Neutralization of the impact of belt speed on printed

This study demonstrates that while belt speed during copper fire-through metallization initially affects the electrical performance of PERC solar cells, subsequent LECO treatment effectively neutralizes this impact, resulting in identical high-efficiency outcomes (20.8%) across different processing speeds.

The Gist

Imagine you are trying to build a super-efficient solar panel. To make electricity flow, you need to connect the silicon "solar engine" to metal wires that carry the power out. For years, we used Silver for these wires because it works great, but it's expensive and scarce. Now, scientists want to use Copper instead because it's cheap and conducts electricity even better.

However, there's a catch: Copper is a bit of a "picky eater." It needs to be fired (heated up quickly) in a very specific way to bond properly with the silicon. If the heating isn't perfect, the connection is weak, and the solar panel loses power.

Here is the story of how the researchers in this paper solved that problem.

The Problem: The "Conveyor Belt" Dilemma

Imagine the solar cells moving down a giant conveyor belt (called a belt furnace) through a super-hot oven to bake the copper onto them.

• The Variable: The speed of the belt determines how long the cells stay in the heat.
  • Slow belt: The cells get a long, gentle bake.
  • Fast belt: The cells get a quick, intense blast of heat.
• The Issue: The researchers tried three different speeds (Slow, Medium, Fast). Before they did anything else, the results were all over the place:
  • The Slow speed didn't bake the copper enough. The connection was weak, like trying to stick two pieces of tape together without pressing them down hard.
  • The Fast speed baked the copper too aggressively. It created a messy, uneven connection, like burning a steak so fast the outside is charred but the inside is raw.
  • Only the Medium speed worked okay, but it was still far from perfect.

In technical terms, the "Series Resistance" (the friction electricity feels trying to get out) was high, and the efficiency varied wildly depending on how fast the belt was moving.

The Detective Work: Looking Under the Hood

The researchers used a super-microscope (SEM/EDS) to look at the tiny interface where the copper meets the silicon.

• They found that on the Fast belt, there was actually more copper sitting right at the surface.
• The Twist: Just having more copper didn't help! It was like having a crowd of people at a door, but the door is jammed. The copper was there, but it wasn't connecting well. It created a "traffic jam" for the electricity, causing it to crowd into a few weak spots and leak out.

The Magic Fix: LECO (The Laser "Tweezer")

This is where the hero of the story enters: LECO (Laser-Enhanced Contact Optimization).

Think of the solar cell after baking as a slightly uneven road. Some parts are bumpy, some are flat, and electricity is struggling to drive over them.

• LECO is like a precision laser "tweezer" or a "road smoother."
• After the cells come out of the oven, they get a quick, precise zap from a laser.
• This laser zap doesn't just heat things up; it locally "welds" the copper to the silicon in the exact right spots, fixing the bad connections and smoothing out the traffic jams.

The Result: Everyone Wins

After the LECO laser treatment, something amazing happened:

1. The Speed Didn't Matter Anymore: Whether the conveyor belt was moving Slow, Medium, or Fast, the final solar cells all performed exactly the same.
2. The Traffic Jam Disappeared: The electricity could flow freely. The "friction" (resistance) dropped dramatically.
3. High Efficiency: All three groups of cells reached a top-tier efficiency of 20.8%.

The Big Takeaway

Before this study, manufacturers were terrified of using copper because if their conveyor belt speed varied even a little, their solar panels would fail. They had to be perfect.

This paper shows that LECO acts as a safety net. It fixes the mistakes made during the baking process. It allows factories to use copper (saving money) without needing to be perfect with their oven settings. It turns a finicky, high-stress process into a robust, reliable one, making cheap, high-efficiency solar power possible for everyone.

In short: They tried to bake copper on solar cells at different speeds, and it was a mess. Then they used a laser to "fix" the connections, and suddenly, the speed didn't matter anymore. Everyone got a perfect solar panel.

Technical Summary

1. Problem Statement

The transition from silver (Ag) to copper (Cu) metallization in industrial silicon solar cells is driven by cost reduction and supply chain stability. However, Cu fire-through metallization is significantly more sensitive to firing conditions than Ag.

• The Challenge: In screen-printed fire-through systems, the formation of the metal-semiconductor contact relies heavily on the thermal profile (specifically the high-temperature dwell time).
• The Specific Issue: Variations in belt speed in a belt furnace alter the dwell time, leading to inconsistent contact microstructures. This creates a narrow "process window" where low series resistance ($R_s$) and low leakage can be achieved simultaneously.
• The Consequence: Without precise control, belt speed variations cause significant performance spread (efficiency, fill factor, and series resistance), making industrial-scale Cu metallization difficult to implement reliably on high-efficiency PERC (Passivated Emitter and Rear Cell) homogeneous emitters.

2. Methodology

The researchers fabricated G1-format industrial crystalline silicon solar cells (p-type, Boron-doped) with a phosphorus-diffused n+ front emitter and a passivated front surface.

• Metallization: Screen-printed Cu busbars and gridlines were used.
• Firing Process: Cells were fired in an industrial inline belt furnace (TPSolar) with a fixed peak temperature (560–600°C).
• Variable: Three distinct belt speeds were tested to modulate the thermal dwell time:
  • BS325: 325 inches per minute (IPM)
  • BS360: 360 IPM
  • BS390: 390 IPM
• Post-Processing: All samples underwent Laser-Enhanced Contact Optimization (LECO) treatment after firing.
• Characterization:
  • Electrical: Illuminated IV measurements (Sinton FCT-450) for $V_{oc}$, $J_{sc}$, Fill Factor (FF), Efficiency, Series Resistance ($R_s$), and Shunt Resistance ($R_{sh}$).
  • Low-Injection Diagnostics: Ideality factor ($n$) at 0.1 sun and recombination current ($J_{02}$) to distinguish between transport limits and parasitic losses.
  • Microstructural Analysis: Cross-sectional Field Emission Scanning Electron Microscopy (FE-SEM) with Energy Dispersive X-ray Spectroscopy (EDS) to analyze Cu/Si interface morphology and elemental composition.

3. Key Results

A. Pre-LECO Performance (Belt Speed Sensitivity)

Before laser treatment, the cells exhibited strong dependence on belt speed:

• BS360 (Middle Speed): Performed best with the lowest $R_s$ (4.56 $\Omega\cdot\text{cm}^2$) and highest FF. This indicated the optimal balance of thermal budget for contact formation.
• BS325 (Slowest): Showed the highest $R_s$ (6.25 $\Omega\cdot\text{cm}^2$) and lower efficiency, suggesting insufficient contact formation due to excessive dwell time or other thermal dynamics.
• BS390 (Fastest): Showed degraded performance despite having a high prevalence of elemental Cu at the interface (per SEM/EDS).
  • Critical Finding: BS390 exhibited a high ideality factor at low injection ($n@0.1\text{sun} = 1.62$) and reduced $R_{sh}$. This indicated parasitic current pathways and current crowding rather than simple resistive loss. The interface was chemically Cu-rich but electrically heterogeneous, leading to localized shunting.

B. Post-LECO Performance (Neutralization)

After LECO treatment, the performance differences between the three belt speeds were effectively eliminated:

• Convergence: The series resistance dropped drastically for all samples to a narrow range of 0.428 – 0.503 $\Omega\cdot\text{cm}^2$.
• Efficiency: All three conditions converged to an efficiency of 20.8%.
• Fill Factor: FF increased to ~80% for all samples, and the gap between FF and pseudo-Fill Factor (pFF) narrowed to ~2–3 percentage points.
• Low-Injection Recovery: The ideality factor at 0.1 sun normalized (approaching 1.16–1.20), and $J_{02}$ decreased, indicating the suppression of parasitic recombination and leakage currents.

4. Key Contributions

1. Demonstration of LECO Efficacy: The study proves that LECO can "neutralize" the impact of firing variations (specifically belt speed) on Cu metallization, effectively broadening the viable process window.
2. Decoupling Chemical vs. Electrical Interface: The research highlights that a high concentration of elemental Cu at the interface (observed in BS390) does not guarantee good electrical contact. It revealed that electrically heterogeneous interfaces (caused by rapid firing) lead to current crowding and low-injection degradation, which LECO successfully repairs.
3. Process Window Expansion: By showing that efficiency can be stabilized at 20.8% across a wide range of belt speeds (325–390 IPM), the work provides a robust pathway for industrial adoption of Cu metallization without requiring ultra-precise, narrow thermal control.

5. Significance

• Industrial Scalability: Copper metallization is critical for reducing the Levelized Cost of Energy (LCOE) in solar PV. This paper addresses a major barrier to its adoption: the sensitivity to firing conditions.
• Reliability: By mitigating current crowding and parasitic leakage through LECO, the technology enhances the long-term stability and reliability of Cu-contacted cells.
• Cost vs. Performance: The findings suggest that manufacturers can use standard industrial belt furnaces with less stringent speed controls, relying on LECO as a "fix-all" step to ensure high-performance, low-cost Cu contacts on high-efficiency PERC cells.

Conclusion: The paper establishes that while belt speed significantly dictates the initial quality of Cu fire-through contacts, LECO acts as a corrective mechanism that homogenizes the metal-emitter interface. This allows for the realization of the cell's intrinsic high diode quality (high pFF) across a broad range of firing conditions, achieving a stable 20.8% efficiency regardless of the initial belt speed variation.