Disc-shaped nanocrystal model for simulating the diffraction peak profile from a one-dimensional superlattice and its application to Pt/AlN superlattice films

On the basis of the film structure of a Pt/AlN multilayer film which exhibits a superlattice peak after annealing, a disc-shaped nanocrystal model is proposed for simulating the diffraction line profile from a one-dimensional superlattice. The proposed model demonstrates that a superlattice peak can arise even from just two disc-shaped nanocrystals and even with a large misorientation (a few degrees), provided both the thickness and the in-plane crystal size of the nanocrystals are of the order of nanometres. Using the model, the superlattice peaks from Pt/AlN superlattices are analysed quantitatively and the effect of annealing on the film is discussed.

Study of HfAlOx Films Deposited by Layer-by-Layer Growth for CMOS High-k Gate Dielectrics

ABSTRACTWe have investigated ternary metal oxide films for high-k gate dielectrics by using a layer-by-layer deposition & annealing method so as to keep the dielectric constant of the film high with no crystallization. HfAlOx films have been prepared by alternating deposition of HfO2 and Al2O3 layers, where it is a key for the improvement of the film quality to understand the intermixing process between two layers by thermal treatments. So, we first discuss the atomic diffusion and structural change of the HfO2/Al2O3 “superlattice” film as a function of the annealing temperature by changing the Hf/Al ratio. In a typical case of the film with HfOx/AlOx=3Å/9Å cycle, a clear superlattice peak is observed below 750 °C by XRD. Above 850 °C, a different type of crystalline structure with no superlattice peaks is observed. These results indicate that the intermixing in HfO2/Al2O3 (3Å/9Å) films occurs between 750 and 850 °C. The intermixing onset temperature increases with increasing Al content in Hf/Al ratio.To further increase the crystallization temperature, nitrogen incorporation into the film is considered, and effects of the nitrogen incorporation into HfAlOx films are studied from the structural and electrical viewpoints. The results indicate that there is a tradeoff between crystallization restriction and leakage current degradation for the nitrogen incorporation into HfAlOx films.

High-temperature stability of epitaxial, non-isostructural Mo/NbN superlattices

The effect of 1000 °C vacuum annealing on the structure and hardness of epitaxial Mo/NbN superlattice thin films was studied. The intensity of superlattice satellite peaks, measured by x-ray diffraction, decreased during annealing while new peaks corresponding to a MoNbN ternary phase appeared. The results are consistent with the Mo–Nb–N phase diagram, which shows no mutual solubility between Mo, NbN, and MoNbN. Even after 3-h anneals and a loss of most of the superlattice peak intensity, the room-temperature hardness was the same as for as-deposited superlattices. The retained hardness suggests that a residual nanocomposite structure is retained even after the formation of the ternary structure.

First Observation of Atomic Long Range Ordering in Metal-Oxide Based ZnMgO Wide Bandgap Heterostructures

We report the first observation of atomic long range ordering in the metal-oxide based wide bandgap (3.68 eV) ZnMgO alloy semiconductor thin films grown on sapphire (0001) by pulsed laser deposition. The ZnMgO system has the hexagonal wurtzite structure. The atomic ordering was deduced from the observations of forbidden x-ray diffraction peaks in the wurtzite structure. We show that under optimum thin film growth conditions, Zn and Mg preferentially occupy hcp sublattice. The relative intensity of the superlattice peak was largest for the Mg content in the range of 10%, which is about 50% of the solubility of Mg in ZnO. The observation of long range order in ZnMgO alloy semiconductor is quite important in the context of understanding issues related to the growth kinetics, surface reconstruction, and phase stability of metal-oxide based heterostructures and devices.

Structural properties of molecular beam epitaxy grown Ni/Pt superlattices

We find that the [Ni3.2nmPt1.6nm] × 15 and [Ni3.2nmPt0.8nm] × 15 multilayers are semicoherent and display a columnar morphology. From both the period of the moir’e fringes and the positions of the diffraction peaks in electronic (plan-view and crosssection geometries) and x-ray diffraction patterns, one deduces that the nickel is relaxed (at least in the error bars of all our measurements), whereas the platinum remains slightly strained (≈−1%). The interfaces are sharp; no intermixing takes place giving rise to neat contrasts in transmission electron microscopy (TEM) and to high intensities of the superlattice peaks in the growth direction in both diffraction techniques. The relaxation of the interfacial misfit occurs partially through misfit dislocations, partially through the strain of platinum. A quasiperiodic twinning occurs at the interfaces, the stacking fault which forms the twin being the most often located at the interface Pt/Ni, i.e., when a Pt layer begins to grow on the Ni layer. The simulation of the θ/2θ superlattice peak intensities takes into account the columnar microstructure. It shows that the roughness is predominantly at medium scale with a fluctuation of about 12.5% for Ni layers and negligible for Pt layers.

Short and Long-Range Ordering of (Ni0.67Cr0.33)1-xFex Alloys Under Electron Irradiation

Abstract(Ni0.67Cr0.33)1-xFex alloys with x = 0, 5 and 10 at.%, have been irradiated at temperatures between 300°C and 500°C with the electrons of a 2.5MeV Van de Graaff accelerator and of a 1MeV High Voltage Electron Microscope (HVEM). All of them have been successively ordered at short and at long range under irradiation. The transition temperature is 525°C for Ni0.61Cr0.34Fe0.05 and 445°C for Ni0.59Cr0.31Fe0.10. An Inconel sample irradiated up to 1 d.p.a. in the 1MeV microscope did not show any superlattice peak although its composition is very close to the 10% iron model alloy. The kinetics of ordering has been followed by resistivity measurements and fitted by the Dienes kinetic model.

Effects of thermal processing on (SimGen)p superlattices

In this work we report an X-ray diffraction study of thermally induced interdiffusion and strain relaxation in molecular beam epitaxy grown (SimGen)p short-period superlattices. Both rapid and furnace thermal annealings in the range 500–700 °C were used to generate structural changes in the various samples. Strain relaxation was studied by measuring the shift of (400) superlattice peaks on annealing. About half of the strain was relieved in the first few minutes of annealing. The remaining strain was relieved at a much lower rate and residual strain persisted even after several hours of heating. Also, the decay of the first order (000) superlattice peak was monitored as a function of annealing time. An initial rapid nonexponential decrease in peak intensity was observed, coincident with the sudden strain relaxation found at short annealing times. A slower, exponential decay rate was observed at longer times, so that a wavelength-dependent interdiffusion coefficient Dλ could be calculated. The variation of Dλ with the superlattice period and the effect of different substrates and buffer layers was studied. Diffusion was faster in structures alternating thin Si and thick Ge layers suggesting that migration of Si into Ge is the dominant diffusion process.

The location of tantalum atoms in Ni3(Al, Ta)

An x-ray powder diffraction method was used to determine the location of tantalum atoms in Ni3Al. A series of Ni3(Al, Ta) alloys were produced with tantalum contents ranging from 0.1 to 3.0 at.%. Fine powders with average particle sizes less than 80 μm were made from melt-spun ribbons by a grinding process. The values of the intensity of the (100) superlattice peak normalized to that of the (200) fundamental peak as a function of tantalum content agreed with the calculated values, assuming that tantalum atoms substitute on aluminum sites, not on nickel sites, and also assuming small amounts of anti-site defects exist in the ordered fee structure of Ni3(Al, Ta) alloys. It is concluded from our experiments that tantalum atoms substitute for aluminum in Ni3Al. The long-range order parameters thus calculated of the Ni3(Al, Ta) alloys are generally above 0.84 and below 0.95, except for Ni75Al24.8Ta0.2, at which composition the long-range order parameter is close to unity.

Improved Kematical X-Ray Rocking Cjrve Analyses

ABSTRACTA new boundary condition is used in the kinematical x-ray diffraction model analysis of rocking curves. The boundary condition is the dynamical reflection amplitude instead of the previously used dynamical intensity for the substrate. It is shown that this boundary condition properly accounts for the angular shift effect in the Bragg peak profile of very thin epitaxial layers and in the 0th-order superlattice peak. We present experimental and simulated rocking curves for various samples. The simulation was performed by using the dynamical diffraction theory, the kinematical model with the new amplitude boundary condition, and the kinematical model with the old intensity boundary condition. We also analyzed the strained resonant tunneling device structure and single strained quantum well samples by x-ray interference technique. Our x-ray interference results showed a good agreement with the nominal values for both experimental and simulated rocking curves.

The Location of Tantalum Atoms in Ni3Al

ABSTRACTAn X-ray powder diffraction method was used to determine the location of tantalum atoms in Ni3Al. A series of Ni3 (Al,Ta) alloys were produced with tantalum content ranging from 0.1 to 3.0 at%. Fine powders with average particle sizes less than 80 pm were made from melt-spun ribbons by a grinding process. The values of the superlattice peak intensities normalized to the intensity of the (200) fundamental peak agreed with the calculated values assuming that tantalum atoms substitute on aluminum sites, not on nickel sites. Thus it is concluded from our experiments that tantalum atoms substitute for aluminum in Ni3Al, except for Ta contents around 0.2 at% where the data can not be interpreted in terms of a simple substitution scheme.