标签： 行业展会

Bottom line: Perovskite photovoltaics cannot reuse the fired silver pastes developed for crystalline silicon. Perovskite absorbers and organic transport layers degrade irreversibly above roughly 150°C, so metallisation must use low-temperature curing pastes that conduct through a percolating silver network rather than sintered metallurgical contact. That single process constraint drives every selection decision: you accept an order of magnitude higher resistivity in exchange for zero damage to the functional stack and resistance to halide-driven degradation.

1. The Thermal Constraint That Defines the Category

Conventional c-Si front silver paste is fired at 750–800°C, using glass frit to etch through the silicon nitride layer and form an ohmic contact. None of that is available in a perovskite device:

• Absorber instability — MAPbI₃-based compositions show phase decomposition and halide migration under sustained thermal stress well below 150°C. FA/Cs-based formulations are more robust but still constrain post-processing.
• Organic transport layers — Spiro-OMeTAD has a glass transition near 120°C; above it morphology collapses and hole transport degrades. C60/BCP electron transport layers are similarly temperature-limited.
• Flexible substrates — PET/PEN webs typically cap processing at or below 120°C.

Practical cure windows for commercial low-temperature pastes are therefore 100–150°C for 10–30 minutes, with some systems curing at 80°C over longer dwell or with UV assistance.

2. Conduction Mechanism and Its Performance Cost

Low-temperature pastes conduct because resin shrinkage during cure forces silver flakes and particles into contact, forming a percolation network. No sintering occurs. Buyers must plan around the consequences:

• Volume resistivity: typically 1×10⁻⁵ to 1×10⁻⁴ Ω·cm, versus 2–3×10⁻⁶ Ω·cm for fired paste and 1.6×10⁻⁶ Ω·cm for bulk silver — roughly 10–100× worse.
• Compensation levers: larger finger cross-section (higher wet print thickness or double printing), shorter current paths, TCO sheet resistance matching, and flake-dominant silver morphology to maximise inter-particle contact area.
• Contact resistance often dominates — the interface with ITO/IZO or a buffer layer frequently matters more than bulk resistivity. Always measure on your actual stack; datasheet resistivity alone is not predictive.

3. Specification Parameters to Lock Down

4. Two Failure Modes Buyers Consistently Underestimate

4.1 Solvent attack on transport layers

Low-temperature pastes carry substantial organic solvent load (glycol ethers, esters, terpineol and similar). These can dissolve or permeate Spiro-OMeTAD, PTAA and other organic layers, producing immediate shunting or shorts after printing. This is the leading cause of failure when introducing a new paste, and it is invisible on any certificate of analysis. Run a compatibility trial on the complete stack — print, cure, then measure dark J–V and shunt resistance — before any commercial discussion.

4.2 Silver–halide reaction and ion migration

Iodide from the perovskite migrates to the silver electrode and forms insulating AgI, driving progressive fill-factor loss; silver can also diffuse into the absorber creating deep-level defects. Mitigations:

• Barrier layers — ALD SnO₂/Al₂O₃, sputtered IZO, or thicker C60/BCP to block the direct diffusion path.
• Carbon or copper-based pastes — carbon electrodes are far more halide-tolerant and cheaper, at the cost of much higher resistivity (10⁻₃ Ω·cm range). Suitable where stability and cost outweigh peak efficiency.
• Silver-coated copper or silver alloys — reduces silver loading and migration exposure simultaneously.

5. Additional Requirements in Perovskite/Silicon Tandems

Tandem architectures are the main industrialisation route and add constraints:

• Double-bounded thermal budget — the bottom cell is already complete, and the perovskite top cell caps all subsequent processing at ≤150°C.
• Optical transparency (for four-terminal or semi-transparent designs) pushes toward narrower fingers with higher aspect ratio to limit shading loss.
• Lower current, higher voltage — a two-terminal tandem operates at roughly half the current of a single-junction silicon cell, so resistive loss in the grid is intrinsically less punishing. This is genuine design headroom that can offset low-temperature paste resistivity. Evaluate total power loss at module level, not resistivity in isolation.
• Interconnection — cured low-temperature paste is generally poorly solderable, so electrically conductive adhesive (ECA) interconnection is usually required and must be qualified together with the paste.

6. Reliability Qualification

Metallisation is a primary contributor to long-term degradation. Test at module level against the IEC 61215 / 61730 framework:

• Damp heat (85°C / 85% RH, 1000 h minimum) — watch series resistance rise and finger corrosion.
• Thermal cycling (−40°C to +85°C, 200 cycles) — CTE mismatch between polymer matrix and TCO drives microcracking and interfacial delamination.
• Light and bias coupled ageing — ion migration only manifests fully under illumination and electric field. Dark storage testing will substantially overestimate lifetime.
• Adhesion retention — peel strength before versus after ageing.

7. Commercial and Supplier Qualification Points

1. Match terms to development stage. Lab, pilot and mass production have very different needs. At pilot scale (100 g to 1 kg lots) a supplier’s formulation responsiveness matters more than unit price; at volume, lot-to-lot consistency and capacity assurance dominate.
2. Lot consistency clauses. Require single-lot supply where possible, and commit the supplier to a maximum lot-to-lot range on resistivity and viscosity — not merely min/max limits. Retain reference samples.
3. Silver-indexed pricing. Silver dominates cost. Agree a formula referenced to a public benchmark (LBMA or SHFE/SGE silver) with a defined reset interval. Fixed pricing invites non-performance when silver rallies.
4. Mandatory CoA content: silver content, cured volume resistivity with stated cure conditions, viscosity, fineness, solids, manufacture date and lot number. Resistivity without cure conditions is meaningless data.
5. Storage and shelf life. Most low-temperature silver pastes require 2–10°C refrigerated storage with 3–6 month shelf life, cold-chain shipping with freeze protection, and equilibration to room temperature before opening and remixing. This is the single most common failure point in cross-border procurement.
6. IP and formulation lock. For custom formulations, define IP ownership, exclusivity period, and an explicit obligation to notify formulation changes. Undisclosed “silent changes” are a recurring cause of production excursions.
7. Dual sourcing. Most perovskite paste suppliers are early-stage companies with real capacity and solvency risk. Begin parallel qualification of a second source during late pilot, not after a disruption.

8. Market Assessment

The perovskite paste segment remains early-stage: supply is fragmented, technical routes have not converged, and reliability data is not yet standardised. That gives buyers unusual leverage on customisation and pricing, but shifts the qualification burden onto you. The practical recommendation is to front-load validation cost — use stack compatibility trials and light-bias ageing to eliminate most candidates before commercial negotiation, and do not manage this category with the c-Si paste playbook of price and resistivity comparison.

This guide is for technical procurement reference. Confirm all parameters against your own device architecture, process window and supplier-verified data.