The Hydrothermal Degradation of Quartz Below the Critical Point
Subcritical pressurized water-heated liquid water maintained under pressure to prevent boiling, typically at 280–350°C and 10–20 MPa-is increasingly used in semiconductor wafer cleaning (high-temperature deionized water rinses), advanced oxidation processes, and hydrothermal synthesis. Unlike supercritical water where silica solubility drops sharply, subcritical water near the critical temperature (374°C) exhibits a maximum in silica solubility. At 300–350°C, the solubility of amorphous silica in pure water reaches 300–500 ppm, approximately 100 times higher than at room temperature. This high solubility drives continuous dissolution of quartz sheaths used as immersion heaters. Simultaneously, the high pressure (10–20 MPa) imposes external compressive stress on the quartz tube, requiring adequate wall thickness to prevent buckling or implosion. The combination of chemical dissolution and mechanical pressure loading creates a unique design challenge: the wall thickness must provide both a corrosion allowance against steady silica removal and sufficient stiffness to withstand the external hydrostatic pressure. This analysis quantifies silica solubility and dissolution kinetics in subcritical deionized water as functions of temperature and pressure, determines the minimum wall thickness for pressure containment, and derives the required initial wall thickness to achieve target service lives of 500–5,000 hours. A selection framework distinguishes clean high-purity water systems from those with trace ionic or organic contaminants that accelerate attack.
Silica Solubility and Dissolution Kinetics in Subcritical Pressurized Water
The equilibrium solubility of fused silica in pure water increases exponentially with temperature up to approximately 360°C, following the relationship log₁₀(S) = A - B/T, where S is solubility in mg/kg (ppm), T is absolute temperature, and A and B are empirical constants. At 280°C (saturation pressure ~6.5 MPa), solubility is approximately 120 ppm. At 320°C (11 MPa), solubility reaches 250 ppm. At 350°C (16.5 MPa), solubility is 380–420 ppm. At 360°C, it peaks near 500 ppm before declining in the supercritical region.
The rate at which quartz approaches this equilibrium is controlled by the mass transfer boundary layer. In flowing systems (typical in semiconductor wet benches or circulation loops), the dissolution rate is given by R = k_m × (C_sat - C_bulk), where k_m is the mass transfer coefficient (typically 0.001–0.005 mm/hour per ppm of driving force, depending on flow velocity). In clean systems where C_bulk is near zero (fresh deionized water), the initial dissolution rate at 320°C is approximately 0.002–0.004 mm/hour per ppm of solubility, giving 0.5–1.0 mm/hour for 250 ppm solubility. This is extremely high-comparable to molten caustic. A 2.0 mm quartz sheath would completely dissolve in 2–4 hours. However, in recirculating systems, C_bulk increases over time, reducing the driving force. In a closed loop with a fixed volume of water and a large quartz surface area, the system approaches saturation, and the dissolution rate approaches zero. In continuous flow systems where fresh water constantly passes over the heater, the bulk concentration remains low, and the dissolution rate remains high.
Experimental data from a once-through flow loop with deionized water at 330°C, 12 MPa, and flow velocity 0.5 m/s show a steady-state dissolution rate of 0.08–0.12 mm/hour-much lower than the initial rate calculation because a thin silica-saturated boundary layer forms at the quartz surface. Nevertheless, this rate is still aggressive: a 2.0 mm wall would perforate in 17–25 hours. At 300°C, the rate drops to 0.02–0.04 mm/hour, giving 50–100 hours of life. At 280°C, the rate is 0.008–0.015 mm/hour, providing 130–250 hours. For practical industrial service requiring thousands of hours, quartz is only viable below 280°C. Above 300°C, even thick walls (5–10 mm) would dissolve within weeks.
The presence of trace ionic contaminants dramatically alters these rates. Silica solubility is suppressed by dissolved salts (the "salting-out" effect). In water with 1 ppm sodium chloride, silica solubility decreases by approximately 10%. In 10 ppm NaCl, it decreases by 30–40%. Thus, in semiconductor applications where ultrapure water (resistivity >18 MΩ·cm) is used, the solubility is maximal and corrosion is fastest. In industrial water with higher ionic strength, quartz dissolution is slower. Conversely, trace organic compounds (e.g., residual photoresist) can form complexes with silicic acid, increasing solubility and accelerating attack. For heater design, the purity of the water is as important as temperature.
Pressure Containment Requirements for Subcritical Quartz Sheaths
The external pressure in subcritical water systems at 280–350°C is the saturation pressure of water at that temperature, ranging from approximately 6.5 MPa at 280°C to 16.5 MPa at 350°C. In pressurized systems with an inert gas overlay, pressures can be even higher (e.g., 20 MPa). The quartz sheath, with atmospheric pressure inside, must withstand this external pressure without buckling or imploding. As derived previously for supercritical service, the critical buckling pressure for a thin-walled tube is P_critical = (E / (1-ν²)) × (t/R)³ / 4. For a 20 mm outer diameter quartz tube (R = 10 mm) at 350°C (E for quartz decreases from 72 GPa at room temperature to approximately 68 GPa at 350°C), the minimum wall thickness for a safety factor of 2 against a pressure of 16.5 MPa is calculated. Setting P_critical / 2 = 16.5 MPa → P_critical = 33 MPa. Solving (68e9 / 0.97) × (t/10)³ / 4 = 33e6 → (70.1e9) × (t/10)³ / 4 = 33e6 → (t/10)³ = (33e6 × 4) / 70.1e9 = 132e6 / 70.1e9 = 0.001883 → t/10 = 0.1235 → t = 1.235 mm. For 20 MPa, t_min = 1.35 mm. For 280°C (6.5 MPa), t_min = 0.9 mm. Thus, pressure containment alone requires wall thickness between 1.0 and 1.5 mm for typical diameters. This is modest and easily satisfied by most quartz heater designs. The corrosion allowance, not pressure, dictates the required thickness for long service life.
How Wall Thickness Balances Corrosion Allowance and Thermal Performance
For subcritical water service, the required initial wall thickness is t_initial = t_pressure + (corrosion rate × desired life). At 280°C with a corrosion rate of 0.010 mm/hour and a desired life of 2,000 hours, corrosion allowance = 20 mm, which is absurdly large. This shows that continuous operation at 280°C is not feasible with quartz unless the water is pre-saturated with silica. In practice, subcritical water systems that use quartz heaters operate in one of two modes: (1) batch or recirculating systems where the water becomes silica-saturated over time, reducing the corrosion rate to near zero after an initial period, or (2) very short exposure times (e.g., rapid thermal processing where the heater is only energized for minutes). For recirculating systems, the initial dissolution may remove 0.2–0.5 mm of quartz during the first few hundred hours, after which the rate drops to <0.001 mm/hour. A 2.0 mm wall with an initial loss of 0.5 mm leaves 1.5 mm, which then degrades very slowly. Such systems can operate for years with quartz heaters. For once-through systems, quartz is not recommended above 250°C.
The thermal penalty of thicker walls in subcritical water is significant because water at 280–350°C has high thermal conductivity (0.6–0.7 W/(m·K)) and low viscosity, allowing convective coefficients of 2,000–5,000 W/(m²·K) at modest flow velocities. The boundary layer resistance is very low (0.0002–0.0005 m²·K/W). The conductive resistance of quartz dominates. For a 1.5 mm wall, R_cond = 0.00109 m²·K/W, U ≈ 1/(0.00109+0.0003)= 720 W/(m²·K). For a 3.0 mm wall, U ≈ 1/(0.00217+0.0003)= 405 W/(m²·K)-a 44% reduction. This penalty is substantial. For applications requiring high heat flux (e.g., rapid wafer heating), thin walls are essential. For lower heat flux applications, thicker walls may be acceptable.
Scenario-Based Selection Matrix for Quartz Sheath Wall Thickness in Subcritical Pressurized Water Service
| Application Scenario & Operating Parameters | Recommended Wall Thickness | Core Rationale with Quantified Trade-Off |
|---|---|---|
| Recirculating high-purity water loop (280°C, 6.5 MPa, closed system with silica saturation, 1-year maintenance) | 2.0 – 2.5 mm, high-purity quartz | Initial dissolution removes ~0.3 mm, then rate drops. 2.0 mm provides 1.7 mm remaining after saturation. U ≈ 550 W/(m²·K). |
| Once-through wafer rinsing (300°C, 8.5 MPa, ultrapure water, continuous flow) | Not quartz – use PFA or sapphire | Dissolution rate 0.04 mm/hour. 2.5 mm fails in 60 hours. Alternative material required. |
| Batch hydrothermal synthesis (320°C, 11 MPa, 24-hour cycles, fresh water each batch, 50 batches/year) | 2.5 – 3.0 mm, as-drawn | Per-batch loss ~0.2 mm (rate 0.008 mm/hour × 24h). 3.0 mm lasts ~15 batches (3 months). Acceptable as consumable. |
| Pressurized DI water preheater (250°C, 4 MPa, closed loop, low flow) | 1.5 – 2.0 mm, standard grade | Below 260°C, corrosion rate <0.005 mm/hour. 2.0 mm provides >400 hours initial life, then slows. U ≈ 700 W/(m²·K). |
| High-temperature deionized water (350°C, 16.5 MPa, any configuration) | Not quartz – use titanium or Inconel | Silica solubility >400 ppm. Dissolution rate >0.10 mm/hour. Quartz impractical regardless of wall thickness. |
Complementary Design Modifications for Subcritical Water Heaters
Three strategies enable quartz use in subcritical water beyond the temperature limits suggested above. First, water pre-saturation: passing the feed water through a packed bed of quartz chips at the operating temperature saturates the water with silicic acid before it contacts the heater. This eliminates the driving force for dissolution, reducing the corrosion rate to near zero. A pre-saturation column adds complexity but allows quartz heaters to operate at 320°C for thousands of hours. Second, pH adjustment: adding trace ammonia (to pH 9–10) increases silica solubility? Actually, silica solubility is minimum near neutral pH and increases in both acidic and alkaline conditions. Maintaining neutral pH (6–7) minimizes solubility. Third, surface coating: chemical vapor deposition of a thin (1–2 µm) silicon nitride or boron nitride layer on the quartz surface acts as a barrier against water diffusion, reducing dissolution by 90% at 300°C. However, coating integrity under thermal cycling is challenging.
Conclusion: Specifying Quartz Wall Thickness for Subcritical Pressurized Water with Clear Temperature Limits
Quartz immersion heaters can be reliably used in subcritical pressurized deionized water at temperatures up to 260–280°C, provided the system is closed-loop or recirculating such that silica saturation occurs. Under these conditions, a 2.0–2.5 mm wall thickness provides adequate initial corrosion allowance (0.3–0.5 mm loss during the initial saturation period) and maintains structural integrity for thousands of hours. The thermal penalty of thicker walls is 20–40% compared to ultra-thin walls, but acceptable for most applications. For once-through systems or temperatures above 300°C, quartz dissolution rates exceed 0.04 mm/hour, making it impractical regardless of wall thickness; alternative materials such as sapphire (single-crystal Al₂O₃), PFA-lined tubes, or titanium sheaths should be specified. When requesting quotations for high-temperature pressurized water heaters, specify whether the system is recirculating or once-through, the water purity (resistivity or conductivity), maximum temperature and pressure, and desired service life. This enables the manufacturer to recommend the optimal wall thickness-typically 2.0–2.5 mm for recirculating systems below 280°C, with clear guidance that quartz is not suitable for continuous once-through service above 260°C.
