Production of Upgraded Metallurgical Grade (UMG) silicon for a low-cost high-efficiency and reliable PV technology

This paper summarizes the comprehensive development of Ferrosolar's upgraded metallurgical grade (UMG) silicon technology, demonstrating that through optimized purification, doping, and defect engineering, UMG-Si can produce high-efficiency, reliable, and environmentally superior solar cells and modules that are competitive with conventional polysilicon.

The Gist

Imagine you want to build a house. You have two choices for your bricks:

1. The "Gold Standard" Bricks: These are made in a super-clean, high-tech factory. They are perfect, but they cost a fortune to make because the process uses a massive amount of electricity and expensive chemicals.
2. The "Recycled" Bricks: These are made from leftover industrial materials. They are cheaper and greener to make, but they usually have more cracks and dirt inside, making them weaker and less efficient for a fancy house.

For years, the solar industry only used the "Gold Standard" bricks (called Polysilicon) because the "Recycled" ones (called Metallurgical Silicon) were too dirty to make good solar panels.

This paper tells the story of a team of scientists and engineers who decided to try a different path. They asked: "What if we could clean up the 'Recycled' bricks just enough to make them work as well as the Gold Standard, but for half the price and with less pollution?"

Here is the story of how they did it, explained simply:

1. The Raw Material: The "Dirty" Silicon

The team started with Metallurgical Grade Silicon. Think of this as a lump of clay that has been used to make steel. It's full of impurities like Boron and Phosphorus (like dust and grease on the clay). In the past, this was useless for solar panels.

2. The Cleaning Process: The "Laundry"

Instead of melting the clay down and rebuilding it from scratch (which is what the expensive Gold Standard method does), they used a series of "metallurgical" tricks to clean it up.

• Slagging: Like skimming the foam off a boiling pot of soup to remove the bad stuff.
• Vacuum Evaporation: Like using a vacuum cleaner to suck out the light, smelly impurities.
• Directional Solidification: Like freezing water slowly so the ice forms perfectly, pushing the dirt to the edges where it can be cut off.

The result? A new material called UMG-Si (Upgraded Metallurgical Grade Silicon). It's still not perfectly pure, but it's clean enough to be useful.

3. The "Tuning" Problem: Fixing the Electrical Flow

Even after cleaning, the material had a problem. It had two types of "dirt" (impurities) that fought against each other electrically. Imagine a highway where some cars are trying to drive forward and others backward; traffic jams would happen, and the solar panel wouldn't generate much power.

The Solution: They added a tiny pinch of Gallium (a special metal) to the mix. Think of this as adding a traffic cop who directs all the cars to move in the same direction. This made the electrical flow smooth and consistent from the top to the bottom of the silicon block.

4. The "Magic" Cleanup: Defect Engineering

Even with the traffic cop, there were still tiny invisible holes (defects) in the material where energy would get lost.

• The Fix: They used a technique called Phosphorus Diffusion Gettering. Imagine throwing a magnet into a pile of iron filings. The magnet pulls all the filings (the bad impurities) out of the pile and holds them tight.
• The Result: This "magnet" step cleaned the inside of the silicon so well that the quality jumped from "okay" to "excellent."

5. Making the Solar Cells: The "Black" Finish

To make the solar panels catch more sunlight, they textured the surface.

• Standard Method: Making tiny pyramids (like a mountain range) on the surface.
• Their Method: Using a chemical trick called MACE to make the surface look like Black Silicon. It's so dark it looks like a black hole for light! This means almost no sunlight bounces off; it all gets absorbed.

6. The Results: Do They Work?

They built solar cells using this new, cheaper, cleaner silicon.

• Efficiency: They matched the performance of the expensive "Gold Standard" cells. Some even reached over 20% efficiency (which is very high for solar panels).
• Durability: They tested them outside for years. They didn't break down faster than the expensive ones. In fact, they lasted just as long.
• Advanced Tech: They even tested these materials on the newest, most advanced solar cell designs (called TOPCon), and they worked great there too.

7. The Green Bonus: Why It Matters for the Planet

The biggest win isn't just the cost; it's the environment.

• Energy: Making the "Gold Standard" silicon is like running a giant factory 24/7. Making UMG-Si uses much less energy.
• Carbon Footprint: Because it uses less energy, it creates far less carbon dioxide.
• The Payback: If you build a solar farm with these new panels, the energy they generate "pays back" the pollution created to build them twice as fast as the expensive panels do.

The Bottom Line

This paper proves that you don't need the most expensive, energy-hungry process to make great solar panels. By cleverly cleaning up cheaper, industrial silicon, the team created a "super-material" that is:

1. Cheaper to make.
2. Greener for the planet.
3. Just as good at making electricity.

It's like discovering a way to turn ordinary clay into gold bricks without needing a magic wand—just some very smart science. This could make solar energy even cheaper and faster to deploy around the world.

Technical Summary

1. Problem Statement

The photovoltaic (PV) industry has historically relied on the Siemens process to produce polysilicon, a method that is energy-intensive, capital-expensive, and involves complex chemical transformations (e.g., trichlorosilane). This creates a bottleneck in silicon supply and limits the cost-reduction potential of solar technology.
Upgraded Metallurgical Grade (UMG) silicon offers a potential alternative by purifying metallurgical silicon through physical metallurgical steps rather than chemical conversion. However, early attempts faced significant challenges:

• Impurity Control: High concentrations of Boron (B) and Phosphorus (P) in raw metallurgical silicon lead to compensation effects, causing non-uniform resistivity and potential type inversion (p-type to n-type) along the ingot.
• Electronic Quality: Metallic impurities and defects result in low carrier lifetimes, limiting solar cell efficiency compared to electronic-grade polysilicon.
• Degradation: Concerns existed regarding Light-Induced Degradation (LID) and Light and Elevated Temperature Induced Degradation (LeTID) in UMG-based cells.
• Commercial Viability: Previous UMG initiatives failed to compete with the rapidly optimizing and cost-reducing polysilicon industry.

2. Methodology

The paper summarizes a two-decade R&D effort by the Ferrosolar joint venture (Ferroglobe and Aurinka PV Group) and academic partners (UPM, ODTÜ, etc.) to optimize the UMG value chain from feedstock to module.

• Feedstock Purification: A multi-step metallurgical process was employed:
  1. Slagging: To reduce Boron content.
  2. Vacuum Evaporation: To reduce Phosphorus content.
  3. Directional Solidification: To crystallize ingots and remove metallic impurities.
• Crystallization Engineering: To address the compensation issue (B vs. P segregation), Gallium (Ga) was added as a p-type dopant during crystallization. This created a uniform resistivity profile (~1 $\Omega\cdot$cm) along the ingot height, preventing type inversion.
• Defect Engineering:
  • Phosphorus Diffusion Gettering (PDG): Optimized thermal processes (specifically at 780°C) were developed to remove bulk metallic impurities.
  • Texturing: Implementation of Metal-Assisted Chemical Etching (MACE) to create "Black Silicon" nanotextures, enhancing light trapping compared to standard pyramidal texturing.
• Cell Fabrication: Three architectures were tested on 156.75 mm wafers:
  1. Al-BSF: Standard industrial process.
  2. PERC: Optimized with MACE texturing, double-stack passivation (a-SiOxNy:H/a-SiNx:H), and ALD Al2O3 rear passivation.
  3. TOPCon (TOPCoRe): Tested at Fraunhofer ISE to evaluate potential for advanced architectures.
• Degradation Testing: Cells underwent LID and LeTID testing (IEC TS 63342), including regeneration steps (light + heat treatment).
• Field Testing: Modules were installed in Ávila and Tudela, Spain, and monitored for 2–3 years against polysilicon reference modules.
• Life Cycle Assessment (LCA): A process-based LCA was conducted using Simapro 9.0 and Ecoinvent 3.5, comparing UMG and polysilicon across different electricity mixes (Spain, EU, US, China) and cell technologies.

3. Key Contributions

• Process Optimization: Demonstrated a complete, industrially scalable UMG-Si manufacturing route that successfully manages B/P compensation via Gallium doping.
• Defect Engineering: Established that sequential PDG steps can boost carrier lifetimes from ~1 $\mu$s (bare wafers) to 722 $\mu$s, making the material electronically comparable to polysilicon.
• Advanced Cell Architectures: Proved that UMG-Si is not limited to basic cells but is compatible with high-efficiency PERC and emerging TOPCon architectures.
• Reliability Validation: Provided long-term outdoor data proving UMG modules do not suffer from unique degradation mechanisms compared to polysilicon.
• Environmental Quantification: Provided a comprehensive LCA showing significant reductions in Carbon Payback Time and Climate Change emissions for UMG technology.

4. Key Results

A. Material Quality & Cell Performance

• Purity: Final UMG-Si contains <0.2 ppmw Boron, 0.1–0.3 ppmw Phosphorus, and <0.5 ppmw total metals.
• Carrier Lifetime: Achieved average effective lifetimes of 300 $\mu$s after one PDG and up to 722 $\mu$s after a second optimized PDG.
• Cell Efficiencies:
  • Al-BSF: Achieved 18.4% (comparable to polysilicon).
  • PERC: Achieved 20.1% (pilot) and up to 20.8% (record) in industrial lines.
  • TOPCon: ELBA analysis predicted efficiencies >22% for UMG substrates.
• Degradation:
  • LID: Reversible via standard regeneration steps; Voc degradation was successfully restored.
  • LeTID: Preventive treatments rendered UMG PERC cells insensitive to LeTID (<3% power loss).

B. Field Performance

• Outdoor monitoring in Spain over 2–3 years showed Performance Ratios (PR) for UMG modules were identical to or slightly better than polysilicon references.
• Degradation rates were measured at ~0.2%/year for polysilicon, while no measurable degradation was found for UMG modules over the same period.

C. Environmental Impact (LCA)

• Climate Change Emissions: UMG technology reduces emissions by >20% compared to polysilicon (e.g., 12.1 vs. 14.95 gCO2eq/kWh in Spain; 16.12 vs. 21.38 gCO2eq/kWh in China).
• Energy Payback Time (EPBT): UMG modules manufactured in Spain have an EPBT roughly 50% lower than polysilicon modules manufactured in China.
• CO2 Payback Time: In high-emission grids (e.g., Germany), UMG systems offset manufacturing emissions in 1.8 years compared to 3.1 years for polysilicon.
• Acidification: UMG shows a 45–76% reduction in acidification potential depending on the manufacturing location and electricity mix.

5. Significance

This paper validates UMG-Si as a viable, low-cost, and environmentally superior alternative to conventional polysilicon.

• Economic: It offers a pathway to reduce CAPEX and manufacturing costs by avoiding the energy-intensive Siemens process.
• Technological: It proves that UMG-Si is no longer a "low-efficiency" niche material but can support high-efficiency PERC and TOPCon architectures, keeping pace with industry evolution.
• Sustainability: The LCA results are critical for the PV industry's decarbonization goals. UMG technology significantly lowers the carbon footprint of solar energy, particularly when manufactured in regions with cleaner energy mixes or when deployed in high-carbon grids where the "payback" is faster.
• Reliability: The demonstration of long-term outdoor stability and resistance to degradation mechanisms removes the primary barrier to commercial adoption.

In conclusion, the sustained R&D effort has successfully demonstrated that UMG-Si can form the basis of a low-CAPEX, low-cost, and environmentally friendly supply chain for the global PV industry.