Selected Technical Reports

Process Measurements Division
Chemical Science and Technology Laboratory
National Institute of Standards and Technology

Program Areas

Multimedia Showcase

	Synthesis and Characterization of Nanostructured Materials
	Models and Data for Semiconductor Processing
	Measurement Technology for Processing of Semiconductors
	Web CKMech - database for chemical kinetic mechanisms
	XSenkplot - graphics postprocessor for chemical kinetics simulations

1. Characterization and modeling of sodium/metal-halide based flame chemistry for production of unagglomerated metals and ceramics.
2. Development of generic methods for vapor-phase production of nanocomposites.

Approach: Our approach has been to use methods for synthesis of nanostructured materials directly in particulate form, with a special emphasis on methods that have the potential to be extended to the production of materials on a bulk scale. The vapor-phase route has been pursued as potentially the most robust generic method to supply the basic requirements for cost effectiveness in manufacture and the resulting material property. Two synergistic approaches are pursued: 1) development and modeling of new prototype methods and chemistries for materials production and investigation of their performance, including the use of advanced in situ diagnostics for characterization, 2)development and application of molecular based models for prediction of nucleation and growth of nanoscale materials.

Results and Future Plans: This years activities include development of an aerosol dynamics model for growth of nanocomposite particles. The model was specifically applied to the growth of a FeOx/SiO2 superparamagnetic nanocomposite previously synthesized by a flame process and characterized by planar laser induced fluorescence (PLIF) and Mie (elastic) scattering. The results support both the viability of this modeling approach as well as providing insight into some of the rate controlling processes occurring during growth of a nano-composite. Molecular dynamics simulation has also been used to understand intra-particle phase segregation, which has been used to explain the morphology seen in the FeOx/SiO2 nanocomposite. Work this year has also included activities in conjunction with Xerox in the formation of magnetic ferrites from a flame process.

The primary activity this year has been the implementation of a novel sodium-flame process for the production of non-oxide particles. The basic chemistry for the process involves the reaction of sodium vapor with metal halides (Na + MClx => M + NaCl), which results in the condensation of small metal clusters encapsulated in salt. The chemistry is highly generic and should be applicable to the synthesis of a wide class of materials (metals, intermetallics, ceramics). The potential strength of the process is not only its generic chemistry, but the fact that a by-product of the reaction, NaCl, serves as a coating to the particles, preventing agglomeration. Furthermore, the synthesis scheme can clearly be classified as green" as no toxic by-products are involved. Most of the effort this year has been directed toward the chemical characterization of a counter flow diffusion reactor incorporating this chemistry for the TiCl4 and BCl3 systems. We have monitored the consumption of Na via planar laser induced fluorescence (PLIF) of Na2, and the production of the particle phase by Mie scattering. A chemical model has been constructed and tested against the experimental result and shows good agreement. Basically the chemistry associated with the formation of Ti, B and TiB2 is very fast (favoring the use of turbulent flow for practical application) and consistent with sequential abstraction of halogens, prior to nucleation. Temperature plays a vital role, in that low temperatures will result in both salt and metal conucleating, which leads to smaller metal particles.

Publications:

1. Zachariah, M.R. and Tsang, W., Theoretical Computation of Thermochemistry, Energetics and Kinetics for High Temperature SixHyOz Reactions," J. Phys. Chem. 99, 5308 (1995).
2. Blaisten-Barojas, E., Liu, L., and Zachariah, M.R., Molecular Dynamics Study of Growth and Coalescence of Nanometer Particles at High Temperatures," High Performance Computing 1995, Grand Challenges in Computer Simulation, Ed. A. Tentner, Society for Computer Simulation, 228-233 (1995).
3. Blaisten-Barojas, E., Liu, L., and Zachariah, M.R., Dynamics of Nanometer SiO2 Particles and their Coalescence Characteristics," MRS Proceedings on Dynamics in Small Confining Systems 386, 173-178 (1995).
4. Zachariah, M.R., Aquino-Class, M., Shull, R.D. and Steel, E., Formation of Superparamagnetic Nanocomposites from Vapor Phase Condensation in a Flame," Nanostructured Materials 5, 383 (1995).
5. McMillin, B.K., Biswas, P., and Zachariah, M.R., In Situ Characterization of Vapor Phase Growth of Iron Oxide-Silica Nanocomposite; Part I: 2-D Planar Laser-Induced Fluorescence and Mie Imaging ," J. Mat. Res. (in press).
6. Zachariah, M.R., Shull, R.D., McMillin, B.K., and Biswas, P., In-Situ Characterization and Modeling of the Vapor Phase Growth of a Superparamagnetic Nanocomposite," ACS Symposium Proceedings Molecularly Designed Nanostructured Materials and Composites (in press).

1. Ab-initio quantum chemistry and reaction-rate-theory derived thermochemical and kinetic data for silicon-based process chemistries.
2. Development of user-friendly data handling and estimation procedures and software tools.
3. Development of software for modeling flow, chemistry, and particle contamination in prototypical thermal CVD reactors.
4. Benchmarking of key data and models against experiments.

Problem: Reactor and process design are currently limited to empirical trial and error approaches which tend to converge to semi-optimized states very slowly. For an industry, such as the semiconductor industry, where changes occur rapidly, this implies that processes are not adequately investigated prior to final implementation. This results in not only more costly, possibly poorer quality products, but also processes which may not be environmentally optimized.

Approach: Process simulation has the potential to improve the design process, and therefore, make it both more effective and more likely to have an impact on the final process within the given time constraints. The potential of process simulation has been enabled by ever-increasing computational power which has evolved to the point that very sophisticated models can be constructed for a variety of complex semiconductor processing steps. However, increasing complexity of models implies greater need for accurate thermochemical and kinetic data, not now available. Our approach is to both use and develop methods for reliably generating, from first principles, thermochemical and kinetic data suitable for process modeling. In some cases, we develop user friendly tools that will aid an experienced process engineer in generating the data necessary for modeling as well as methods and tools for efficient post-processing of simulation results. Furthermore, procedures for handling, collating, and disseminating the information in an efficient and timely manner are an important component to the task. Finally, the reliability and utility of the information we generate is both a function of the inherent quality of the data (which must be verified) and the demonstration of its utility to the user community. The latter point requires that models of prototypical processes be developed for which benchmarking studies can conducted.

Results: This year we focused on the development of fundamental thermodynamic and kinetic data through the application of ab-initio molecular orbital methods and reaction rate theory. Data has been generated on the thermochemistry of a large body of neutral fluoro-oxyhydrocarbon species (110 molecules). Such data has been compared with literature values where available and has shown excellent agreement to heats of formation ( < 10 kJ/mol). These results were used as the basis for construction of a reaction mechanism for the high temperature chemistry of fluorocarbons, which relied heavily on the use of transition state quantum chemical calculation as well as reaction rate theory methods. The data have application to both plasma etching chemistries as well as combustion of the off gases from reactors. Data obtained with these methods has also been used to determine thermochemistry and kinetics for the silicon-oxy-hydride system important to the chemistry of SiO2 film formation. Both sets of computed chemistries suggest that these methods have remarkable potential to change the character and scope of process modelling by enabling extremely rapid, relative to experiment, generation of large bodies of high quality data.

A new reaction-rate-theory computer program has been developed that solves the time dependent master equation and uses the RRKM formalism. The program enables the computation of rates for any arbitrary combination of recombination, isomerization, chemical activation and decomposition reaction channels. We have been implementing this program in a user-friendly format that should allow researchers, not fully versed in the methodology, to use the software effectively. We have also developed a first version of a program that enables more effective handling of large reaction data sets. The software will allow one to rapidly screen large reaction rate and thermodynamic data sets for the data needed for a specific modeling task.

As part of our effort to benchmark certain chemistries, we have been making high temperature kinetic and mechanistic studies for tetraethoxysilane decomposition (TEOS). This is the primary precursor used in both thermal and plasma enhanced CVD growth of silicon dioxide films. Models, currently being developed elsewhere, require fundamental data on both the kinetics of TEOS decomposition as well as qualitative mechanistic information. Results to date have determined the thermal decomposition rate of TEOS and a likely mechanism for the overall chemistry and are consistent with our quantum chemical computations.

We have developed a model for a rotating disk CVD reactor that incorporates flow, chemistry, and aerosol formation and transport. The code is a modification of the Sandia SPIN code and will be used to develop a model for particle formation during CVD. The principal goal of this work is to assess the effects of changes in reactor operating conditions on the levels of wafer contamination, some of which we expect is formed via an aerosol route. The model will be tested against an experimental spinning disk reactor currently being built for us at MIT/Lincoln Laboratories. This reactor will be suitable for both industrially realistic film growth and simultaneous optical characterization for benchmarking our model. Also conducted this year were studies of wafer contamination during processing in barrel-type CVD reactors. This involved development of a model for the transport of micron-sized particles which indicated that such reactors may be susceptible to the formation of particle attractors near the wafer surface. These attractors entrain particles of a certain size and hold them stationary in the flow. A novel treatment of the flow has involved the use of dynamical system theory to analyze particle behavior in this vicinity.

Publications:

1. Blaisten-Barojas, E., Liu, L., and Zachariah, M.R., Molecular Dynamics Study of Growth and Coalescence of Nanometer Particles at High Temperatures," High Performance Computing 1995, Grand Challenges in Computer Simulation, Ed. A. Tentner, Society for Computer Simulation, 228-233 (1995).
2. Blaisten-Barojas, E., Liu, L., and Zachariah, M.R., Dynamics of Nanometer SiO2 Particles and their Coalescence Characteristics," MRS Proceeding on Dynamics in Small Confining Systems 386, 173-178 (1995).
3. Zachariah, M.R. and Tsang, W., Theoretical Computation of Thermochemistry, Energetics and Kinetics for High Temperature SixHyOz Reactions," J. Phys. Chem. 99, 5308 (1995).
4. Bedanov, V., Tsang, W., and Zachariah, M.R., Master Equation Analysis of Thermal Activation Reactions: Reversible Isomerization and Decomposition," J. Phys. Chem. 99, 11452 (1995)
5. Amato-Wierda, C., Zachariah, M.R., and Burgess, Jr., D.R.F, Kinetic and Mechanistic Studies of the Thermal Decomposition of Tetraethoxysilane (TEOS) in a Flow Reactor Coupled to a Molecular Beam Mass Spectrometer," Proceedings of the Electrochemical Society (1995).
6. Zachariah, M.R., Tsang, W., Westmoreland, P.R., and Burgess, Jr., D.R.F., Theoretical Predication of Thermochemistry and Kinetics of Reactions of CF2O with Hydrogen Atom and Water," J. Phys. Chem. 99, 12512 (1995).
7. Allendorf, M.D., Melius, C.F., Ho, P., and Zachariah, M.R., Theoretical Study of the Thermochemistry of the Si-O-H System," J. Phys. Chem. 99, 15285 (1995).
8. Zachariah, M.R., Westmoreland, P.R., Burgess, Jr., D.R.F., Theoretical Prediction of Thermochemistry and Kinetics of Halocarbons," to appear in ACS 611, Halon Replacements: Technology and Science, Chapter 27, 358-373 (1995).
9. Burgess, Jr., D.R.F., Zachariah, M.R., Tsang, W., and Westmoreland, P.R., Thermochemical and Chemical Kinetic Data for Fluorinated Hydrocarbons," Prog. Energ. Comb. Sci. (in press).
10. Berry, R.J., Burgess, Jr., D.R.F., Nyden, M.R., Zachariah, M.R., and Schwartz, M., Halon Thermochemistry: Ab Initio Calculations of the Enthalpies of Formation of Fluoromethanes," J. Phys. Chem. 99, 17145-17150 (1995).
11. Zachariah, M.R., Westmoreland, P.R., Burgess, Jr., D.R.F., and Melius, C.F., Thermochemical Data for C1 and C2 Hydrofluorocarbons and Oxidized Hydrofluorocarbons: BAC-MP4 ab-Initio Predictions of Stable and Radical Species," J. Phys. Chem. (in press).
12. Tsang, W., Bedanov, V., and Zachariah, M.R., Master Equation Analysis of Thermal Activation Reactions: Energy Transfer Constraints on Fall-Off Behavior in the Decomposition of Reactive Intermediates with Low Thresholds," J. Phys. Chem. (in press).

Problem: Design and operation of processing reactors is changing from a completely empirical basis to one strongly dependent upon process and reactor modeling. The semiconductor industry's long term objective, reflected in the SIA's 1994 National Technology Roadmap for Semiconductors, is model-based design and control of manufacturing processes. Validation of the models is critical for their acceptance and use by the industry and demands many improvements and advances in our measurement capabilities. Although presently used reactor models simulate physical parameters well, the necessary addition of the complex chemistries found in commonly used manufacturing processes is only just beginning. This will require new measurements which couple external control variables such as pressure, gas flows, operating voltages, currents, and temperatures to spatially and temporally resolved chemical information for both the gas phase and wafer surface. To support model-based control algorithms, next generation reactors will also require new process sensing techniques, again with an emphasis on chemical information.

Approach: Data taken on systems with characteristics similar to production process reactors have the greatest value for verification of model predictions and extension of existing or development of new measurement methods. Simulation of production process reactors in the laboratory is a powerful approach to providing the research community with relevant process conditions and ranges of operating parameter variation. Reference reactors are a means to satisfy this need. They provide a common basis for comparison of results obtained by models and a well defined basis for comparison of measurement results for both existing measurement techniques and newly developed ones. In the plasma processing community the Gaseous Electronics Conference Reference Discharge Cell (GEC cell) satisfies this need for low-density, plasma-based processes. For chemical vapor deposition (CVD) processes reference reactors are not yet available, although NIST is currently engaged in developing a prototype which may serve as a reference reactor specification. Application and improvement of existing measurement techniques using radio frequency (RF) discharges in the GEC cell is a major effort. Measurement of species concentration, ionized species constituency and energy, RF voltage and current and their relation to plasma conditions, and development of measurement methods for surface physical and chemical state are areas of effort. Many of these are non-intrusive to the plasma and have potential for future application as process sensors for control applications.

Results and Future Plans: Current, voltage, and RF power characteristics of the electrical discharge are primary plasma reactor control parameters and have been the subject of much theoretical and experimental effort. An important goal is development of equivalent circuit models of the discharge and reactor that relate electrical properties (current, voltage, and impedance) to densities, energies, and fluxes of charged species. Both models and improved measurement methods and algorithms have been developed using the GEC cell. Ion energies obtained from electrical measurements and models have been compared with direct ion-energy measurements. Good agreement between the two methods have been attained.

Measurement of plasma chemical concentrations in two dimensions provides data needed to test recently developed multi-dimensional models. We have demonstrated that 2-D laser-induced fluorescence imaging of species formed by reactive plasmas in the GEC reference cell can provide such data. In this study, we employed an Ar plasma with small amounts of added CF4, O2, and Cl2, conditions commonly used for etching silicon. The plasma excitation was monitored by optical emission and by laser induced fluorescence of an excited Ar metastable state. The chemical state of the gas was monitored via the fluorescence of the CF2 radical, a reactive intermediate in the decomposition mechanism yielding etchant fluorine species. The technique of planar laser induced fluorescence (PLIF), in which a planar laser excitation beam and a gated, image intensified CCD camera are used to measure the fluorescent species, was used to determine the concentration profiles in the vertical plane of the discharge.

A number of future developments are planned. These include (1) in-situ wafer surface state measurement based on spectroscopic ellipsometry with application both to plasma etching and CVD reactors, (2) development of a prototype spinning disk CVD reactor to serve as a reference reactor, and (3) design and fabrication of a second generation plasma reactor similar to the high density plasma reactors planned for the 300 mm wafer size.

Publications:

1. McMillin, B.K. and Zachariah, M.R., "2-D Images of CF2 Density in CF4/Ar Plasmas by Laser-Induced Fluorescence in a GEC rf Reference Cell," Special Issue of the IEEE Transactions on Plasma Science (in press).
2. McMillin, B.K. and Zachariah, M.R., "2-D Imaging of CF2 Density by Laser-Induced Fluorescence of CF4 Etching Plasmas in the GEC rf Reference Cell," in Proc. of the 12th International Symposium on Plasma Chemistry 1, 539-544 (1995).
3. McMillin, B.K. and Zachariah, M.R., "2-D Laser-Induced Fluorescence Imaging of Metastable Density in Low-Presser rf Argon Plasmas with Added O2, Cl2 and CF4," J. Applied Physics (in press).
4. McMillin, B.K. and Zachariah, M.R., "Two-Dimensional Argon Metastable Density Measurements in a Radio Frequency Plasma Reactor by Planar Laser-Induced Fluorescence Imaging," J. Applied Physics 77, 5538-5544 (1995).
5. Sobolewski, M.A., "Current and Voltage Measurements in the Gaseous Electronics Conference RF Reference Cell," J. of Research of the National Institute of Standards and Technology 100, 341-351 (1995).
6. Sobolewski, M.A., "Electrical Characteristics of Argon Radio-Frequency Glow Discharges in an Asymmetric Cell," IEEE Transactions on Plasma Science (in press).

Program Areas

Multimedia Showcase

If you are interested	See	[Reports] [Related Links]
If you have any comments	Contact	[dburgess@nist.gov]
This page was last	Updated	December 14, 1996
This page is located at	URL	http://www.nist.gov/cstl/div836/836.03/technical.html