We spent six months chasing a dielectric breakdown problem on a customer’s PET backsheet line. The film passed all incoming QC on thickness and flame retardancy—V-0 at 50 phr, no problem. But under damp heat testing at 85°C and 85% relative humidity, the wet leakage current crept up after 500 hours. By 1000 hours, three out of ten samples failed the IEC 61730 wet leakage current test at 1500 V. The formulation hadn’t changed. The MDH supplier had.
The first thing we did was pull the batch records and run ICP-OES on the incoming MDH. The culprit wasn’t particle size or surface area—those were within spec. It was the ionic impurity profile. The previous supplier’s MDH had chloride at 180 ppm and sodium at 90 ppm. Our material runs chloride ≤100 ppm and sodium ≤50 ppm. That 80 ppm difference in chloride, combined with moisture and voltage, creates a conductive path through the film. It’s not a theory. It’s what we saw under cross-section SEM: dendritic silver migration from the electrode interface, tracking right along the MDH particle boundaries.
Why 100 ppm Chloride Is a Hard Ceiling, Not a Guideline
The wet leakage current test in IEC 61730 isn’t a quick pass-fail. It’s a 1000-hour soak at 85°C/85% RH with 1500 V DC bias applied continuously. Under those conditions, any mobile ion—chloride, sodium, calcium—dissolves into the adsorbed water layer at the polymer-filler interface. Chloride is the worst because it’s monovalent and highly mobile. Once it migrates to the anode, it accelerates electrochemical corrosion of the metal grid.
We ran a controlled experiment using three MDH grades with identical particle size (D50 1.2 µm) and BET (8 m²/g), but with chloride content varying from 45 ppm to 210 ppm. At 1000 hours, the 45 ppm batch showed leakage current of 0.8 µA. The 210 ppm batch hit 12 µA and arced through. The 100 ppm batch held at 1.5 µA—still within the 2 µA limit, but with no margin. In production, you don’t want to run at the limit. You want the buffer. That’s why we spec chloride ≤100 ppm. In practice, our production QC data over the last three years shows an average chloride of 62 ppm across 240 batches.
Particle Morphology and Dispersion: The Platy Hexagonal Advantage
Thin-film backsheets run at 100–150 µm total thickness. The MDH layer is typically 30–50 µm. If the filler particles aren’t dispersing isotropically, you get pinholes. A single agglomerate above 5 µm creates a stress concentration point that turns into a micro-crack during lamination. That crack then becomes a moisture ingress path.
Our MDH has a controlled platy hexagonal morphology. The aspect ratio runs about 4:1 to 6:1. In a properly compounded masterbatch—twin-screw extrusion at 200 rpm, melt temperature 220°C—these platelets orient parallel to the film surface. That orientation increases the tortuosity path for moisture diffusion. We’ve measured water vapor transmission rates (WVTR) on 50 µm films at 38°C/90% RH: 8.2 g/m²/day with platy MDH versus 11.4 g/m²/day with irregular-particle MDH at the same 50 phr loading. That’s a 28% reduction in moisture ingress, directly from particle shape.
The catch is that you can’t achieve that dispersion with a low-shear process. If you’re using a Banbury mixer with a single-pass drop, you’ll see agglomerates. We recommend a two-pass masterbatch process: first pass at 180°C to wet out the particles, second pass at 220°C with a high-shear screw design to break down any remaining clusters. Our internal QC uses laser diffraction on the masterbatch pellets—anything above 5 µm at D99 triggers a reprocess.
Decomposition Temperature and Extrusion Stability
PET backsheet extrusion runs at 270–280°C. Most commodity MDH grades start decomposing at 300–320°C. That’s too close for comfort. If your MDH starts releasing water vapor during extrusion, you get bubbles, surface haze, and a drop in mechanical properties.
Our MDH has a decomposition onset temperature ≥340°C, measured by TGA at 10°C/min in nitrogen. That gives you a 60°C safety margin above the extrusion temperature. We’ve run 100 kg trial batches on a 75 mm single-screw extruder with a flat die at 280°C melt temperature. No foaming, no bubble formation, and the film surface gloss held at 92% (Gardner 60°). Compare that to a standard MDH with 315°C onset—we saw visible haze at 270°C melt temperature, and the gloss dropped to 78%.
The downside of higher decomposition temperature is that it typically requires a more controlled calcination process, which costs more. But for PV backsheets, the cost of a field failure—module replacement, labor, warranty claims—dwarfs the raw material cost difference. One customer calculated that switching from a commodity MDH to our solar film grade added $0.12 per square meter. Their field failure rate dropped from 0.8% to 0.05% over two years. That math works.
Color Stability and Iron Content
Multi-layer backsheets often use a white MDH-filled layer as a reflector. If the MDH has high iron content, the film yellows after UV exposure. ASTM D1925 yellowness index shift of more than 5 units is visible to the eye in a white film.
Our MDH specs iron ≤30 ppm. In production, we average 18 ppm. We’ve seen competitive materials run 80–120 ppm iron. After 2000 hours of UV exposure per IEC 61730 (UV preconditioning), our film shows a yellowness index shift of 2.1 units. The 80 ppm iron film shifted 8.4 units. That’s not just cosmetic—yellowing indicates polymer degradation, which means the mechanical properties are dropping too. Tensile elongation at break on the high-iron film dropped from 85% to 42% after UV aging. Ours went from 85% to 71%.
The iron comes from the raw ore and the grinding media. We mine our own magnesite, which gives us control over the feed stock. The calcination kilns use alumina-lined chambers, not steel, to avoid iron pickup. It’s not something you can fix with a post-treatment—once the iron is in the crystal lattice, it’s there.
Technical Specifications & Performance Data
| Parameter | Specification | Test Method | Typical Production Value |
|---|---|---|---|
| Purity (Mg(OH)₂) | ≥99.5% | Complexometric titration | 99.7% |
| Particle Size D50 | 0.8–1.5 µm | Laser diffraction (Malvern Mastersizer 3000) | 1.2 µm |
| Specific Surface Area (BET) | 6–12 m²/g | ISO 9277 | 8.5 m²/g |
| Loss on Ignition (LOI) | 30–31% | ISO 12743 | 30.4% |
| Chloride Content | ≤100 ppm | Ion chromatography | 62 ppm (avg, 240 batches) |
| Sodium Content | ≤50 ppm | ICP-OES | 28 ppm |
| Iron Content | ≤30 ppm | ICP-OES | 18 ppm |
| Calcium Content | ≤100 ppm | ICP-OES | 55 ppm |
| Decomposition Onset Temp | ≥340°C | TGA (10°C/min, N₂) | 345°C |
| Whiteness (Hunter L*) | ≥95 | ISO 2470 | 96.2 |
Frequently Asked Questions
Q: How does calcium oxide content affect dielectric strength in EVA encapsulant, not just MDH in backsheets?
A: Calcium oxide in any form—hydroxide, carbonate, or oxide—is a mobile ion source under damp heat. In EVA encapsulants, calcium levels above 200 ppm have been linked to PID (Potential Induced Degradation) in our customer qualification data. The mechanism is the same: calcium ions migrate under bias, forming a conductive layer at the cell surface. We spec calcium ≤100 ppm in our MDH specifically to avoid this issue, even though the MDH is in the backsheet layer, not the encapsulant. Moisture migration through the backsheet carries those ions into the encapsulant over time.
Q: Can I use standard flame retardant grade MDH in a PV backsheet if I add a coupling agent?
A: You can, but you’re masking the root problem. A silane coupling agent will improve dispersion and reduce agglomerates, but it won’t remove the ionic impurities. We’ve tested this: a 1% vinyl silane treatment on a 180 ppm chloride MDH reduced the wet leakage current at 500 hours from 8 µA to 3 µA. At 1000 hours, it was back up to 7 µA. The coupling agent slows moisture ingress, but it doesn’t stop chloride migration once the water is there. You’re better off starting with low-impurity MDH.
Q: What’s the maximum MDH loading you can achieve in a PET backsheet without losing mechanical properties?
A: We’ve run trials up to 65 phr in PET with no loss in tensile strength (maintained at 45 MPa), but elongation at break drops from 120% at 50 phr to 80% at 65 phr. The practical limit for most backsheet applications is 55 phr if you need to pass the elongation requirements in IEC 61730 (≥50% retention after aging). At 50 phr, you get V-0 at 1.6 mm, which is the standard thickness for backsheets. Going higher doesn’t improve flame retardancy enough to justify the embrittlement.
Q: How do you verify MDH dispersion quality in the masterbatch before film extrusion?
A: We use a 50 µm thin-film press-out test. Take 5 g of masterbatch pellets, press at 200°C and 10 MPa for 2 minutes between PET release films. Examine the film under transmitted light at 100x magnification. Count agglomerates above 5 µm per square centimeter. Our pass criterion is fewer than 5 agglomerates per cm². If you see more than 10, the dispersion process needs adjustment—either increase screw speed, raise melt temperature by 10°C, or add a second pass. We also run laser diffraction on the masterbatch pellets dissolved in hot xylene, but the press-out test is faster and catches agglomerates that laser diffraction might miss.
