Levels Of Metal Impurities Research Articles

Infrared-reflection characterization of sintered SiC thermoelectric semiconductors with the use of a four-component effective medium model

Infrared (IR) reflection spectra of sintered SiC thermoelectric semiconductors are studied as a function of doping concentration as well as of sintering temperature. Attention is focused on the changes in the reststrahlen band. The peak and dip frequencies remain constant, implying we are probing stable optical phonon mode in the crystalline grains through the interaction with IR photons. In contrast, variation of spectral shape, including the peak and background reflectance seems to give valuable information regarding the quality of the sintered samples and dopant distribution within the crystal grains and grain boundaries. Measured spectra are analyzed by the use of a four-component effective medium model. In this model we assume the presence of: (1) pores, (2) intergranular materials, (3) crystalline grains, and (4) metallic inclusions (at higher doping levels of metallic impurities), with volume fractions fi(i=1–4), respectively. For frequency dependent complex dielectric functions we employed: (a) a single oscillator model appropriate to the single crystal for the crystalline grains, (b) a Drude model for the metallic inclusions, and assumed (c) another single oscillator model for the intergranular material. The parameters of the latter are determined to give a best fit to the spectrum of a low doping level sample. This model gives excellent overall fit to measured spectra with the ratio f2/f3 and the damping coefficients for the oscillators as adjustable parameters when the experimental estimates for the fraction of pores f1 and metallic inclusions f4 are used. This is remarkable in view of the situation that similar data could not, so far, be explained in terms of the simple Maxwell–Garnet model. The results suggest that careful analyses of spectral shape of IR reflection in the optical phonon resonance region can be used to get information on the intergranular to grain ratio or on the damping coefficients, which are related to the quality of sintered samples or to the distribution of doped impurities among grains and grain boundaries. The background reflectance also give information on the presence of metallic inclusions although the effects of scattering may affect the interpretation. These results combined with the possibility of local probing seems to imply that IR reflectance can offer a versatile means to check the quality of the sintered samples.

The Mineralogy of Bronze Age Copper Ores from the British Isles: Implications for the Composition of Early Metalwork

Numerous analytical studies during the latter half of this century have contributed to the compilation of a large compositional database of Early to Middle Bronze Age copper‐based artefacts revealing distinctive impurity patterns which appear to change over time. However, attempts to relate these data to copper ore sources proved problematic in the absence of firm evidence for the location of prehistoric copper mines. Over the last fifteen years this situation has changed dramatically with the discovery of numerous Early and Middle Bronze Age copper mines in England and Wales. This study is the first attempt at a comprehensive mineralogical survey of the principal mines investigated to date in order to define the likely composition of the copper ores as mined in antiquity for comparison with the artefact database. The study suggests that the majority of these mines can only have produced essentially pure copper. Only one mine, Ross Island, is likely to have produced copper with a significant level of impurities. The relative purity of the known ore sources is contrasted with significant levels of various metallic impurities among the analysed artefacts, leading to the conclusion that metal circulation and mixing may have been more extensive than previously thought even during the earliest part of the Bronze Age.

Electrochemical and Radiochemical Study of Copper Contamination Mechanism from HF Solutions onto Silicon Substrates

The mechanism of copper contamination of silicon wafers from dilute HF solutions containing ultratrace levels of metallic ion impurities, was investigated using a new electrochemical cell, which proved to act as a very efficient sensor for in situ characterization. Upon copper contamination, the open‐circuit potential was observed to shift rapidly toward more positive values at a rate nearly proportional to the copper concentration. All potential/time curves tend to reach a plateau, while quantitative measurements using radioactive tracers revealed that during a few tens of minutes, copper ions were continuously reduced on the silicon surface. Results are interpreted in terms of the mixed‐potential theory and lead to the conclusion that copper nuclei act as a catalyst which enhances the cathodic activity for proton reduction. The model was supported by atomic force microscopy observations which showed the initiation of corrosion pits around the nuclei.

Lead from Carthaginian and Roman Spanish Mines Isotopically Identified in Greenland Ice Dated from 600 B.C. to 300 A.D.

The pollution history of the atmosphere of the Northern Hemisphere is recorded in the levels of heavy metal impurities in Greenland ice. The possibility also exists of using natural variations in the abundances of lead isotopes to trace the source of this pollution. Lead isotopes have now been measured in ancient Greenland ice with a lead concentration as low as 0.9 pg/g. The results show a depression in the 206Pb/207Pb ratio between 600 B.C. and 300 A.D., giving unequivocal evidence of early large-scale atmospheric pollution by this toxic metal. This ratio changes from ∼1.201 in ∼8-kyr-old ice to ∼1.183 about 2 kyr ago. Isotopic systematics point to the mining districts in southwest and southeast Spain as the dominant sources of this lead, giving quantitative evidence of the importance of these mining districts to the Carthaginian and Roman civilizations. Lead with a Rio Tinto-type signature represents ∼70% of the lead found in Greenland ice between ∼150 B.C. and 50 A.D. after correcting for the contribution from rock dust indexed to aluminium concentrations.

5618655 Process of reducing trace levels of metal impurities from resist components : Davidson James M Corydon, IN, UNITED STATES assigned to Olin Corporation

5618655 Process of reducing trace levels of metal impurities from resist components : Davidson James M Corydon, IN, UNITED STATES assigned to Olin Corporation

Copper Contamination Mechanism of Silicon Substrates from HF Solutions

The contamination of silicon wafers from dilute HF solutions containing ultratrace levels of metallic ion impurities is a subject of constant interest. The mechanism of copper electroless deposition from HF onto monocrystalline silicon was investigated using a new electrochemical cell, which proved to be a very sensitive detector for in situ characterization of silicon surfaces. Upon addition of copper trace amounts, the open-circuit potential was observed to shift rapidly towards more positive values at a rate nearly proportional to the copper concentration. All potential/ time curves tend to reach a limiting value of the potential, while quantitative measurements of radioactive tracers revealed that during a few tens of minutes, copper ions were continuously reduced on the silicon surface. Electrochemical potentials and voltammetric measurements were interpreted in terms of the mixed potential theory and led to the conclusion that copper nuclei act as a catalyst which enhances the cathodic activity for protons reduction. The model was supported by AFM observations which demonstrated the initiation of corrosion pits around the nuclei.

Carbothermally prepared nanophase SiC/ Si3N4 composite powders and densified parts

Nanophase SiC/Si3N4 composite powders were synthesized by the carbothermal nitridation of SiO2. These powders have desirable characteristics of high quality with oxygen contents on the order of 1.5 to 2 wt. %, surface area of ∼ 10 m2/g, submicron α-Si3N4, low metallic impurity levels, and a homogeneous distribution of the nanophase SiC phase. High-resolution TEM analysis has shown that the content and size of the nanophase SiC can be varied from 0.5 to 50 wt. % and 25 to 500 nm, respectively, through proper control of raw materials and reactor conditions. To determine how the nanophase SiC reinforcement affects the mechanical properties of Si3N4, densified components were fabricated using both pressureless and pressure-assisted densification methods. TEM analysis revealed that the nanophase SiC particles are distributed both intergranularly and intragranularly throughout the Si3N4, matrix. By controlling the sintering additive package and the sintering conditions, the ratio of inter– to intragranular SiC can be adjusted. Mechanical property measurements at elevated temperatures showed a dramatic improvement in high-temperature strength and creep resistance over components made with commercially available powders.

A New Potentiometric Sensor for the Detection of Trace Metallic Contaminants in Hydrofluoric Acid

Detection of ultratrace levels of metallic ion impurities in hydrofluoric acid solutions was demonstrated using a silicon‐based sensing electrode. The sensor's operation principle is based on direct measurements of the silicon open‐circuit potential shift generated by the charge‐transfer reaction between metallic ions and the silicon‐based sensing surface. For instance, the silicon‐based sensor is capable of detecting parts per‐trillion to parts‐per billion level of ions in HF solutions with a detection sensitivity of ca. +150 mV shift per decade change of . The new sensor can have practical applications in the on‐line monitoring of microelectronic chemical processing.

Improving Metal Oxide Semiconductor Device Performance Through the Use of Ion Exchange‐Purified HF

Ion‐exchange purification systems were employed at two different semiconductor metal oxide semiconductor (MOS) wafer fabs to purify the dilute hydrofluoric acid solutions used in cleaning silicon wafers. One fab performed the cleans in immersion baths while the other fab used a spray cleaner. Both fabs observed improvements in device performance, as measured in split‐lot gate oxide integrity tests; after they began using the ion exchange‐purified . One fab also observed a significant improvement in refresh characterization time, and a 5% yield improvement at multiprobe. These improvements are believed due to the lower levels of metallic impurities in the ion exchange‐purified . Impurity levels of over 30 elements in the treated are routinely below 1 ppb.

Total Reflection X-Ray Fluorescence Spectroscopy: Analysis Of Gaas And InGaAs

AbstractThis article explores applications of total reflection x-ray fluorescence (TRXRF) to GaAs processes. The applications include determination of surface contamination and InGaAs layer thicknesses. Surface contamination can deteriorate device performance and can occur in starting substrates and in subsequent processing. We demonstrate that TRXRF is a quick, nondestructive method for identifying sulfur contamination on incoming wafers and low levels of metallic impurities from device fabrication. Variable angle TRXRF has been used to determine heterostmeture film thickness, measuring films of InGaAs on GaAs as thin as 4 nm.

Near‐surface damage and contamination of silicon following electron cyclotron resonance etching

Damage and contamination produced after electron cyclotron resonance (ECR) etching of Si using CF4 gas has been studied using electrical characterization, Rutherford backscattering spectroscopy (RBS), secondary ion mass spectroscopy (SIMS), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and etch pit density measurement techniques. Due to the small dc self‐bias voltage generated across the plasma sheath, ECR etching is expected to produce low damage and contamination levels. RBS measurements show that ECR etching does indeed produce less structural damage than that produced by conventional reactive ion etching (RIE). It is found that the damage and contamination levels from an ECR etching process are actually reduced by the addition of radio frequency (rf) power to the wafer. The metallic impurity levels are shown to be greatly reduced by covering the stainless steel wall of the ECR source near the resonance region with an anodized Al liner. The plasma density in the resonance region of the reactor during ECR processes is much higher than that during RIE processes. Therefore, the ECR processes produce heavy metal contamination, which is mainly from the portion of the stainless steel wall of the reactor in contact with the plasma. Schottky diodes fabricated on the etched samples exhibit high leakage currents implying some damage and/or impurities are present in the near‐surface region. Relationships that exist among the generation current of the metal–oxide–silicon (MOS) capacitors, the etch pit density and the metallic impurity level were studied. Some wafers were exposed to an Ar ECR/RIE plasma to compare the effects of pure physical sputtering and ion‐assisted chemical etching, as when CF4 was used. A possible explanation for the observed behavior is given.

Studies of impurity behaviour in TFTR

Central medium and low Z impurity concentrations and Zeff have been measured by X-ray spectrometry in Tokamak Fusion Test Reactor discharges during three periods of operation. These were (1) startup period, (2) Ohmic heating, and (3) Ohmic heating portion of the two neutral beam periods, distinguished mainly by different vacuum vessel internal hardware and increasing plasma current and toroidal field capability. Plasma parameters spanned minor radius a = 0.41−0.83 m, major radius R = 2.1−3.1 m, current Ip = 0.25−2.0 MA, line averaged electron density n̄e = (0.9–4.0) × 1019 m−3, and toroidal magnetic field BT = 1.8−4.0 T. The metal impurities came mostly from the limiter. At low densities titanium or nickel approached 1% of ne during operation on a TiC-coated graphite or Inconel limiter, respectively. Lower levels of Cr, Fe, and Ni (⪅0.1%) were observed with a graphite limiter at similarly low densities; these elements were removed mainly from stainless steel or Inconel hardware within the vacuum vessel during pulse discharge cleaning or plasma operation on an Inconel limiter and then deposited on the graphite limiter. Hardware closest to the plasma contributed most to the deposits. Subsequent discharges slowly eroded the deposits from the limiter and reduced the metal impurity levels in the plasma. Both total low Z and metal impurity concentrations, relative to ne, decreased approximately exponentially with ne and increased with Ip. The metals showed much larger variations than did low Z elements, dropping to ≳10−5 ne at high density. Low Z impurities, mainly carbon and oxygen, ranged from 10% of ne at low ne to 3% at high ne and usually dominated Zeff and power loss. In low density plasmas with a TiC-coated graphite or Inconel limiter, the respective metals contributed 30–50% of Zeff and ⪅10–20% of the central input power. The metal contributions to Zeff and radiated power were much lower with the graphite limiter.

Universal alignment of transition metal impurity levels in III–V and II–VI compound semiconductors

Existence of a universal energy reference level E HO (Hasegawa-Ohno's hybrid orbital energy level) is pointed out for transition metal (TM) deep levels in tetrahedral semiconductors. When viewed from this reference level, each TM energy level E(TM) in major III–V and II–VI compound semiconductor satisfies a simple universal relation of E( TM) = E HO ± ( m + 1 2 ) Δ E with m = 0, 1, 2, --- and the empirical energy separation quantum ΔE = 0.244 eV for II–VI compounds and 0.240 eV for III–V compounds. The so-called vacuum level pinning of TM levels is not generally valid. The band line-up based on the new TM level alignment is in excellent agreement with experiment.

On a direct connection of the transition metal impurity levels to the band edge discontinuities in semiconductor heterojunctions

Our recent proposal of using the transition metal impurity levels to predict the isovalent heterojunction (HJ) band-edge discontinuities is further discussed. It is shown that for Ga1−xAlxAs/GaAs heterojunctions most of the recent discontinuity data follow within experimental error the prediction of the ΔEcb:ΔEvb= 0.64:0.36 discontinuity ratio derived from the Fe2+ level position in Ga1−xAlxAs compound. Predictions of valence-band discontinuities for the other III–V and II–VI HJ systems are also given.

ChemInform Abstract: IMPROVED HIGH‐PURITY ARC‐FURNACE SILICON FOR SOLAR CELLS

AbstractDurch Reduktion von SiO2 mit Kohle im 100 kW Lichtbogenofen dargestelltes Si besitzt eine B‐Verunreinigung von 1.8 ppma.

Improved High‐Purity Arc‐Furnace Silicon for Solar Cells

Single‐crystal silicon solar cells having efficiencies within 1% of those prepared in semiconductor‐grade silicon have been fabricated in twice‐recrystallized, high purity metallurgical silicon. Produced in a 100 kW experimental arc furnace by carbothermic reduction of pure silica with pelletized carbon black, the best “solar‐grade” silicon contained 50–100 ppmw Al, 50–100 ppmw Fe, 10 ppmw Ti, 1.8 ppma B, and 2.1 ppma P. Following two Czochralski (Cz) recrystallizations, the metal impurity levels in the final ingot were indistinguishable from those in once Cz‐recrystallized semiconductor‐grade silicon, while the boron and phosphorus contents were 1.5 ppma and 0.6 ppma, respectively. The average efficiency of 4 cm2 test solar cells prepared in twice‐recrystallized “solar‐grade” materials was 13.7%, compared to 14.4% for semiconductor‐grade controls. The average efficiency of unencapsulated, production‐type 100 mm diam solar cells made in this material was 11.8%. Compared with earlier experiments, this represents a nearly 2% improvement in efficiency, which is attributed largely to the factor of two lower boron content in the starting silicon.

Magnetic Properties of Gem-Quality Synthetic Diamonds

Measurements of the magnetic properties of four gem-quality synthetic diamonds indicate that higher levels of permanent magnetism can be induced in the synthetics than could be induced in a number of gem-quality natural stones. As received, the colored and included synthetics showed much higher levels of magnetism than the natural stones, but one near-colorless synthetic had less magnetism than many of the natural stones. Most of the magnetism of the natural stones is due to surface contamination, which can only be removed by rigorous acid cleaning. After cleaning, the natural diamonds were not only less magnetic than all of the synthetic stones, but they were also more resistant to demagnetization. Significant levels of metallic impurities were detected in most of the synthetics. These results on a limited number of samples suggest that natural diamonds can be distinguished from synthetics on the basis of their magnetic properties, but that in the case of near-colorless, inclusion-free stones, such tests may require sophisticated instrumentation.

Zone refining of rare earth metals: Lanthanum, cerium and gadolinium

Three rare earth metals were zone refined under low contamination conditions. Where possible, effective distribution coefficients for the major impurities were determined, and these were used to estimate the ultimate potential for zone refining as a purification technique for the rare earth metals. It is concluded that zone refining will be of most value for refining with respect to certain common metallic impurities. The metallic impurity levels at one end of a zone-refined gadolinium rod are believed to be the lowest ever recorded for a rare earth metal.

On the theory of the deep levels of transition metal impurities in semiconductors

The scattering of band electrons accompanied by the excitation of an unfilled d shell of an impurity atom is considered to be the origin of the deep levels associated with transition-metal impurities in semiconductors. In order to describe the energy states of the system, the Anderson model, generalized for the case of the electron spectrum with the gap, is used. The activation-energy dependence of the parameters of the model is found. It is shown that this mechanism is able to explain the existence of deep levels with symmetries which coincide with that of an unfilled impurity shell.

The effect of base-metal impurities on the surface tension of mercury

The surface tension of mercury in air has been found, by the drop-weight method, to be 476.0 ± 2.0 dynes/cm at 25°C. The values of the surface tensions of seven different grades of mercury, which varied from commercial to polarographic, were measured and found to have the same values (within experimental error). The sample grades were analyzed and were found to contain very similar base metal impurity levels, i.e., <2 ppm of mercury. The “dulling” of mercury, and its “tailing” on glass surfaces in air is apparently due to the formation of base-metal oxides. Even very pure mercury exposed to dry air (uncontaminated) will gradually “dull,” and “tail” on glass, after long exposure times. However, when mercury surfaces are kept in vacuo, or in an inert gas, they retain their metallic lustre indefinitely. It is possible that base-metal impurities as high as 5–10 ppm of mercury can be tolerated if the mercury surface is kept in vacuo, or in an inert gas.