USP-NF vs. QB/T 2346 — The Dual-Standard Challenge for Toothpaste Silica Exporters

Created on Today

Every silica exporter who has dealt with Korean, Japanese, European, or American buyers knows the moment. Your domestic COA — perfectly compliant with QB/T 2346-2015 — lands on the customer's desk, and the reply comes back: "Please also provide USP-NF compliance data." Or worse: "Your chloride is 0.15%, our limit is 0.1%. Please revise."

Suddenly you are not just selling precipitated silica. You are navigating two monographs with different test methods, different limits, and different philosophies — and your customer expects both on a single Certificate of Analysis. This article breaks down every meaningful difference between USP-NF Silicon Dioxide and China's QB/T 2346, and explains how to engineer and document a product that satisfies both.

USP-NF vs QB/T 2346 dual standard comparison for precipitated silica exporters

Two Standards, Two Purposes

QB/T 2346-2015 is a Chinese recommended industry standard for toothpaste-grade silica. It is performance-oriented: it defines three product series (abrasive, comprehensive, thickening) and sets specs around particle size, whiteness, oil absorption, water absorption, apparent density, and friction value — the properties that determine how silica behaves in a toothpaste formulation.

USP-NF Silicon Dioxide is a pharmacopeial monograph. It is safety- and purity-oriented: it defines silica as an excipient, sets a 99% purity floor, caps chloride, sulfate, arsenic, and heavy metals, and specifies exact test conditions (145°C for 4 hours, 1000°C for not less than 1 hour). It does not care about RDA or oil absorption because those are formulation variables, not safety parameters.

When a Korean buyer asks for both, they want a product that is formulation-ready and excipient-pure at the same time.

Parameter-by-Parameter Comparison

Parameter	USP-NF Silicon Dioxide	QB/T 2346-2015	Dual-Compliance Target
SiO₂ content	≥99.0% (ignited basis)	≥96–98% (by grade)	≥99.0%
Loss on drying	≤5.0% at 145°C, 4 h	≤10% at 105°C	≤5.0% at 145°C, 4 h
Loss on ignition	≤8.5% at 1000°C, ≥1 h	≤8.5% at 1000°C	≤8.5% at 1000°C, 1 h
pH	4–8 (1:20 slurry)	5.5–8.5 (5% slurry, by grade)	6.0–8.0 (5% slurry)
Chloride	≤0.1%	≤0.3%	<0.1%
Sulfate	≤0.5%	Not separately set	<0.5%
Arsenic	≤3 ppm	≤3 mg/kg	≤3 ppm
Heavy metals (as Pb)	≤30 ppm (Method I)	≤15 mg/kg	≤15 ppm (QB/T tighter)
Lead (Pb)	Not separately set in older USP	≤15 mg/kg	≤5 ppm (Korean overlay)
Mercury (Hg)	Not specified	Not specified	≤1 ppm (Korean overlay)
Total salts	Not specified	≤2.0%	≤2.0% (and Cl/SO₄ caps)
Iron (Fe)	Not specified	≤350–500 mg/kg	≤350 mg/kg
Whiteness	Not specified	≥93–96% (by grade)	≥96%
325-mesh sieve passing	Not specified	≥98%	≥98%

quality control laboratory testing silica purity washing conductivity and heavy metals

The Five Hardest Problems (and How to Solve Them)

Problem 1: The 145°C Loss-on-Drying Test

This is the single most common failure point. QB/T 2346 tests at 105°C; USP-NF tests at 145°C for 4 hours. The 40-degree difference drives off not just free moisture but also loosely bound water and some surface silanol groups. A silica that shows 5% LOD at 105°C can show 6.5–7.5% at 145°C.

Solution: Adjust the drying and milling step to achieve lower outgoing moisture (target 4–5% at 145°C). This may require higher dryer temperature, longer residence time, or a flash-drying stage. Verify with a USP-style 145°C/4 h test in your own lab before every shipment.

Problem 2: Getting from 98% to 99% SiO₂

The gap between 98% and 99% is almost entirely residual sodium sulfate and sodium chloride from the precipitation reaction. QB/T 2346 allows total salts up to 2.0%; USP-NF's chloride and sulfate caps effectively push total salts below ~0.7%.

Solution: Add counter-current washing stages or optimize filter-cake washing. Use deionized water for the final wash. Monitor conductivity of the wash filtrate — when it approaches the feed water conductivity, salts are effectively removed.

Problem 3: Chloride at <0.1%

QB/T 2346 allows 0.3% chloride; USP-NF and Korean buyers want <0.1%. Chloride comes from residual NaCl in the cake. The same washing improvements that raise purity will lower chloride, but you must verify because chloride and sulfate wash out at different rates.

Solution: Test chloride at every wash stage. If chloride plateaus above 0.1%, consider a repulp wash (reslurry the cake in DI water and filter again) rather than simply extending the final rinse.

Problem 4: pH Window 6.0–8.0

USP-NF allows pH 4–8; QB/T 2346 allows 5.5–8.5 depending on grade. Korean buyers tighten to 6.0–8.0. The pH of precipitated silica is determined by residual acid/base balance and surface chemistry. Over-washing with acid-treated water can drop pH below 6; insufficient washing leaves it above 8.

Solution: Control the final wash water pH and monitor the cake pH after drying. If pH drifts low, a mild alkaline adjustment during the final rinse may be needed — but this must not reintroduce sodium salts.

Problem 5: Mercury and Lead Documentation

Neither USP-NF (older revisions) nor QB/T 2346 requires mercury testing, but the Korean Food Additives Code caps Hg at 1 ppm. Lead is 15 mg/kg under QB/T 2346 but 5 ppm under Korean requirements.

Solution: Add Hg and Pb (plus As) to your routine heavy metal panel using ICP-MS or atomic absorption. Source sodium silicate from low-heavy-metal suppliers. If you share production lines with industrial grades, implement changeover cleaning and line-clearance procedures.

How to Issue a Dual-Standard COA

A COA that satisfies Korean, European, or American buyers should include the following elements:

1. Product name: "Food Additives Silicon Dioxide (Precipitated Silica) for Toothpaste" — this language signals food-additive grade

2. Reference standards: List both "USP-NF" and "QB/T 2346-2015" (and optionally the Korean Food Additives Code)

3. Test methods with conditions: "Loss on drying, 145°C, 4 h" — not just "Loss on drying ≤5%"

4. Individual chloride and sulfate results — not only total salts

5. Heavy metal panel: Pb, As, and Hg with actual values and detection limits

6. Microbiological results: Total aerobic count, mold & yeast, and absence of specified pathogens (E. coli, P. aeruginosa, S. aureus)

7. Performance data: RDA, oil absorption, water absorption, whiteness, and apparent density for the specific grade

dual standard Certificate of Analysis COA for toothpaste silica export approval

Final Advice for Exporters

Dual-standard compliance is not a QC problem — it is a production philosophy. If your line is engineered for QB/T 2346 commodity output, you will not consistently hit USP-NF limits by testing harder. You need to build the purity into the process: cleaner raw materials, more washing, tighter drying, and routine heavy metal monitoring.

The suppliers that win Korean and global orders are the ones whose COAs already show the numbers — not the ones who promise to "adjust for your order."

Contact

Leave your information and we will contact you.