSe2 (CIGS)-based thin-film solar cells [1]. Current-voltage analysis indicates that improved performance in the record device is due to reduced recombination. The reduced recombination was achieved by terminating the processing with a Ga-poor (In-rich) layer, which has led to a number of devices exceeding the prior (19.5%) efficiency record. This paper documents the properties of the high-efficiency CIGS by a variety of characterization techniques, with an emphasis on identifying near-surface properties associated with the modified processing.

We use the combination of high-resolution electron microscopy and density-functional theory to study the atomic structure and electronic effects of grain boundaries in polycrystalline photovoltaic materials such as Si, CdTe, CuInSe 2 , and CuGaSe 2 . We find that grain boundaries containing dislocation cores create deep levels in Si, CdTe, and CuGaSe 2 . Surprisingly, however, they do not create deep levels in CuInSe 2 . We further find that the presence of Ga in grain boundaries in CuInSe 2 generates deep levels. These results may explain the fact that Si and CdTe solar cells usually require special passivation, whereas CuInSe 2 solar cells do not. The passivation of grain boundaries in Si and CdTe is also studied. We find that grain boundaries in CdTe can be passivated very well by Cl, Br, and I.

This report demonstrates the synthesis and properties of large-area well aligned air-stable ZnO/ZnSe core/shell nanowires on a large area transparent conducting substrate (TCO) by combining chemical vapor deposition (CVD) and pulsed laser deposition (PLD) techniques. the ZnO nanowire were uniformly and perpendicularly grown on the TCO substrate and the ZnSe shell with a thickness of 5–8 nm were epitaxially grown on the ZnO nanowires core. Photoluminescence (PL) studies indicate that the core/shell nanowires are relatively immune to the defects present at the core/shell interface due to the large lattice mismatch. Such a core/shell nanowire array represents a novel architecture that could lead to a stable and efficient as well as low-cost solar cell technology for solar energy harvesting.

We present the results of systematic studies of optical enhancements by textured Ag/ZnO back reflectors in a-SiGe:H and nc-Si:H solar cells. First, the back reflector materials were characterized by AFM, XRD, and TEM. The results showed that the ZnO layers deposited by sputtering exhibit an increased surface texture with the deposition temperature and film thickness. The material structure showed a strong (002) preferential orientation. The large columnar crystal structure determines the surface texture. Second, the solar cell performance was correlated to the back reflector structure. We found that with a thin ZnO layer, a textured Ag layer results in more light scattering than a flat Ag layer. However, when a thick ZnO layer is used, a flat Ag layer can produce similar or more light scattering than a textured Ag layer. Third, we developed a method to estimate the optical enhancement for a-SiGe:H and nc-Si:H solar cells on various structures of Ag/ZnO back reflectors. Comparing the quantum efficiency data from solar cells made using the same recipe but one on a flat stainless steel substrate and another on a textured Ag/ZnO BR substrate revealed that the optical enhancement for the long wavelength light can be as high as 20 to 30. Compared to the theoretical value of 4n2, there is still scope for further improvement

We present a systematic study on the correlation of hydrogen dilution profiles to structural properties materials and solar cell performance in nc-Si:H solar cells. We deposited nc-Si:H single-junction solar cells using a modified very high frequency (VHF) glow discharge technique on stainless steel substrates with various profiles of hydrogen dilution in the gas mixture during deposition. The material properties were characterized using Raman spectroscopy, X-TEM, AFM, and C-AFM. The solar cell performance correlates well with the material structures. Three major conclusions are made based on the characterization results. First, the optimized nc-Si:H material does not show an incubation layer, indicating that the seeding layer is well optimized and works as per design. Second, the nanocrystalline evolution is well controlled by hydrogen dilution profiling in which the hydrogen dilution ratio is dynamically reduced during the intrinsic layer deposition. Third, the best nc-Si:H single-junction solar cell was made using a proper hydrogen dilution profile, which caused a nanocrystalline distribution close to uniform throughout the thickness, but with a slightly inverse nanocrystalline evolution. We have used the optimized hydrogen dilution profiling and improved the nc-Si:H solar cell performance significantly. As a result, we have achieved an initial active-area cell efficiency of 9.2% with a nc-Si:H single-junction structure, and 15.4% with an a-Si:H/a-SiGe:H/nc-Si:H triple-junction solar cell structure.

Metastable and stable phases in rapidly quenched, and quenched and annealed quaternary Al-Si-Mn-Fe alloys have been identified using electron and X-ray powder diffraction techniques. It has been found that the constituent phases in rapidly quenched specimens a (Al6Mn), b (Al82Mn17Fe1), c (Al86Si4Mn10), d (Al75Si7Mn14Fe4), e (Al78Si4Mn12Fe6) and f (Al75Si6Mn10Fe9) are metastable icosahedral and decagonal phases and stable crystalline Al, Al6Mn, alpha -AlMnFeSi and Al13Fe4 phases. The constituent phases in the quenched and annealed Al-Si-Mn-Fe specimens are crystalline Al, Al6Mn, Al11Mn4, alpha -AlMnFeSi and Al13Fe4 phases. The relative contents of these phases are dependent on the composition of each alloy.

The formation mechanisms that lead to the layered M-modulated InMO3(ZnO){n} structures (M=In, Ga, and Al; n=integer) are revealed and confirmed by first-principles calculations based on density functional theory. We show that all ground state structures of InMO3(ZnO){n} satisfy the octahedron rule for the InO2 layers; they contain an inversion domain boundary located at the M and Zn fivefold trigonal bipyramid sites and maximize the hexagonality in the (MZn{n})O{n+1} layers. They also obey the electronic octet rule. This understanding provides a solid basis for studying and understanding the physical properties of this group of homologous materials.

A systematic study of nine binary and ternary spinel oxides formed from Co, Al, and Fe is presented by means of density functional theory. Analysis of the structural, magnetic, and electronic properties through the series of materials is carried out. Preference for the octahedral spinel sites are found in the order Fe

The classic grain-boundary (GB) model concludes that GBs in polycrystalline semiconductors create deep levels that are extremely harmful to optoelectronic applications. However, our first-principles density-functional theory calculations reveal that, surprisingly, GBs in CuInSe2 (CIS) do not follow the classic GB model: GBs in CIS do not create deep levels due to the large atomic relaxation in GB regions. Thus, unlike the classic GB model, GBs in CIS are electrically benign, which explains the long-standing puzzling fact that polycrystalline CIS solar cells with remarkable efficiency can be achieved without deliberate GB passivation. This benign electrical character of GBs in CIS is confirmed by our scanning Kelvin probe microscopy measurements on Cu(In,Ga)Se2 chalcopyrite films.

We combine first-principles density functional theory, material synthesis and characterization, and photoelectrochemical (PEC) measurements to explore methods to effectively reduce the band gap of ZnO for the application of PEC water splitting. We find that the band gap reduction of ZnO can be achieved by N and Cu incorporation into ZnO. We have successfully synthesized ZnO:N thin films with various reduced band gaps by reactive RF magnetron sputtering. We further demonstrate that heavy Cu-incorporation lead to both p-type doping and band gap significantly reduced ZnO thin films. The p-type conductivity in our ZnO:Cu films is clearly revealed by Mott-Schottky plots. The band gap reduction and photoresponse with visible light for N- and Cu-incorporated ZnO thin films are demonstrated.

Using scanning transmission electron microscopy and nanoprobe X-ray energy-dispersive spectroscopy, we studied the locations of Cu in CdTe solar cells and test interfaces with intentionally introduced Cu sources. We found three primary locations of Cu: 1) back-contact region, 2) CdTe layer, and 3) CdTe/CdS junction areas. In the back-contact region, Cu diffused from back-contact can convert the Te-rich layer into Cu-Te compounds. In the CdTe layer, a higher concentration of Cu is found to distribute along grain boundaries, but not in twin boundaries and stacking faults. In the CdTe/CdS junction area, Cu is found uniformly in the CdS layer. However, significant segregation of Cu into CdS/TCO interfaces is also found.

High open-circuit voltage (Voc) silicon heterojunction (SHJ) solar cells are fabricated in double-heterojunction a-Si:H/c-Si/a-Si:H structures using low temperature (p-type c-Si float-zone wafers, we used an amorphous n/i contact to the top surface and an i/p contact to the back surface to obtain a Voc of 667 mV in a 1 cm2 cell with an efficiency of 18.2%. This is the best reported p-type SHJ voltage. In our labs, it improves over the 652 mV cell obtained with a front amorphous n/i heterojunction emitter and a high-temperature alloyed Al back-surface-field contact. On n-type c-Si float-zone wafers, we used an a Si:H (p/i) front emitter and an a-Si:H (i/n) back contact to achieve a Voc of 691 mV on 1 cm2 cell. Though not as high as the 730 mV reported by Sanyo on n-wafers, this is the highest reported Voc for SHJ c-Si cells processed by the HWCVD technique. We found that effective c-Si surface cleaning and a double-heterojunction are keys to obtaining high Voc. Transmission electron microscopy reveals that high Voc cells require an abrupt interface from c-Si to a-Si:H. If the transition from the base wafer to the a-Si:H incorporates either microcrystalline or epitaxial Si at c Si interface, a low Voc will result. Lifetime measurement shows that the back-surface-recombination velocity (BSRV) can be reduced to ~15 cm/s through a-Si:H passivation. Amorphous silicon heterojunction layers on crystalline wafers thus combine low-surface recombination velocity with excellent carrier extraction.

Solid-phase crystallization (SPC) rates are compared in amorphous silicon films prepared by three different methods: hot-wire chemical vapor deposition (HWCVD), plasma-enhanced chemical vapor deposition (PECVD), and electron-beam physical vapor deposition (e-beam). Random SPC proceeds approximately 5 and 13 times slower in PECVD and e-beam films, respectively, as compared to HWCVD films. Doping accelerates random SPC in e-beam films but has little effect on the SPC rate of HWCVD films. In contrast, the crystalline growth front in solid-phase epitaxy experiments propagates at similar speed in HWCVD, PECVD, and e-beam amorphous Si films. This strongly suggests that the observed large differences in random SPC rates originate from different nucleation rates in these materials while the grain growth rates are relatively similar. The larger grain sizes observed for films that exhibit slower random SPC support this suggestion.

The surface structures of ionic zinc-blende CdTe (001), (110), (111), and (211) surfaces are systematically studied by first-principles density functional calculations. Based on the surface structures and surface energies, we identify the detrimental twinning appearing in molecular beam epitaxy (MBE) growth of II-VI compounds as the (111) lamellar twin boundaries. To avoid the appearance of twinning in MBE growth, we propose the following selection rules for choosing optimal substrate orientations: (1) the surface should be nonpolar so that there is no large surface reconstructions that could act as a nucleation center and promote the formation of twins; (2) the surface structure should have low symmetry so that there are no multiple equivalent directions for growth. These straightforward rules, in consistent with experimental observations, provide guidelines for selecting proper substrates for high-quality MBE growth of II-VI compounds.

We conducted cathodoluminescence (CL) spectrum imaging and electron backscatter diffraction on the same microscopic areas of CdTe thin films to correlate grain-boundary (GB) recombination by GB “type.” We examined misorientation-based GB types, including coincident site lattice (CSL) Σ = 3, other-CSL (Σ = 5–49), and general GBs (Σ > 49), which make up ∼47%–48%, ∼6%–8%, and ∼44%–47%, respectively, of the GB length at the film back surfaces. Statistically averaged CL total intensities were calculated for each GB type from sample sizes of ≥97 GBs per type and were compared to the average grain-interior CL intensity. We find that only ∼16%–18% of Σ = 3 GBs are active non-radiative recombination centers. In contrast, all other-CSL and general GBs are observed to be strong non-radiative centers and, interestingly, these GB types have about the same CL intensity. Both as-deposited and CdCl2-treated films were studied. The CdCl2 treatment reduces non-radiative recombination at both other-CSL and general GBs, but ...

Using sunlight and semiconductors to split water in a photoelectrochemical (PEC) process to generate hydrogen provides a viable way to produce renewable fuel. Metal oxide photoelectrodes are of particular interest due to their low cost and relatively high stability in aqueous media. Spinel Co-Fe-Al oxide system has shown the potential for PEC hydrogen production. The electronic and optical properties of Co based spinel oxides, CoX2O4 (X =Al, Ga, In) are calculated using first-principle density functional theory. We show that the desirable properties for solar photoconversion can be obtained by controlling the alloy compositions. The ability to manipulate the electronic and optical properties is attributed to the different s-orbital energies and sizes of the cations. Our calculated bandgaps as function of composition provide detailed practical guidance for synthesizing Co based spinel oxides with electronic and optical properties necessary to achieve high efficiency PEC decomposition of water for hydrogen production by sunlight. Our experimental work demonstrated the tunable band gaps and the PEC responses of the Co based spinel oxides.

The crystal structures of Ru2C were extensively investigated using the structure searching method coupled with first-principles calculations. In contrast to the previously proposed P-3m1 phase, two energetically more favorable structures with space groups P-31m and P63/mmc were found, which are stable at pressure ranges of 0-32 GPa and 32-50 GPa, respectively. The dynamical stabilities of both phases at ambient and high pressures are confirmed by the phonon dispersions. Further calculations indicate that the two new predicted phases are ultra-incompressible metals due to a strong covalent Ru-C bond.

Thin-film solar cells based on polycrystalline Cu(In,Ga)Se2 (CIGS) and CdTe photovoltaic semiconductors have reached remarkable laboratory efficiencies. It is surprising that these thin-film polycrystalline solar cells can reach such high efficiencies despite containing a high density of grain boundaries (GBs), which would seem likely to be nonradiative recombination centers for photo-generated carriers. In this paper, we review our atomistic theoretical understanding of the physics of grain boundaries in CIGS and CdTe absorbers. We show that intrinsic GBs with dislocation cores exhibit deep gap states in both CIGS and CdTe. However, in each solar cell device, the GBs can be chemically modified to improve their photovoltaic properties. In CIGS cells, GBs are found to be Cu-rich and contain O impurities. Density-functional theory calculations reveal that such chemical changes within GBs can remove most of the unwanted gap states. In CdTe cells, GBs are found to contain a high concentration of Cl atoms. Cl atoms donate electrons, creating n-type GBs between p-type CdTe grains, forming local p-n-p junctions along GBs. This leads to enhanced current collections. Therefore, chemical modification of GBs allows for high efficiency polycrystalline CIGS and CdTe thin-film solar cells.

Recent theoretical and experimental reports have shown that the perovskite CH3NH3PbI3 exhibits unique ambipolar self-doping properties. Here, we show by density-functional theory calculation that its sister perovskite, CH3NH3PbBr3, exhibits a unipolar self-doping behavior—CH3NH3PbBr3 presents only good p-type conductivity under thermal equilibrium growth conditions. We further show that despite a large bandgap of 2.2 eV, all dominant defects in CH3NH3PbBr3 create shallow levels, which partially explains the ultra-high open-circuit voltages achieved by CH3NH3PbBr3-based thin-film solar cells. Our results suggest that the perovskite CH3NH3PbBr3 can be both an excellent solar cell absorber and a promising low-cost hole-transport material for lead halide perovskite solar cells.

Atomic-resolution Z-contrast images demonstrate unambiguously that the annealed, metalorganic chemical vapor deposition derived transparent In2O3–ZnO films have a polytypoid microstructure, consisting of ZnO slabs of variable width separated by single In–O octahedral layers. These In–O layers induce a polarity inversion in the two adjacent ZnO layers, which is reversed again by a mirror domain boundary inside each ZnO slab.

The atomic structure of ordered domains in Ba(Mg1/3Nb2/3)O3 and La-doped and undoped Pb(Mg1/3Nb2/3)O3 is studied by high-resolution Z-contrast imaging. The ordered domain structure in both doped and undoped Pb(Mg1/3Nb2/3)O3 is determined to be in agreement with the charge balanced random-layer model and inconsistent with the space-charge model. It is shown that La doping in Pb(Mg1/3Nb2/3)O3 enhances not only the domain size but also the degree of ordering.

The asymmetry doping problem has severely hindered the potential applications of many wideband gap (WBG) materials. Here, we propose a possible approach to overcome this long-standing doping asymmetry problem for WBG semiconductors. Our approach is based on the reduction of the ionization energies of dopants through introduction and effective doping of mutually passivated impurity bands, which can be realized by doping the host with passive donor-acceptor complexes or isovalent impurities. Our density-functional theory calculations demonstrate that this approach provides excellent explanations for the n-type doping of diamond and p-type doping of ZnO, which could not be understood by previous theories. In principle, this approach can be applied to any WBG semiconductors and therefore will open a broad vista for the application of WBG materials.

Current studies have attributed the remarkable performance of polycrystalline ${\mathrm{CuInSe}}_{2}$ (CIS) to anomalous grain-boundary (GB) physics in CIS. The recent theory predicts that GBs in CIS are hole barriers, which prevent GB electrons from recombining. We examine the atomic structure and chemical composition of (112) GBs in $\mathrm{Cu}(\mathrm{In},\mathrm{Ga}){\mathrm{Se}}_{2}$ (CIGS) using high-resolution $Z$-contrast imaging and nanoprobe x-ray energy-dispersive spectroscopy. We show that the theoretically predicted Cu-vacancy rows are not observed in (112) GBs in CIGS. Our first-principles modeling further reveals that the (112) GBs in CIS do not act as hole barriers. Our results suggest that the superior performance of polycrystalline CIS should not be explained solely by the GB behaviors.

Solid state crystallization of hydrogenated amorphous silicon (a-Si:H) prepared by hot-wire CVD is studied in solid phase epitaxy mode. By using a novel optical method combined with cross-sectional TEM and SIMS, a reduction of epitaxial growth speed is observed with increase in a-Si:H film thickness. Namely, in films thinner that 0.9 micron, solid phase epitaxy velocity depends linearly on film thickness. As the film thickness increases beyond 1 micron, the average velocity of solid phase epitaxy decreases considerably with respect to that in thinner films. In this regime, its velocity becomes also time-dependent: initial slow propagation of crystallization front gets considerably accelerated after the front has traveled above 400nm. SIMS depth profiles of hydrogen shows considerably more residual hydrogen in thicker films after the start of solid phase epitaxy. In addition, prolonged pre-dehydrogenation at lower temperatures results in the increase in the average epitaxy speed in thicker films. These phenomena are likely related to delayed hydrogen outdiffusion in thicker films, which also leads to time-dependent speed of the solid-phase epitaxy front. Thus, the excess residual hydrogen in CVD films reduces the rate of solid-phase crystalline growth, similarly to earlier results on H-implantation and indiffusion.

We have achieved an independently-confirmed 17.8% conversion efficiency in a 1-cm2, p-type, float-zone silicon (FZ-Si) based heterojunction solar cell. Both the front emitter and back contact are hydrogenated amorphous silicon (a-Si:H) deposited by hot-wire chemical vapor deposition (HWCVD). This is the highest reported efficiency for a HWCVD silicon heterojunction (SHJ) solar cell. Two main improvements lead to our most recent increases in efficiency: 1) the use of textured Si wafers, and 2) the application of a-Si:H heterojunctions on both sides of the cell. Despite the use of textured c-Si to increase the short-circuit current, we were able to maintain the same 0.65 V open-circuit voltage as on flat c-Si. This is achieved by coating a-Si:H conformally on the c-Si surfaces, including covering the tips of the anisotropically-etched pyramids. A brief atomic H treatment before emitter deposition is not necessary on the textured wafers, though it was helpful in the flat wafers. It is essential to high efficiency SHJ solar cells that the emitter grows abruptly as amorphous silicon, instead of as microcrystalline or epitaxial Si. The contact on each side of the cell comprises a thin (< 5 nm) low substrate temperature (~100°C) intrinsic a-Si:H layer, followed by a doped layer. Our intrinsic layers are deposited at 0.3-1.2 nm/s. The doped emitter and back-contact layers were deposited at a higher temperature (>200°C) and grown from PH3/SiH4/H2 and B2H6/SiH4/H2 doping gas mixtures, respectively. This combination of low (intrinsic) and high (doped layer) growth temperatures was optimized by lifetime and surface recombination velocity measurements. Our rapid efficiency advance suggests that HWCVD may have advantages over plasma-enhanced (PE) CVD in fabrication of high-efficiency heterojunction c-Si cells; there is no need for process optimization to avoid plasma damage to the delicate, high-quality, Si wafers.

We present first-principles total-energy calculations of water adsorption on the $\mathrm{Zn}\mathrm{O}(10\overline{1}0)$ surface. We find that on ideal $\mathrm{Zn}\mathrm{O}(10\overline{1}0)$ surfaces, the fully molecular chemisorption is most energetically favorable at coverage $\ensuremath{\leqslant}0.5\phantom{\rule{0.3em}{0ex}}\mathrm{ML}$, whereas at 1 ML, the mixed adsorption with a $(2\ifmmode\times\else\texttimes\fi{}1)$ superstructure is most energetically favorable. However, the mixed adsorption with a $(2\ifmmode\times\else\texttimes\fi{}2)$ superstructure is only 0.07 eV higher per ${\mathrm{H}}_{2}\mathrm{O}$ molecule than the mixed adsorption with a $(2\ifmmode\times\else\texttimes\fi{}1)$ superstructure. Therefore, we expect that domains with mixed $(2\ifmmode\times\else\texttimes\fi{}1)$ and $(2\ifmmode\times\else\texttimes\fi{}2)$ structures may also exist at elevated temperatures. Moreover, we find that oxygen vacancies in the surface significantly affect the adsorption behavior. For an oxygen vacancy density $\ensuremath{\leqslant}25%$, the dissociative adsorption is energetically preferred at water coverage $\ensuremath{\leqslant}0.25\phantom{\rule{0.3em}{0ex}}\mathrm{ML}$ and the mixed adsorption is energetically favorable at water coverage $\ensuremath{\geqslant}0.5\phantom{\rule{0.3em}{0ex}}\mathrm{ML}$. The further increase of oxygen vacancy density promotes the dissociative adsorption.

We present first-principles total-energy calculations of adsorption and reaction of molecular oxygen on the $\mathrm{ZnO}(10\overline{1}0)$ surface. We find that, on stoichiometric $\mathrm{ZnO}(10\overline{1}0)$ surfaces, the adsorption is fully molecular with a small adsorption energy and the dissociation is energetically unfavorable. On a partially reduced $\mathrm{ZnO}(10\overline{1}0)$ surface, the dissociative adsorption is energetically preferred, with one of the resulting oxygen atoms filling the original oxygen vacancy, while the other is in between two neighboring Zn atoms, forming a bridging oxygen adatom. This bridging oxygen may diffuse along the Zn-O dimer row and capture and fill another oxygen vacancy. The dissociative adsorption influences the electronic properties of the $\mathrm{ZnO}(10\overline{1}0)$ surface.

We use a high-throughput (combinatorial) hot-wire chemical vapor deposition system to passivate the crystal silicon surface and to grow heterojunction silicon solar cells. We study the effectiveness of crystal surface treatments by atomic H or/and NH/sub x/ radicals, followed by the growth of thin hydrogenated amorphous silicon (a-Si:H) films. Treatment and layer properties such as times, thicknesses and gas mixtures can be continuously graded, creating a two-dimensional sample with each variable varying in one direction. This results in high-throughput optimization of the processes. Effective carrier lifetime is measured by photoconductive decay to evaluate the effectiveness of the surface passivation by surface treatments. The effective carrier lifetime increases from about 5 /spl mu/s without passivation to about 24 /spl mu/s with an optimized surface treatment and thickness a-Si:H on single-sided c-Si. Transmission electron microscopy reveals that a-Si:H, a mixed phase, or epitaxial growth of thin-film Si depending upon the surface treatment. Improvement in effective carrier lifetime correlates to with an immediate a-Si:H growth on c-Si, rather than a mixed phase and epitaxial Si growth. We have obtained an efficiency of 13.4% on a non-textured single-sided heterojunction solar cell on a p-type CZ-Si processed with optimized surface treatment.

A mechanism for the transport of H into a Si solar cell during plasma-enhanced chemical vapor deposition (PECVD) of a hydrogenated silicon nitride (SiN:H) layer and its subsequent fire-through metallization process is described. The PECVD process generates process-induced traps, which "store" H at the surface of the solar cell. This stored H is released and diffuses rapidly into the bulk of Si during the high-temperature metallization-firing process. During the ramp-down, the diffused H associates with impurities and defects and passivates them. The firing step partially heals up the surface damage. The proposed model explains a variety of observations and experimental results.

We measure times for complete solid phase crystallization (SPC) of hydrogenated amorphous silicon (a-Si:H) thin films that vary eight orders of magnitude, from a few ms to a few days. The time-to-crystallization activation energy is consistent with literature values of approximately 3.4 eV but the prefactor is markedly different for hot-wire chemical vapor deposition (HWCVD) films than for plasma-enhanced (PE) CVD films. The crystallized films were 0.3 – 2 μm thick, and deposited by high deposition rate (10-100 Å/s) HWCVD or standard PECVD onto glass substrates. We annealed these a-Si:H films over a wide temperature range (500 to 1100 °C) using techniques including simple hot-plates and tube furnaces, rapid thermal annealing by a tungsten-halogen lamp, and microwave electromagnetic heating at 2.45 GHz (magnetron) and 110 GHz (gyrotron).

The microstructure of high-efficiency CdTe solar cells on commercial SnO2/soda-lime glass was investigated by scanning transmission electron microscopy. The CdTe solar cells have a structure of soda-lime glass/SnO2/ZTO/CdS:O/CdTe. We found no interdiffusion between the SnO2 layer and ZTO layer. Weak diffusion of Zn from the ZTO layer into the CdS:O layer was observed; however, the diffusion was not uniform. Interdiffusion also occurred at the CdTe/CdS:O interface. In the back-side of the CdTe, a thin layer of Te was found, which formed during the nitric-phosphoric etching. In addition, a very thin layer of CdHgTe was observed at the CdTe/Te interface.

Using the combination of high-resolution transmission electron microscopy, first-principles density-functional total-energy calculations, and image simulations, we studied the atomic structure and passivation effects of double-positioning (DP) twin boundaries in CdTe. The DP twin boundaries are found to contain more Te dangling bonds than Cd dangling bonds, resulting in energy states in the bandgap that are detrimental to the electronic properties of CdTe. We found that I, Br, Cl, S, and O atoms present passivation effects on the DP twin boundaries to differing degrees, whereas H does not passivate the boundaries. Of all these impurities, I and Cl atoms present the best passivation effects on the DP twin boundaries. The superior passivation effects are realized by either terminating the Cd atoms with dangling bonds, or substituting the Te atoms with dangling bonds in the DP twin boundaries in CdTe by Cl and I atoms.

The energetics and electronic structures of basal-plane stacking faults in wurtzite (WZ) ZnO are studied using first-principles density-functional total energy calculations. All the basal-plane stacking faults are found to have very low formations energies. They also introduce a downward shift at the conduction-band minimum (CBM). However, plane-averaged charge densities of the CBM state reveal that the CBM states are not very localized, indicating that these stacking faults should be electronically inert. The high concentration of these stacking faults can result in embedded zinc-blende (ZB) ZnO surrounded by WZ materials. The WZ/ZB interface exhibits a type-II lineup with $\ensuremath{\Delta}{E}_{V}\ensuremath{\approx}0.037\phantom{\rule{0.3em}{0ex}}\mathrm{eV}$ and $\ensuremath{\Delta}{E}_{C}\ensuremath{\approx}0.147\phantom{\rule{0.3em}{0ex}}\mathrm{eV}$.

Quasicrystals are solids that exhibit symmetries long thought forbidden in nature. Since their discovery in a rapidly solidified Al–Mn alloy in 1984, the central issue in the field has been to understand why they form. Are they energetically stable compounds or stabilized by entropy? In recent years, major strides have been made in determining atomic structure, largely by direct imaging using advanced electron microscopy. One system is now known to be energetically stabilized, and quasicrystals are therefore firmly established as a new physical state of matter. They represent a unique packing of atomic clusters some tens of atoms in size, with substantial localized fluctuations, referred to as phasons. Understanding phasons may in future allow their unique macroscopic properties to be tailored for useful materials applications.

We present first-principles total-energy calculations of domain wall energies for inversion domain boundaries and their interaction with point defects in ZnO. We find that the inversion domain boundaries (IDB's) with Zn-Zn and O-O wrong bonds are energetically unstable, and they transform into the so-called ${\mathrm{IDB}}^{*}$ structures. Although the ${\mathrm{IDB}}^{*}'\mathrm{s}$ do not induce electronic states in the bandgap, they are very attractive to electrically active compensating point defects. The segregation of such charged point defects builds up electrical potentials for minority carriers in doped ZnO, which are harmful to optoelectronic applications.

In the conventional CdS/CdTe device structure, the poly-CdS window layer has a bandgap of /spl sim/2.4 eV, which causes absorption in the short-wavelength region. Higher short-circuit current densities (J/sub sc/) can be achieved by reducing the CdS thickness, but this can adversely impact device open-circuit voltage (V/sub oc/) and fill factor (FF). Also, poly-CdS film has about 10% lattice mismatch related to the CdTe film, which limits the improvement of device V/sub oc/ and FF. In this paper, we report a novel window material: oxygenated amorphous CdS film (a-CdS:O) prepared at room temperature by rf sputtering. The a-CdS:O film has a higher optical bandgap (2.5-3.1 eV) than the poly-CdS film and an amorphous structure. The preliminary device results have demonstrated that J/sub sc/ of the CdTe device can be greatly improved while maintaining higher V/sub oc/ and FF. We have fabricated a CdTe cell demonstrating an NREL-confirmed J/sub sc/ of 25.85 mA/cm/sup 2/ and a total-area efficiency of 15.4%.

We investigate the structure and energetics of native and oxygen-related defects in MgB 2 . First-principles total-energy calculations reveal that the nonstoichiometry in MgB 2 can be very small. Oxygen impurities have high solubility in MgB 2 at O-rich growth conditions. However, the presence of excess Mg during the growth can dramatically reduce oxygen incorporation by forming MgO precipitates. Oxygen impurities are found to exist in atomic form in MgB 2 . Interstitial oxygen atoms are very mobile, with an activation energy of 1.3 eV, indicating that they can be annealed out ex situ in Mg vapor.

We report on our investigation of the microstructure and chemical composition at the CdTe/Te-rich interfaces generated by NP-etching polycrystalline and single-crystalline CdTe films and followed with HgTe-graphite pasting and thermal annealing. We find that after this process, a thin layer of CdxHg1-xTe forms between CdTe and Te-rich layers, giving a structure like CdTe/CdxHg1-xTe/Te. High-resolution electron microscopy reveals that the CdxHg1-xTe layer has an epitaxial relationship with the CdTe. No CdxHg1-xTe layer has been observed in bromine/methanol-etched samples or samples with intentionally deposited Te layers.

We report on the effects of doping on the growth of ZnO nanostructures. We find that the incorporation of dopants, such as Si, Li, and Na, usually lead to reduced quality of ZnO nanostructures. Although doping with Li and Na results in the formation of facets on ZnO nanostructures, doping of Si often reduces dramatically the quantity of ZnO nanostructures. The formation of facets on ZnO nanostructures is found to significantly affects the cathodoluminescence emissions of the nanostructures.

We investigate local structural variations in Al72Ni20Co8 quasicrystals using first principles density-functional total-energy calculations. We find that local chemical fluctuation can cause large structural variations on the central rings of the 2-nm clusters. Such structure variations may result in ambiguities at the central rings in high-resolution electron microscopy images of Al72Ni20Co8 decagonal quasicrystals.

Surface reconstructions involving curved graphitic overlayer are proposed to be the missing transition state between the clean (2\ifmmode\times\else\texttimes\fi{}1)- and fully hydrogenated (1\ifmmode\times\else\texttimes\fi{}1)-diamond {111} surfaces. First-principles calculations show that surface graphitization causes changes in the H vibration frequency, in good agreement with experiment. Localized graphitization may also introduce voidlike planar defects that increase the surface-to-bulk ratio. We find that phosphorus solubility is significantly enhanced near the surface region. This provides an explanation for the recent experimental success on n-type diamond.

We investigate the structure, energetics, and effects of stacking faults in ${\mathrm{MgB}}_{2}$ by means of first-principles density-functional calculations. Two structurally different stacking faults were determined with displacements of $\mathbf{R}=[01/20]$ and [2/3, 1/3 0], and formation energies calculated to be 956 and 932 ${\mathrm{e}\mathrm{r}\mathrm{g}/\mathrm{n}\mathrm{m}}^{2}$, respectively. Significant expansions were found at both stacking faults. The partial density of states showed that the B atoms in these stacking faults present large reductions in the density of states around the Fermi energy. However, due to their low density, these stacking faults may not cause a significant decrease of the transition temperature of ${\mathrm{MgB}}_{2}.$

We propose an effective co-alloying approach to creating an intermediate band (IB) in bulk ZnTe. First-principles calculations show that a donor–acceptor co-alloying scheme can produce an IB within the band gap of ZnTe and that the position of the IB can be tuned by choosing appropriate donor–acceptor pairs. We find that the position of the IB is governed by the atomic d-orbital energies of dopants, and also by the charge transfer between donor and acceptor atoms and their subsequent intra- and inter-Coulomb interactions. Therefore, the location of the IB can be manipulated by selecting donors and acceptors with desirable d-orbital energies and electronegativities.