Why Does High Purity Quartz Sand Require Ultra-Low Phosphorus Content?

Among the trace elements controlled in high-purity quartz sand, phosphorus (P) occupies a uniquely dangerous position: it is one of the three primary n-type dopants used intentionally in silicon semiconductor manufacturing — alongside arsenic and antimony — and therefore any phosphorus that leaches from a quartz crucible into molten silicon acts as an uncontrolled dopant, altering the electrical properties of the growing crystal in ways that are difficult to predict or compensate for.

At JACAN, we process high-purity quartz sand for semiconductor crucible and photovoltaic applications, where phosphorus control is a critical quality parameter — often specified at sub-ppm levels and, for the most demanding inner-layer crucible materials, at tens of ppb. Below is a detailed technical explanation of why ultra-low phosphorus content is non-negotiable for high-purity quartz sand.

1. Phosphorus in Quartz: Occurrence and Persistence

Phosphorus enters quartz through geological hydrothermal processes and, like aluminum and boron, tends to incorporate into the quartz crystal lattice rather than remaining as separable mineral inclusions:

Lattice Incorporation

Phosphorus can substitute for silicon in the SiO₄ tetrahedral framework (P⁵⁺ replacing Si⁴⁺), forming strong covalent bonds within the crystal structure. This lattice-bound phosphorus:

• Cannot be removed by physical separation methods (crushing, magnetic separation, flotation).
• Is only partially accessible to standard acid leaching.
• Requires specialized high-temperature chemical treatment (e.g., chlorination roasting, where phosphorus is converted to volatile PCl₃ or PCl₅) for significant reduction.

Difficulty of Removal

Because of its lattice incorporation, phosphorus is one of the most difficult impurities to reduce in quartz beneficiation. The practical implication is that ore selection is the primary phosphorus control strategy — only deposits with inherently low phosphorus content can meet semiconductor and high-end PV specifications, and beneficiation can only make marginal improvements.

2. The Diffusion Path: From Crucible to Silicon Melt

The critical mechanism by which quartz phosphorus affects semiconductor and PV silicon is high-temperature dissolution and diffusion during Czochralski (CZ) crystal growth:

The Process

1. Polycrystalline silicon feedstock is loaded into a fused quartz crucible and heated to ~1,450°C.
2. At this temperature, the inner surface of the quartz crucible slowly dissolves into the molten silicon — an inevitable process that also provides the SiO₂ layer preventing crucible sticking.
3. Phosphorus contained in the quartz glass is released into the silicon melt as the crucible dissolves.
4. Phosphorus has a segregation coefficient (k₀) of ~0.35 in silicon — meaning it readily incorporates into the growing crystal (though less efficiently than boron, k₀ ≈ 0.8).
5. As the pull progresses and melt volume decreases, phosphorus concentration in the melt increases, causing axial resistivity variation along the ingot.

Why This Is Catastrophic

Unlike iron or copper — which create deep-level defects but do not fundamentally change the carrier type — phosphorus is a shallow donor in silicon. Even trace amounts:

• Shift the resistivity of the crystal.
• Can compensate intentional p-type dopants (boron) in p-type silicon.
• Can add to intentional n-type dopants (phosphorus, arsenic) in n-type silicon.
• Cause the grown crystal to deviate from its target doping concentration and conductivity type.

The result is ingots that fail resistivity specifications, wafers with inconsistent electrical properties, and devices with unpredictable performance.

3. Phosphorus as an n-Type Dopant: Quantified Impact

To understand why phosphorus control is so critical, it is essential to recognize how potent a dopant phosphorus is in silicon:

Doping Concentration vs. Resistivity

A typical p-type PV wafer has a target resistivity of 1–3 Ω·cm, corresponding to a boron concentration of approximately 1–5 × 10¹⁵ atoms/cm³. For n-type wafers (TOPCon, HJT), target resistivity is similar, with phosphorus at comparable concentrations.

Phosphorus from the crucible, even at sub-ppm levels in the quartz, can translate to 10¹³–10¹⁴ atoms/cm³ in the silicon — representing 1–10% of the total dopant budget. This is enough to:

• Shift resistivity by several percent.
• Cause measurable axial variation along the ingot.
• Reduce the fraction of the ingot meeting tight resistivity specifications.

Heavily Phosphorus-Doped Silicon

For context, heavily phosphorus-doped silicon (used as power device substrates) reaches concentrations of 10¹⁹–10²⁰ cm⁻³, with resistivity as low as ~1 mΩ·cm. This demonstrates the enormous dynamic range of phosphorus as a dopant — and why even trace crucible-derived phosphorus matters at the low end of the doping scale.

Constitutional Supercooling

At high phosphorus concentrations, constitutional supercooling becomes a growth defect risk — causing dopant striations, dislocations, and even polycrystalline growth. While crucible-derived phosphorus alone rarely reaches these levels, it adds to the intentional dopant burden, pushing the melt closer to the supercooling threshold and reducing process margin.

4. Differential Impact: p-Type vs. n-Type Silicon

Phosphorus contamination affects p-type and n-type silicon differently — and the industry’s shift to n-type is reshaping phosphorus requirements:

p-Type Silicon (PERC, Al-BSF)

• Boron is the intentional dopant; phosphorus is an unwanted compensating donor.
• Crucible phosphorus reduces the net hole concentration (p = N_A − N_D), increasing resistivity beyond target.
• Phosphorus-boron compensation also reduces carrier mobility and lifetime through ionized impurity scattering.
• The axial increase in phosphorus (as the crucible dissolves) causes resistivity to rise along the ingot — opposite to the normal boron segregation trend, creating complex non-linear resistivity profiles.
• Typical p-type crucible phosphorus specification: ≤ 0.1–0.5 ppm.

n-Type Silicon (TOPCon, HJT, XBC)

• Phosphorus is often the intentional dopant (or arsenic is used, with phosphorus as a contaminant).
• Crucible phosphorus adds to the intentional dopant, reducing resistivity below target.
• For n-type wafers requiring high minority carrier lifetime (> 500 μs), any uncontrolled doping variation degrades lifetime and cell efficiency.
• n-type crucible inner-layer phosphorus specification: ≤ 0.04 ppm (40 ppb), with advanced specifications pushing lower.

The Industry Trajectory

As the PV industry transitions from p-type PERC to n-type TOPCon (now > 50% of new installations), and as semiconductor nodes shrink, the demand for ultra-low-phosphorus quartz sand is accelerating. n-type crucible inner layers now approach or exceed semiconductor-grade phosphorus specifications.

5. Impact on Semiconductor Device Performance

Beyond crystal growth, phosphorus contamination affects downstream device fabrication:

Resistivity Uniformity

Semiconductor devices require tightly controlled wafer resistivity — often within ±5% across a wafer and ±10% across an ingot. Uncontrolled phosphorus from the crucible:

• Creates axial resistivity gradients (seed-to-tail variation).
• Can cause radial non-uniformity if crucible dissolution is asymmetric.
• Forces device manufacturers to bin wafers by resistivity, reducing yield and increasing cost.

Dopant Compensation in p-Type Devices

In p-type MOSFETs and bipolar devices, phosphorus compensation:

• Shifts threshold voltage (V_t).
• Reduces channel mobility.
• Increases leakage current in reverse-biased junctions.
• Causes device-to-device variability that is unacceptable for advanced logic and memory chips.

Gettering Interactions

Phosphorus is sometimes used intentionally as a gettering agent in silicon processing — phosphorus-doped glass (PSG) layers attract and trap metal impurities. Uncontrolled phosphorus from the crucible can interfere with intentional gettering schemes, creating unpredictable impurity distributions in the wafer.

Phosphorus-Oxygen Precipitation

In heavily phosphorus-doped CZ silicon, phosphorus can interact with oxygen (also from the crucible) to form P-O precipitates during thermal processing. These precipitates:

• Act as recombination centers, reducing carrier lifetime.
• Cause wafer warpage and slip dislocation.
• Degrade the performance of power devices and high-efficiency solar cells.

6. Phosphorus Effects on Quartz Glass Itself

Phosphorus also degrades the performance of the quartz crucible and process ware:

Devitrification Acceleration

Like aluminum and alkali metals, phosphorus enhances the devitrification (cristobalite formation) tendency of quartz glass at high temperatures. During crystal growth:

• A controlled cristobalite layer on the crucible inner surface is beneficial (prevents sticking, reduces particle generation).
• Excessive, uncontrolled devitrification caused by phosphorus creates rough surfaces, particle shedding, and crucible degradation — leading to dislocations in the growing crystal and reduced crucible lifetime.

Viscosity and Thermal Properties

Phosphorus oxide (P₂O₅) is a glass former that, when present in quartz glass:

• Lowers the softening temperature.
• Modifies thermal expansion coefficient.
• Can cause phase separation at high phosphorus concentrations.

While phosphorus levels in high-purity quartz are too low to dramatically change bulk properties, even trace phosphorus can influence surface crystallization behavior — the critical interface between crucible and silicon melt.

Electrical Conductivity

Phosphorus incorporation can slightly increase the electrical conductivity of quartz glass — problematic for quartz components used in RF and plasma processes (e.g., etch chambers, CVD chambers) where dielectric integrity is required.

7. Industry Standards for Phosphorus in High-Purity Quartz

The phosphorus specifications for high-purity quartz sand vary by application and grade:

Photovoltaic-Grade Quartz Sand

Semiconductor-Grade Quartz

Reference Standards

• IOTA-CG (photovoltaic crucible grade, Sibelco): P = 0.12 ppm (certified typical value); actual measured values in purified sand can reach 0.095 ppm.
• IOTA-4 / IOTA-STD (semiconductor grade): P controlled at sub-ppm levels, with exact values proprietary.
• JC/T 1048-2018 (quartz crucible for single crystal silicon): T-grade specifies 13-element total ≤ 20 ppm, with P implicitly controlled within this total.
• GB/T 32649-2016 (PV high-purity quartz sand): P listed among controlled trace elements.
• 4N8 grade (99.998% SiO₂): Total impurities < 20 ppm, with P typically < 0.05–0.1 ppm.

Summary Table

8. The Purification Challenge: Removing Phosphorus from Quartz

Phosphorus’s lattice-bound nature makes it one of the most difficult impurities to remove:

Ineffective Methods

• Magnetic separation: No effect (phosphorus is non-magnetic).
• Flotation: Removes phosphorus-bearing minerals (apatite) but not lattice-bound phosphorus.
• Standard acid leaching (HCl, H₂SO₄): Minimal effect on lattice phosphorus.
• Scrubbing / attrition: Removes only surface-bound phosphorus.

Specialized Methods

• Chlorination roasting: Heating quartz with Cl₂ or HCl at 800–1,000°C converts phosphorus to volatile PCl₃ (boiling point 76°C) or PCl₅ (sublimes at 167°C), which are removed in the gas phase. This is the most effective method for lattice phosphorus but requires corrosion-resistant equipment and careful emissions control.
• Alkali roasting + leaching: Can liberate some phosphorus but introduces alkali contamination that must then be removed.
• High-temperature vacuum treatment: Partial reduction of phosphorus in some forms.

The Practical Reality

For semiconductor and high-end PV quartz sand, the most cost-effective phosphorus control strategy is raw material selection — choosing deposits with inherently low phosphorus. This is why the Spruce Pine deposit (North Carolina, USA) and a small number of other global deposits command such premiums: their geological formation conditions resulted in inherently low phosphorus incorporation.

9. Phosphorus vs. Boron: The Dopant Pair

Phosphorus and boron are the two most critical dopant-type impurities in quartz sand, and they require parallel but distinct control strategies:

The key insight: for p-type silicon, boron is partially tolerated (it’s the dopant) but phosphorus must be minimized; for n-type silicon, phosphorus is partially tolerated but boron must be minimized. In practice, both are controlled to sub-ppm levels because the cost of crucible contamination far exceeds the cost of purification.

10. The JACAN Perspective

At JACAN, our role in the high-purity quartz supply chain focuses on preserving phosphorus (and all dopant-type impurity) levels through contamination-free processing:

• All-ceramic grinding media: Our alumina and zirconia grinding systems ensure that no phosphorus or other contaminants are introduced during size reduction. Certain grinding aids and ceramic formulations can contain phosphorus — we use phosphorus-free media and process chemicals where required.
• Precision classification: Consistent particle size distribution supports uniform crucible fusion, ensuring that phosphorus (and all impurities) are distributed homogeneously rather than concentrated in size fractions.
• Process integrity: Strict batch segregation prevents cross-contamination between industrial-grade and high-purity quartz sand, ensuring that low-phosphorus material is not exposed to higher-phosphorus material during processing.
• Traceability: Full batch traceability supports customers’ ICP-MS verification, enabling them to confirm phosphorus (and all trace elements) at incoming inspection.

While we do not perform phosphorus removal (the domain of chemical beneficiation specialists), our processing ensures that the ultra-low phosphorus content achieved through ore selection and specialized beneficiation is not compromised during grinding and classification.

High-purity quartz sand requires ultra-low phosphorus content for five converging reasons:

1. Uncontrolled n-type doping: Phosphorus is a potent shallow donor in silicon. When it leaches from a quartz crucible into molten silicon during CZ growth, it acts as an unintended dopant — shifting resistivity, compensating p-type boron, or adding to n-type dopant concentrations. Even sub-ppm phosphorus in the crucible translates to 10¹³–10¹⁴ atoms/cm³ in the silicon, representing 1–10% of the dopant budget.
2. Resistivity non-uniformity: As the crucible dissolves during growth, phosphorus concentration in the melt increases axially, creating resistivity gradients along the ingot that reduce the usable fraction and force wafer binning.
3. Device performance degradation: Phosphorus compensation shifts threshold voltage, reduces mobility, increases leakage, and degrades carrier lifetime — effects that are unacceptable for advanced semiconductors and high-efficiency solar cells.
4. Quartz glass degradation: Phosphorus accelerates devitrification (cristobalite formation) at the crucible inner surface, causing particle generation, reduced crucible life, and crystal defects.
5. Industry transition to n-type: As PV shifts to n-type TOPCon and HJT, and as semiconductor nodes advance to 7 nm and below, phosphorus specifications tighten from ≤ 0.1 ppm to ≤ 0.04 ppm (40 ppb) — and to ppb levels for the most advanced applications — driving demand for synthetic quartz inner layers and the highest-purity natural deposits.

The standard for photovoltaic crucible sand is P ≤ 0.1 ppm (IOTA-CG specifies 0.12 ppm typical), while semiconductor and advanced n-type crucible inner layers require P ≤ 0.04 ppm (40 ppb). Achieving these levels depends primarily on ore selection — because phosphorus, lattice-bound and chemically persistent, is among the hardest impurities to remove by beneficiation.

For the high-purity quartz industry, the principle is clear: phosphorus is not just another impurity — it is an electrical dopant, and its presence in the crucible directly rewrites the doping profile of the silicon crystal grown within it. Controlling phosphorus to ultra-low levels is therefore not a quality refinement — it is a fundamental requirement for producing the silicon that powers modern electronics and solar energy.