Determination of Carbon Content in Photovoltaic Silicon Wafers: Methods, Standards, and Applications

Carbon content determination in photovoltaic silicon wafers is a critical quality control measurement that directly impacts solar cell efficiency, minority carrier lifetime, and module reliability. Excess carbon in crystalline silicon promotes defect formation, SiC precipitate growth, and wafer breakage during downstream processing. This article provides a comprehensive technical overview of carbon testing standards, sampling strategies, analytical methods, and practical considerations for photovoltaic manufacturers and accredited third-party laboratories.

Technical Overview

Carbon is one of the most significant impurities in crystalline silicon used for photovoltaic applications. It originates primarily from polysilicon feedstock, crucible graphite components, and the argon atmosphere during the Czochralski (CZ) or directional solidification (DS) ingot growth process. Because the equilibrium solubility of carbon in silicon is low (approximately 3.5 × 10¹⁷ atoms/cm³ near the melting point), excess carbon readily precipitates as silicon carbide (SiC) inclusions or interacts with oxygen and intrinsic point defects to form complexes such as carbon-oxygen pairs.

In solar cell manufacturing, elevated carbon levels are associated with several performance and yield risks, including reduced minority carrier lifetime, increased recombination activity at defects, degraded mechanical strength, and wafer or cell breakage during high-temperature diffusion and firing steps. Carbon concentrations above the saturation limit can also produce inclusions that cause wire-saw breakage during wafering. Accurate quantification of carbon content is therefore an essential element of incoming material inspection, process qualification, and failure analysis in the photovoltaic value chain.

Carbon in silicon exists in two principal forms that must be distinguished analytically: substitutional carbon atoms occupying lattice sites, which are measurable by infrared absorption spectroscopy, and total carbon, which includes precipitated SiC and carbon-rich inclusions, typically measured by combustion-based techniques. The carbon concentration is conventionally reported in units of atoms/cm³ (1 ppma ≈ 5 × 10¹⁶ atoms/cm³ for silicon).

Testing Samples

Monocrystalline Silicon Wafers

Wafers sliced from CZ-grown monocrystalline ingots, including p-type, n-type, and gallium-doped variants. Carbon in monocrystalline material is predominantly substitutional, and concentration gradients along the ingot axis must be considered when selecting sample positions. Both as-cut and polished wafers can be tested, although polished or lapped surfaces with defined thickness are preferred for infrared transmission measurements.

Multicrystalline and Cast-Mono Silicon Wafers

Wafers produced by directional solidification typically exhibit higher and less uniform carbon distribution than monocrystalline material, with concentrations increasing from bottom to top of the ingot and toward the crucible walls. Sampling should account for brick position and grain structure. These wafers more frequently contain SiC inclusions, making total carbon analysis particularly relevant.

Polysilicon Feedstock

Virgin polysilicon in chunk, granular, or rod form from the trichlorosilane or silane process. Feedstock carbon specification directly controls the baseline impurity level in the resulting ingot and wafer. Granular polysilicon from fluidized bed reactors requires special attention due to potential surface carbon contamination.

Rod Scrap, Topcuts, and Recycled Material

Reclaimed silicon streams, broken wafers, and ingot residuals reintroduced into the crucible. These materials carry elevated risk of accumulated carbon and SiC particles and should be characterized before blending into production feedstock.

Testing Items

• Substitutional carbon concentration (Cs), reported in atoms/cm³ or ppma
• Total carbon content (including precipitated SiC), reported in ppma or ppm by weight
• Carbon distribution mapping across wafer area or ingot height
• Interstitial oxygen-to-carbon ratio and carbon-oxygen complex assessment
• SiC inclusion detection and density evaluation for cast silicon
• Carbon concentration verification against incoming material specifications
• Comparison of carbon levels before and after gettering, diffusion, or annealing processes

Testing Methods

Infrared Absorption Spectroscopy (FTIR)

Fourier-transform infrared spectroscopy is the industry-standard method for measuring substitutional carbon in silicon, performed in accordance with SEMI MF1391 (Test Method for Substitutional Atomic Carbon Content of Silicon by Infrared Absorption). The technique detects the characteristic carbon absorption band at approximately 607 cm⁻¹ (16.46 µm), measured in transmission mode using a carbon-free float-zone silicon reference. Calibration coefficients defined in the standard convert peak absorbance to concentration, with a typical detection limit near 5 × 10¹⁵ atoms/cm³ (0.1 ppma) for double-side polished wafers of standard thickness.

Sample requirements include parallel polished surfaces, thickness typically between 0.4 mm and 4 mm (thicker samples improve sensitivity), and room-temperature measurement conditions. Peak deconvolution is required to separate the carbon band from the overlapping two-phonon lattice absorption and the oxygen-related band at 515 cm⁻¹. Multicrystalline samples with grain boundaries and high dislocation density require careful interpretation, as free-carrier absorption and scattering can distort the baseline.

High-Frequency Combustion Infrared Absorption

For total carbon determination, including precipitated SiC and carbon inclusions, samples are combusted in an induction furnace with accelerators, and the evolved CO₂ (or CO) is quantified by infrared detection. The procedure follows the general principles of ASTM E1019 and related combustion analysis standards. This method is suitable for as-cut wafers, feedstock chunks, and materials where infrared transmission is impractical. Detection limits are on the order of a few tenths of a ppm by weight, and results are validated using certified reference materials.

Secondary Ion Mass Spectrometry (SIMS)

SIMS provides depth-profiled carbon measurement with detection limits down to approximately 10¹⁶ atoms/cm³. It is used primarily for research, failure analysis, and verification of near-surface carbon contamination from processing media or packaging. SIMS complements FTIR where spatially resolved information is required.

Sample Preparation and Quality Assurance

Reliable results depend on controlled preparation: surfaces must be cleaned to remove organic residues, wafer thickness must be measured accurately at the analysis point, and reference samples must be traceable. Accredited third-party laboratories typically report results with measurement uncertainty, apply round-robin validated procedures, and maintain conformance with SEMI and ASTM documentary standards as well as laboratory accreditation requirements such as ISO/IEC 17025.

Applications

• Incoming quality control of polysilicon feedstock against purchaser carbon specifications
• CZ and DS ingot growth process optimization, including hot-zone design, crucible coating, and argon flow management
• Grading and segregation of ingots by height to contain high-carbon material in lower-value blocks
• Yield improvement programs targeting wire-saw breakage caused by SiC inclusions
• Solar cell efficiency studies correlating carbon-related defects with minority carrier lifetime and conversion efficiency
• Qualification of recycled silicon and new crucible or graphite supplier materials
• Failure analysis of wafer cracking, cell breakage, and shunting in finished modules
• Support for research on high-efficiency architectures such as TOPCon, HJT, and n-type cells, where defect control is more stringent

FAQ

What carbon concentration is considered acceptable for photovoltaic silicon wafers?

Typical procurement specifications limit substitutional carbon to below 1 × 10¹⁷ atoms/cm³ (approximately 2 ppma) for monocrystalline material, with tighter limits for high-efficiency n-type products. For multicrystalline silicon, total carbon levels up to several ppma may be tolerated depending on the cell technology, but concentrations approaching the solubility limit increase the risk of SiC precipitation and mechanical defects.

Why measure substitutional and total carbon separately?

FTIR measures only carbon atoms incorporated substitutionally into the lattice, while combustion analysis captures all carbon forms, including SiC precipitates. A wafer may show acceptable substitutional carbon yet contain harmful SiC inclusions detectable only through total carbon or inclusion inspection. Both measurements together give a complete picture of material quality.

Can as-cut wafers be tested directly by infrared spectroscopy?

As-cut surfaces scatter infrared radiation severely, so samples must be polished or lapped on both sides for FTIR measurement. As-cut or textured wafers are better suited to combustion analysis or cross-sectional techniques when total carbon is the parameter of interest.

How does carbon affect solar cell performance?

Carbon acts as a recombination center when complexed with oxygen and vacancies, reducing minority carrier lifetime and diffusion length. SiC precipitates can cause soft shinethrough, shunting in finished cells, and mechanical weakness leading to breakage during handling, stringing, and module lamination.

What standards govern carbon measurement in silicon?

The principal documentary standards include SEMI MF1391 for substitutional carbon by infrared absorption, ASTM E1019 for combustion-based carbon determination, and the corresponding regional standards maintained by standards organizations such as GB and DIN for silicon material characterization. Testing should be performed by laboratories accredited to ISO/IEC 17025 with demonstrated competence in these methods.