Identification of formation of amorphous Si phase in SiOxNy films produced by plasma enhanced chemical vapor deposition

Basic Information

• Paper ID: 2510.14701
• Title: Identification of formation of amorphous Si phase in SiOxNy films produced by plasma enhanced chemical vapor deposition
• Authors: M. V. Voitovych, A. Sarikov, V. O. Yukhymchuk, V. V. Voitovych, M. O. Semenenko
• Classification: cond-mat.mtrl-sci cond-mat.mes-hall
• Research Institutions: Institute of Semiconductor Physics, National Academy of Sciences of Ukraine, and other institutions
• Paper Link: https://arxiv.org/abs/2510.14701

Abstract

This study investigates the formation characteristics of amorphous silicon (a-Si) phase inclusions in silicon-rich silicon oxynitride (SiOxNy) films prepared by plasma-enhanced chemical vapor deposition (PECVD) using combined Raman scattering and infrared absorption spectroscopy. Raman scattering results demonstrate the presence of a-Si phase in the investigated films when the relative silicon content exceeds a threshold of approximately 0.4. The a-Si content correlates with the hydrogen concentration in the films, with hydrogen presence detected through characteristic infrared absorption bands of Si-H bending vibrations (~660 cm⁻¹) and stretching vibrations (composite bands in the ~1900-2400 cm⁻¹ range). The study proposes that infrared spectroscopy combined with low-wavenumber region analysis serves as an effective method for identifying phase composition in silicon-rich silicon oxynitride films.

Research Background and Motivation

Problem Background

1. Material Significance: SiOxNy films are of considerable importance in modern microelectronic and optoelectronic device fabrication, with their structure determining material properties
2. Compositional Complexity: As silicon content increases, the morphology and structure of SiOxNy films undergo fundamental changes, potentially forming variable-sized amorphous silicon clusters embedded in a dielectric matrix
3. Detection Challenges: Existing phase composition analysis methods have limitations in identifying amorphous silicon phases

Research Motivation

1. Theoretical Refinement: Existing theoretical models primarily consider random bonding distributions of Si, O, and N atoms and require updating to include the possibility of amorphous silicon clusters
2. Practical Requirements: Providing monitoring methods for PECVD preparation of films with specific phase compositions
3. Methodological Innovation: Developing simple, non-destructive phase composition identification techniques

Core Contributions

1. Established Threshold Relationships: Determined the critical threshold at approximately 0.4 relative silicon content above which amorphous silicon phase formation begins
2. Proposed Novel Detection Method: Demonstrated that the absorption band at ~660 cm⁻¹ in infrared spectroscopy can be used to identify amorphous silicon phases
3. Revealed Correlations: Discovered direct correlation between amorphous silicon content and hydrogen concentration in films
4. Developed Analysis Techniques: Established phase composition identification methods based on low-wavenumber region analysis of infrared spectra
5. Provided Theoretical Support: Furnished experimental evidence for understanding phase composition formation mechanisms in PECVD-prepared silicon oxynitride films

Methodology Details

Experimental Design

Sample Preparation:

• Nine SiOxNy films with different compositions prepared using PECVD
• Thickness: 300±5 nm
• Substrates: Double-side polished boron-doped CZ silicon wafers (for infrared testing) and 1 mm thick sapphire plates (for Raman testing)
• N₂O/SiH₄ flow rate ratio range: 0.06-9

Characterization Techniques

Raman Spectroscopy:

• Excitation wavelength: λ = 457 nm
• Power density: <10³ W/cm² (to avoid structural damage)
• Equipment: MDR-23 spectrometer equipped with Andor iDus 401A CCD detector
• Normalization: Based on amorphous silicon band (~480 cm⁻¹) intensity

Infrared Spectroscopy:

• Measurement range: 400-4000 cm⁻¹
• Resolution: 2 cm⁻¹
• Number of scans: 100
• Measurement accuracy: ~0.5%
• Equipment: PerkinElmer BX-II spectrometer

Data Processing Methods

Spectral Deconvolution: Mathematical deconvolution of composite absorption bands using Gaussian functions to separate contributions from different Si-H bonding configurations:

• H1 (2252±2 cm⁻¹): H-Si(O₃) complex
• H2 (2150±4 cm⁻¹): H-Si(Si₂O) complex
• H3 (2050±4 cm⁻¹): H-Si(Si₃) complex

Bond Length Calculation: Si-H bond lengths calculated using empirical formula:

νSi-H (dSi-H)³ = 7074 cm²

Experimental Setup

Sample Parameters

Evaluation Metrics

• Raman peak position and full width at half maximum
• Infrared absorption band integrated intensity
• Hydrogen concentration (calculated from infrared absorption)
• Correlation between relative silicon content and amorphous silicon phase content

Experimental Results

Raman Spectroscopy Analysis

Key Findings:

1. Characteristic Peak Identification: Two asymmetric bands observed at ~476 cm⁻¹ and ~660 cm⁻¹
2. Phase Composition Analysis: The 476 cm⁻¹ peak corresponds to the TO mode of amorphous silicon with full width at half maximum Γ₁ = 66 cm⁻¹
3. Threshold Effect: Significant increase in TO band integrated intensity when relative silicon content exceeds 0.4

Infrared Spectroscopy Analysis

Si-H Stretching Vibration Region (1900-2400 cm⁻¹):

• Hydrogen concentration range: ~10²¹-10²² cm⁻³
• Shows upward trend with increasing silicon content, with saturation observed in high-silicon-content samples

Si-H Bending Vibration Region (~660 cm⁻¹):

• Peak shift: From 665 cm⁻¹ (sample #3) to 635 cm⁻¹ (sample #9)
• Intensity variation: Positive correlation with relative silicon content
• Key Finding: This absorption band exhibits linear relationship with Raman TO mode intensity

Quantitative Relationships

Threshold Characteristics:

• Excess silicon threshold: ~0.2
• Linear Relationship: Above the threshold, amorphous silicon content increases linearly with excess silicon concentration

Related Work

Theoretical Foundation

1. Random Bonding Model: Traditional theory primarily considers mixed Si-O tetrahedra and Si-N pyramids
2. Phase Separation Theory: Formation mechanisms of silicon clusters in silicon-rich films
3. Spectroscopic Characterization Methods: Limitations of conventional methods such as X-ray photoelectron spectroscopy and grazing-incidence X-ray diffraction

Innovation of This Study

Compared to existing work, this study is the first to systematically apply low-frequency region infrared spectroscopy analysis to quantitative identification of amorphous silicon phases and establish clear threshold relationships.

Conclusions and Discussion

Main Conclusions

1. Threshold Mechanism: Confirmed relative silicon content of 0.4 as the critical value for amorphous silicon phase formation
2. Detection Method: Established amorphous silicon phase identification method based on infrared spectroscopy absorption band at ~660 cm⁻¹
3. Structure-Property Relationship: Revealed direct correlation between hydrogen content and amorphous silicon phase content
4. Practical Value: Provided simple and effective in-situ monitoring means for PECVD process optimization

Limitations

1. Sample Range: Only investigated films of specific thickness (300 nm)
2. Substrate Effects: Insufficient exploration of substrate influence on phase formation
3. Kinetic Mechanisms: Lack of detailed analysis of phase formation kinetic processes

Future Directions

1. Theoretical Model Updates: Incorporating amorphous silicon clusters into thermodynamic phase diagram models
2. Process Optimization: Developing customized preparation processes based on phase composition control
3. Application Expansion: Exploring potential applications in photovoltaic and optoelectronic devices

In-Depth Evaluation

Strengths

1. Methodological Innovation: First systematic use of low-frequency infrared spectroscopy analysis to identify amorphous silicon phases; method is simple and non-destructive
2. Experimental Sufficiency: Dual characterization techniques (Raman and infrared) with mutual verification enhance conclusion reliability
3. Quantitative Analysis: Established clear threshold relationships and quantitative correlations with practical value
4. Theoretical Contribution: Provided important experimental evidence for understanding phase formation mechanisms in PECVD films

Weaknesses

1. Mechanism Explanation: Lacks in-depth theoretical analysis of microscopic mechanisms for amorphous silicon phase formation
2. Parameter Range: Relatively limited experimental parameter range; generalizability requires verification
3. Application Verification: Lacks performance verification in actual devices

Impact

1. Academic Value: Provides new phase composition analysis methods for materials science
2. Practical Value: Offers convenient quality control means for semiconductor industry
3. Reproducibility: High standardization of experimental methods, easily reproducible

Applicable Scenarios

1. Industrial Production: In-situ monitoring and quality control of PECVD processes
2. Research Applications: Phase composition design of novel silicon oxynitride materials
3. Device Optimization: Material performance regulation for photovoltaic cells and optoelectronic devices

References

The paper cites 28 important references covering various aspects of SiOxNy film preparation, characterization, and applications, providing solid theoretical foundation and experimental reference for the research.

Overall Assessment: This is an experimental research paper with significant contributions to materials science. Through systematic spectroscopic investigation, the authors established a novel method for identifying amorphous silicon phases in PECVD-prepared silicon oxynitride films, providing valuable tools for understanding and controlling phase composition of these important functional materials. The research methodology is scientifically rigorous, and the results possess important theoretical significance and practical value.