We have examined the damage produced by Si-ion implantation into strained Si1-xGex epilayers. Damage accumulation in the implanted layers was monitored in situ by time-resolved reflectivity and measured by ion channelling techniques to determine the amorphization threshold in strained Si1-xGex, (x = 0.16 and 0.29) over the temperature range 30-110°C. The results are compared with previously reported measurements on unstrained Si1-xGex, and with the simple model used to describe those results. We report here data which lend support to this model and which indicate that pre-existing strain does not enhance damage accumulation in the alloy layer.

The hydrogen contents of four natural kyanite samples were determined by N-15 nuclear reaction analysis and used to calibrate an IR spectroscopic method for a more convenient quantitative H analysis of kyanite. Hydrogen is present as the OH- ion and (expressed as ppm H2O by weight) ranges from near zero up to 230 ppm. Its content is best determined from integrated absorbance of the OH bands in the 3200-3450 cm(-1) range. Approximate concentrations can be determined from measurements of either summed peak heights or integrated areas of spectra obtained from just the two principal optical directions in the cleavage plane. The present calibration leads to estimates of the OH concentration in kyanite that are about a factor of 18 lower than the earlier calibration of Beran and Gotzinger (1987).

Olivine is an important host of hydrogen in the Earth's upper mantle, and the OH abundance in this mineral determines many important physical properties of the planet's interior. To date, natural and experimentally hydrated olivines have been analyzed by uncalibrated spectroscopic methods with large (±100%) uncertainties in accuracy. We determined the hydrogen contents of three natural olivines by 15N nuclear reaction analysis and used the results to calibrate the common infrared (IR) spectroscopic method for quantitative hydrogen analysis of olivine. OH content (expressed as parts per million H2O by weight) is 0.188 times the total integrated absorbance of the fundamental OH stretching bands in the 3750–3100 cm−1 region. The results indicate that an upward revision of some previous determinations by factors of between 2 and 4 is necessary. The most hydrous naturally occurring mantle‐derived olivine analyzed to date contains 240 ppm wt. H2O. Retrospective application of this calibration to experimentally hydrated olivines may be limited by spectral differences in some cases and by the previous use of nonpolarized IR spectra.

Paterson (1982) developed a method to determine the absolute OH concentration in minerals from the intensity of OH absorptions in the IR spectrum that has been widely applied to geophysically important earth materials in nature and in laboratory experiments. Subsequent studies have shown the desirability of mineral-specific calibrations of the IR spectrum. Here we report a new calibration of the IR spectrum of OH in olivine that implies greater absolute concentrations. Polarized infrared spectra, covering all three optic orientations were collected on three gem-quality olivines of high-pressure natural origin. Absolute H concentrations were determined by N nuclear reaction analysis at the Institut für Kernphysik, U. Frankfurt on the same samples, or pieces of the same sample demonstrated by FTIR to be homogeneous. The experimental setup involves a custom constructed shielded high vacuum sample chamber and a pre-analysis surface sputtering technique that yield background levels for H (blanks) equivalent to 2 ± 2 ppm H2O by weight. H contents of the three samples were measured as 16 ± 5, 140 ± 20 and 220 ± 20 ppmw H2O, with uncertainty derived from counting statistics and the variable concentrations measured in depth profiles. A linear correlation between integrated IR intensity and absolute H content was obtained, giving the calibration H2O (ppmw) = 0.188 ± 0.012 times the total integrated absorbance of the OH bands between 3650 and 3100 cm, per cm sample thickness. The molar absorptivity I = 29500 ± 1900 is similar to polarization-corrected values for garnet, but lower than values for pyroxenes (Bell et al 1995). This calibration yields OH concentrations higher by a factor of 2.3 than the method of Paterson (1982) applied to our polarized spectra. Greater deviations result if unpolarized spectra are used. Using the new calibration, the most hydrous mantle olivine measured to date contains 240 ppm H2O, while the experimentally determined solubility of OH in olivine at its high pressure limit (Kohlstedt et al., 1996) is approximately 0.62 wt.%. Cite as: Eos Trans. AGU, 82(47), Fall Meet. Suppl., Abstract T21C-07, 2001

A systematic evaluation was made of the performance and efficiency of tris(cyclopentadienyl)cerium, Ce(CPD), as potential dopant source in atomic layer epitaxy (ALE) and chemical vapor deposition (CVD) of Ce-doped strontium sulfide (SrS:Ce) for thin-film electroluminescent display applications. In situ growth and characterization studies were carried out, without a vacuum break, of the adsorption and associated dissociation mechanisms of Ce(CPD) within the substrate temperature window typically employed in ALE and CVD SrS:Ce. Associated findings were compared to results from tetrakis(2,2,6,6-tetramethyl-3,5-heptadionato)cerium, cerium, Ce(tmhd)(4), which was used as a comparative performance baseline. In this respect, X-ray photoelectron spectroscopy (XPS) and time-of-flight secondary ion mass spectrometry studies indicated that, within the typical thermal budget used in ALE and CVD of SrS:Ce films, the Ce(CPD) source decomposed more efficiently than its Ce(tmhd)(4) counterpart, as supported by the observation of reduced hydrocarbon-based surface contamination and a cleaner Ce phase. It was concluded that Ce(CPD) might be a better candidate than Ce(tmhd)(4) for applications as Ce dopant source in ALE and CVD SrS:Ce films. (C) 2001 The Electrochemical Society.

Results are presented from a systematic study to develop and optimize a metallorganic chemical vapor deposition process for the growth of high brightness cerium-doped strontium sulfide (SrS:Ce) thin films for electroluminescent (EL) display applications. Growth of SrS:Ce was investigated in the temperature range from 400 to 530 degrees C using strontium(2,2,6,6-tetramethyl-3,5-heptadionato)trimer (Sr(tmhd)(2) trimer), tetrakis(2,2,6,6-tetramethyl-3,5-heptadionato)cerium (Ce(tmhd)(4)), and hydrogen sulfide (H2S) as reactants. Various Sr and H2S reactant flows and associated partial pressures were examined to explore corresponding effects on the him's physical, chemical, and optical properties. Film structural and compositional properties were analyzed by Rutherford backscattering spectrometry, nuclear reaction analysis for hydrogen profiling, X-ray diffraction, X-ray photoelectron spectroscopy, scanning electron microscopy, and atomic force microscopy. EL measurements were carried out on SrS:Ce-based dielectric-sulfur-dielectric stacks. The results of these studies yielded key correlations between process parameters and film texture, grain size, and EL performance. in particular, the highest EL performance was obtained for films with a predominantly (200) orientation, grain size larger than 1.0 mu m, and Ce dopant level similar to 0.14 atom %. A brightness of 51 cd/m(2) and efficiency of 0.22 lm/W were observed, as measured at 40 V above threshold voltage and 60 Hz frequency. (C) 2000 The Electrochemical Society. S0013-4651(99)06-079-6. All rights reserved.

In this paper, results are presented from a systematic investigation which aimed to evaluate the performance and efficiency of tetrakis(2,2,6,6-tetramethyl-3,5-heptadionato)cerium(IV), or Ce(tmhd)(4), as the cerium dopant source in the atomic layer epitaxy (ALE) of strontium sulfide (SrS) thin films for thin film electroluminescent (TFEL) display applications. For this purpose, in situ growth and characterization studies of the adsorption and associated nucleation mechanisms of Ce(tmhd), on SrS surfaces were performed in a clustered system under tightly controlled ultraclean conditions. The growth experiments were carried out in a specially designed processing chamber under pulsed deposition conditions that emulated the ALE process. The samples were then transferred in situ, without exposure to air, to various characterization chambers where compositional and chemical analyses were readily performed. In particular, chemical evaluation by X-ray photoelectron spectroscopy and time-of-flight secondary ion mass spectroscopy showed incomplete decomposition of the Ce(tmhd), molecule with the observation of high carbon levels corresponding to ligands from the precursor. It is believed that these ligands sterically hinder subsequent adsorption and nucleation of the Sr source by poisoning the substrate surface, and therefore require repeated cycles of exposure to pulses of the Sr source and associated reactants to ensure their complete removal. It is thus suggested that the incomplete decomposition of the Ce(tmhd), molecule is the primary cause for its poor efficiency as Ce source precursor, an observation supported by experimental findings from actual ALE manufacturing of SrS:Ce films for TFEL applications.

SrS:Ce is an important material for full color electroluminescent (EL) flat panel displays. Using a combination of SrS:Ce/ZnS:Mn and appropriate color filters high quality full color displays have been demonstrated [1] . Major issues for commercially viable process integration of SrS:Ce are the combination of high luminance, high growth rate, and process temperatures below 600°C for compatibility with low cost glass substrates. This work describes the process development and optimization of metal-organic chemical vapor deposition (MOCVD) of SrS:Ce. MOCVD is a promising candidate for deposition of SrS:Ce because it can provide the required growth rates and allows control of crystal structure and stoichiometry. Growth of SrS:Ce was performed in the temperature range from 400°C to 530°C using Sr(tmhd) 2 , Ce(tmhd) 4 , and H 2 S as precursors. The structure of the SrS:Ce was found to be strongly dependent on the H 2 S flow. A brightness of 15 fL and an efficiency of 0.22 lm/W has been achieved (40 V above threshold voltage, 60 Hz AC). Film analysis included Rutherford backscattering (RBS), X-ray diffraction (XRD), atomic force microscopy (AFM), and EL measurements. Results on the correlation between process parameters, film structure, grain size and EL performance will be presented.

The OH content of titanite single crystals from different geological occurrences was studied by FTIR microspectroscopy and by nuclear reaction analysis (NRA). Calibrating the OH absorption band at 3485 cm(-1) against the OH content determined by NRA, the integral molar absorptivity epsilon(i) amounts to 67001 * mole(-1) * cm(-2). The H2O contents of the studied low-(Al,Fe) titanites range from 0.09 to 0.36 wt.%. The structural incorporation of OH is not correlated with the minor element content. Variable OH concentrations of single crystals from alpine fissures indicate fluctuations of the activity of hydrous components during crystallization. Titanite can be regarded as an essential ''hydrous'' mineral phase of magmatic rocks.

We investigated in detail the strain relaxation behaviour of metastable tensile-strained epilayers on Si(001) by comparing the layers before and after an annealing step using a variety of different diagnostic methods. The dominant strain-relieving mechanism is the formation of carbon-containing interstitial complexes and/or silicon carbide nanoparticles, similar to the behaviour of carbon in silicon under thermodynamical equilibrium conditions (concentrations below the solid bulk solubility limit). We did not observe any carbon out-diffusion. To grow material suitable for device applications, all carbon atoms should be incorporated substitutionally. There is only a very narrow temperature window for perfect epitaxial growth of such layers, limited on one side by the possible formation of interstitial carbon complexes and on the other side by the deterioration of epitaxial growth at low temperatures. The carbon concentration should not exceed a few per cent to avoid strain-driven precipitation.

As integrated circuit sizes decrease below 0.25 microns, device performance will no longer improve at the same rate as for past generations because of RC interconnect delay which becomes significant as compared to the intrinsic gate delay. The parallel approaches to partially address this fundamental problem are to use a lower resistance metal (i.e., copper instead of aluminum) and to use a dielectric material with a dielectric constant significantly below that of dense silica (∼4). Recently, considerable progress has been made in development of thin films of nanoporous silica (also known as aerogels or low density xerogels) for these ILD and IMD applications. Advantages of these materials include high thermal stability, small pore size, and similarity to conventional spin-on deposition processes, spin-on glass precursors and final material (silica). The dielectric constant of nanoporous silica can be tailored between ∼1 and 3 which allows its’ implementation at multiple technology nodes in integrated circuit manufacture starting with the 0.18 micron node. Research and development efforts at Nanoglass over the last several years have focused on; 1) simpler and more reproducible deposition processes, 2) a more complete understanding of processing-property relationships for this material, 3) scale-up of manufacturing to yield a range of precursor products with stability for at least six months and very high purity, and 4) working with customers to integrate this material into both aluminum/gapfill and copper/damascene process flows. Nanoglass has now developed a new process which considerably reduces the number of process steps and allows independent control of both film thickness and porosity. The current status of process and precursor development and device integration efforts for nanoporous silica is discussed.

Rutherford scattering of MeV 4He ions at forward angles has been used to determine the thickness and composition of single Si1 − xGex layers in Si. With scattering angles of about 20° the obtained depth resolution is up to 25% better than with standard RBS at glancing backward angles. The large scattering cross section at forward angles allows for the use of a small solid angle while maintaining good count rates and short acquisition times. Geometrical broadening of the energy spectra, due to the finite acceptance angle of the detector, is thus negligible. The kinematic factor is close to 1 for almost all elements in this geometry. Layers with different composition are therefore only distinguishable by the differences in spectrum height. The best accuracy for the stoichiometry is obtained by combining the measured energy loss in a layer with the areal density of the heavier element in that layer, which can be determined by standard RBS. The results are compared to RBS and X-ray diffraction measurements. The influence of multiple scattering is illustrated and a comparison to depth resolution calculations with the DEPTH code is made.

Resonant backscattering of 4He ions using the 12C(4He,4He) 12C resonance at 4.265 MeV has been used to analyze C in MBE grown Si1−yCy layers and in ion implanted Si with C concentrations of the order of one at.%. With ion channeling the ratio of substitutional to interstitial C sites in the Si1−yCy layers was determined and the obtained values for the tetragonal distortion are compared with results from X-ray diffraction measurements. The measured C amount of an ion implanted sample used for calibration was in good agreement with the nominal amount, while the range distribution was about 15% shallower than predicted by TRIM calculations. Strong deviations from the Rutherford scattering cross section are observed for scattering from Si and the influence due to background from nuclear reactions with Si on the C analysis is discussed.

Strained layer Si/Si0.79Ge0.21 superlattices consisting of 16 alternating 19.0 nm Si0.79Ge0.21 / 18.5 nm Si layers have been amorphized by Si ion irradiation, then implanted with H ions to nominal atomic concentrations of 1%, 0.1% and 0.05% within the amorphized region. Subsequent solid phase epitaxy (SPE) at a regrowth temperature of 575°C was monitored in situ by time resolved reflectivity (TRR) measurements, while changes in the H distribution were measured by elastic recoil detection analysis (ERDA). Analysis was supplemented by Rutherford backscattering spectrometry (RBS), x-ray double crystal diffraction and reflectivity (DCD/XRF) and transmission electron microscopy (TEM). TRR data reveals a decrease in the initial SPE rate in the Si substrate from 4.9 Å/sec (no H) to 2Å/sec for 1% H concentration as well as a rate decrease as the interface enters the Si/SiGe layers. TRR also indicates an increased roughness in the crystal/amorphous interface with increasing H concentration. ERDA reveals that a significant fraction of the implanted H is stable in the amorphous region for the anneal times (10-30 min) at 575°C, while in the regrown lattice the H concentration has dropped below 20 ppm, near the detection limit of the ERDA. DCD shows almost no strain in the regrown structures. TEM and RBS channeling techniques reveal degradation in the crystal quality of epitaxially regrown structures and a large concentration of strain relieving defects originating near the second deepest of eight SiGe layers in all regrown structures. XRF indicates decreasing sharpness of the regrown Si/SiGe interfaces with increasing H concentration.

A setup for sensitive hydrogen analysis with the H(15N, αγ) 12C reaction has been developed. The detector for the reaction γ-rays is a bore-hole BGO scintillation counter with an efficiency of 27%. The natural-radiation background count rate is reduced by lead and anticoincidence shielding to 25 counts/h. This detector system is used in combination with a specially designed UHV chamber which enables measurements under vacuum in the 10−10 mbar region. Various kinds of background arising from interaction of the ion beam with parts of the beam line and with contaminations on the sample surface (H, D, C) as well as possibilities for background quantification and reduction have been investigated for 15N ion energies up to 12 MeV. Profiling examples demonstrating a detection sensitivity of better than 10 at.ppm are presented.

It has been shown that the 15N nuclear reaction technique for hydrogen profiling can be used for the characterization of polymer films. With an optimized sample geometry, a depth resolution of 5 nm near the sample surface is achieved. A special γ-ray detector allows radiation damage in the sample to be minimized. Since the 15N technique discriminates between the hydrogen isotopes H and D, a suitable contrast between the different components of a polymer film is achieved by deuterating one component of the polymer. This has been utilized in the study of the surface-induced lamellar ordering within diblock copolymer films.

A newly developed measuring set-up for analysis of hydrogen in solids at a concentration level of 10 at.ppm is described. Main features are a special λ ray detector and provisions for reduction of different kinds of background. Various factors determining the sensitivity are discussed and the capability of the measuring set-up is illustrated by several examples.

The 15N nuclear reaction technique for hydrogen analysis is applied for the first time to study hydrogen depth profiles in polymer thin films. Relevant details of this technique are given and in particular the important question of depth resolution is discussed. By using an optimized geometry a depth resolution of the order of 10 nm is achieved. Energy straggling data for 15N ions in polystyrene have been deduced from measured H-profiles. The ion dose necessary for the analysis is found to be low enough that radiation damage in the sample is of minor influence. Since the 15N technique discriminates between the isotopes H and D a suitable contrast between different components of a film system is obtained by deuterating one of them. This is illustrated by several examples, including interdiffusion of polystyrene layer systems and surface enrichment of one component of a diblock copolymer.