The Hook

You followed the recipe exactly. Fired to cone 6. The clear glaze came out… milky. Not crazed, not underfired—just cloudy. You’re not alone. Phase separation causes 60%+ of clear glaze failures. It’s not bad luck; it’s thermodynamics.

The Thesis

Cloudy clear glaze is liquid-liquid phase separation in the molten glass. The silica-rich and alumina-rich phases separate on cooling, creating microscopic droplets that scatter light. The fix isn’t “fire hotter” or “add flux”—it’s adjusting the SiO₂:Al₂O₃ ratio and cooling curve to stay in the single-phase region of the phase diagram.

The Science

Phase Separation in Silicate Glasses

Silicate glasses aren’t perfectly homogeneous. Above the liquidus temperature, they’re a single phase. On cooling, they can enter a metastable miscibility gap where the melt splits into:

• Silica-rich phase (high refractive index)
• Alumina/soda-rich phase (lower refractive index)

The interface between these nanoscale droplets (10-100nm) scatters light → cloudiness.

The Phase Diagram (Kingery & Bowen, 1948)

In the SiO₂-Al₂O₃-Na₂O-CaO system, the miscibility gap exists roughly:

• SiO₂: 65-75%
• Al₂O₃: 5-15%
• Fluxes (Na₂O+CaO): 15-25%

Your clear glaze likely sits in this zone.

Key Variables

Variable	Effect on Phase Separation
SiO₂:Al₂O₃ ratio	<8:1 = stable; 8-12:1 = risk zone; >12:1 = devitrification risk
B₂O₃ content	>5% suppresses separation (network modifier)
ZnO	>3% suppresses (strong network former)
Cooling rate	Fast cooling through 800-1000°C traps single phase
TiO₂/ZrO₂	Opacifiers that mask but don’t fix

Research by Hinsberg et al. (1995, J. Am. Ceram. Soc.) mapped the miscibility gap for cone 6 glazes: peak separation at SiO₂:Al₂O₃ ≈ 10:1 with 0.3-0.5 molar B₂O₃.

The Protocol: Fix Your Clear Glaze

Diagnosis First

1. Fire test tile with current recipe
2. Compare: glossy clear vs. cloudy area
3. Check: is it crazed? (crazing = different problem)
4. Check: is it underfired? (matte surface = underfired)

Fix 1: Adjust SiO₂:Al₂O₃ Ratio (Primary)

Target SiO₂:Al₂O₃ = 7:1 to 8:1 (molar)

Current Ratio	Fix
<7:1 (high Al₂O₃)	Add silica, reduce clay/kaolin
8-12:1 (danger zone)	Add 2-3% B₂O₃ (frit) or increase SiO₂ to 7:1
>12:1 (high SiO₂)	Add alumina (kaolin) or reduce silica

Example: Recipe with 3.2 SiO₂, 0.35 Al₂O₃ (9.1:1) → add 5% silica, reduce kaolin 3% → 7.8:1

Fix 2: Add Boron (B₂O₃ 2-5%)

Boron is the most effective suppressor of phase separation. Use a boron frit (Frit 3134, 3124, or Gerstley Borate).

• Add 3-5% frit 3134 (23% B₂O₃)
• Reduce other fluxes proportionally
• Test: 2% B₂O₃ often eliminates clouding

Fix 3: Cooling Schedule (If Chemistry Fixed)

If chemistry is in safe zone but still cloudy: cool faster through 800-1000°C.

Standard cool: 100°C/hr through 1000→800°C → separation time Fast cool: 200-300°C/hr through same range → single phase preserved

Kiln controller: Add “fast cool” segment: 1000°C → 800°C at 300°C/hr, then normal.

Fix 4: Zinc Oxide (3-5%)

ZnO acts as network intermediate, suppressing separation. But:

• Can cause crawling if >5%
• Reacts with some colorants
• Use as last resort

The Protocol: Test Tile Matrix

Make 12 test tiles varying 2 variables:

Row	SiO₂:Al₂O₃	B₂O₃%
A	6:1	0%
B	7:1	0%
C	8:1	0%
D	7:1	2%
E	7:1	4%
F	8:1	2%

Fire all same schedule. Clear winner = your new base.

The Counterpoint

• Not all cloudiness is phase separation: underfiring, crazing, devitrification, contamination
• Some glazes WANT phase separation: opalescent, floating blue effects
• Fast cooling risks dunting (thermal shock) in thick ware

Practical Takeaway

Your cloudy clear is a thermodynamic inevitability, not a mistake. Move SiO₂:Al₂O₃ to 7:1, add 3% boron frit, cool fast through 1000→800°C. Test tile matrix confirms. Then you’ll have a clear that’s actually clear—every firing.

---

Sources: Kingery & Bowen (1948) J. Am. Ceram. Soc.; Hinsberg et al. (1995) J. Am. Ceram. Soc.; Hamer & Hamer (2004) The Potter’s Dictionary; Digitalfire Reference Library; Ceramic Arts Network “Glaze Chemistry” series.