Quartz-rich sand can provide feedstock for sodium silicates. However, its SiO₂ content does not establish suitability for a particular plant. The processing route, consistent dosing, particle-size fractions and finished-product criteria need to be examined together.
In Belgium, Silmaco describes silicate manufacturing in Lanaken and identifies quartz as a starting material. That industrial presence provides a concrete context for this subject, without establishing a commercial relationship with Minerallum or proving demand for a particular origin. 1
This guide explains how to qualify sand for that transformation. It separates information from the sand supplier, decisions by the silicate manufacturer and properties measured on the resulting solution.
1. Define the product before the feedstock
“Sodium silicate” covers different compositions. CEES distinguishes the molar SiO₂/Na₂O ratio from the weight ratio used in many industrial specifications. A number on a document is meaningful only when its definition is explicit. 2
Starting with the finished product avoids turning a mineral analysis into a performance promise. The manufacturer needs to specify whether the feedstock is intended for fusion, dissolution in an alkaline solution or an intermediate preparation. It should also identify the checks determining acceptance of its silicate.
A usable enquiry therefore combines the proposed route, product form, analytical acceptance criteria and trial boundaries. Simply requesting “sand for silicate” leaves those four questions open and makes competing offers difficult to compare.
2. Distinguish the two processing routes
CEES describes a route involving fusion of sand with an alkali carbonate, and a hydrothermal route using a hydroxide solution. In the first, the resulting silicate glass can subsequently be dissolved; in the second, the reaction produces a solution. 3
These are process categories, not interchangeable prescriptions. For qualification, the plant needs to define the relevant route and retain its reference parameters. Testing sand through another route does not demonstrate its behaviour in the target installation.
Also distinguish received feedstock from the material actually introduced into the process. Delivered sand may undergo drying, grinding or blending. If preparation is part of qualification, its effects and costs belong in the comparison. Feedstock performing well after additional preparation is not equivalent to directly usable feedstock.
3. Interpret the SiO₂/Na₂O ratio correctly
The weight ratio divides the mass of SiO₂ by that of Na₂O on the same calculation basis. The molar ratio compares their amounts of substance. Writing only “ratio 3” therefore does not allow the composition to be reconstructed.
Consider a hypothetical solution containing 30 kg of SiO₂ and 10 kg of Na₂O per 100 kg of solution. Its weight ratio is 30/10 = 3. Calculating the molar ratio requires dividing each mass by its oxide’s molar mass before taking the quotient. The two numerical values are different.
In this example, the two oxides sum to 40% of the solution. That sum equals total dry solids only under the simplifying assumption that no other contribution is present. An industrial data sheet should define its dry-solids measurement rather than replace a measurement with this addition.
4. Keep ratio, concentration and density separate
Adding 100 kg of water to the previous hypothetical 100 kg solution gives 200 kg of mixture. The weight fractions become 15% SiO₂ and 5% Na₂O, but their ratio remains 3. This changes concentration, not the ratio between the oxides.
Density does not define the ratio either. PQ identifies the ratio as influencing several silicate properties, including viscosity and density. This does not make one density measurement sufficient to identify an unknown composition. 6
When comparing solutions, request their analytical contents, defined ratio and measurement conditions separately. Density measured at a stated temperature can help control a product; it does not automatically replace the analysis required to identify it.
5. Separate total purity from processing behaviour
Total SiO₂ content describes sand composition. It does not directly measure the fraction transformed during a particular trial. A high chemical result and satisfactory conversion answer different questions.
The protocol should define “conversion”: disappearance of an initial phase, transfer of silica into solution, material recovered after separation or another suitable indicator. Using the same word for different balances prevents a difference from being interpreted.
Document analytical sample preparation too. A result expressed on a dry basis is not directly comparable to a wet sand mass. Our guide to silica sand moisture examines that distinction.
Sound qualification therefore combines composition, mass basis and transformation result instead of treating purity as a substitute for a process trial.
6. Examine grading without seeking a universal size
Hoang and coauthors studied the effect of sand particle size on alkaline dissolution in research subsequently concerned with nanosilica synthesis. Their optimisation shows that grinding and dissolution need to be assessed together under the study’s conditions. It does not establish a general specification for every silicate plant. 5
For an industrial project, the useful question is which distribution produces the intended result with the preparation actually available. Comparing just two average particle sizes can conceal a difference in coarse fraction or fines content.
Define several control points on the curve, document the measurement method and identify the fractions prepared for the trial. The guide to silica sand particle-size distribution helps organise that assessment.
7. Compare received and prepared sand
A curve supplied for the commercial product describes a particular stage. After drying, grinding or separation, the material may have another distribution. The trial report must show which distribution was actually tested.
A simple arrangement retains three identifiers: received lot, prepared subsample and introduced material. Where several fractions originate from the same lot, their relationship should remain visible. Results from two preparations should not be presented as two independent lots.
Also measure mass retained after preparation. Improved conversion may accompany material loss or an additional operation. The decision should cover the whole chain, without assigning the initial sand a performance obtained only after modification.
8. Build the balance on a consistent dry basis
Consider a teaching example unrelated to any commercial grade. A 1,000 kg lot contains moisture equal to 4% of its wet mass. Its dry mass is 1,000 × 0.96 = 960 kg. If dry-basis analysis gives 98% SiO₂, the theoretical total SiO₂ contribution is 960 × 0.98 = 940.8 kg.
This calculation establishes neither the amount actually transformed nor finished-product purity. It simply prevents the entire 1,000 kg delivered mass from being assigned to available silica. A process balance still needs the trial result and measured outputs.
When lots are blended, calculate their dry contributions separately before adding them. An arithmetic average of contents may be incorrect when masses and moisture differ. Percentages need to retain their calculation basis through to the final result.
9. Connect impurities to product criteria
An elemental list becomes useful when each analysis answers a question. The manufacturer may define composition, appearance or undissolved-material criteria; the supplier then needs to explain which information is available for the proposed lot.
Avoid turning a limit borrowed from glass, foundry or another application into an automatic silicate limit. The process and finished product are different. A requirement should connect to the recipient’s specification and an agreed analytical method.
If bulk analysis and residue observations diverge, retain both results. One describes the sample before transformation; the other concerns a particular output. Qualification becomes more precise by investigating the difference rather than removing whichever result appears less favourable.
10. Distinguish conversion, residue and solution quality
EPA’s sodium silicate profile describes, among other steps, dissolution of silicate glass and subsequent solution filtration in the furnace route. These stages explain why sand characteristics and filtered-solution characteristics are not equivalent information. 4
A comparative trial can track the transformation indicator, residue mass and selected solution characteristics separately. Improving one indicator is insufficient if another becomes incompatible with the industrial objective.
Identify residues by stage. Untransformed solid, material separated during filtration and a deposit appearing later do not automatically represent the same phenomenon. Combining them as “losses” removes part of the diagnostic evidence.
11. Prepare a comparison isolating the sand change
The first trial should compare reference and candidate feedstocks with documented preparation. If grinding, proportions and processing conditions all change, the result concerns a new combination rather than the sand change alone.
Define beforehand which parameters remain constant and which can be adjusted in a second phase. This separates initial compatibility from process optimisation. A material rejected under the first arrangement may warrant another study, but that study needs to be identified separately.
Include genuinely independent repetitions. Three analyses of one solution assess part of analytical repeatability; they do not represent three separate transformations. The report should define its experimental unit before presenting a mean.
12. Compare costs within identical boundaries
Price per delivered tonne is only one part of a comparison. A suitable assessment separates dry feedstock received, preparation, separated losses, transformation and accepted product. The chosen boundaries must be identical for every material.
In a hypothetical example, if 100 kg of introduced dry material leads to 90 kg of useful contribution under the agreed indicator, the quotient is 0.90. It does not become an overall manufacturing yield when product mass also includes sodium and water. Defining the numerator matters as much as the result.
This distinction avoids two errors: calculating performance from the total mass of a diluted solution, or comparing raw and prepared feedstocks without counting preparation. The balance must remain understandable even when an offered price looks attractive.
13. Write a usable sand specification
The purchasing document can combine identified origin, measured composition, particle-size distribution and defined moisture. Each value should state its unit, calculation basis, method and relevant lot.
Then separate receiving criteria from industrial qualification criteria. Material meeting the purchase specification may still require the planned trial before adoption. Conversely, one favourable result does not turn incomplete analysis into routine delivery control.
The specification should identify information accompanying every lot and situations triggering another comparison: preparation changes, modified origin or deviation from an agreed criterion. Decide those triggers with the recipient rather than adding them after a problem occurs.
14. Interpret residue mass
Wet-weighed residue is not directly comparable to the dry silica introduced. The report should specify any washing, its condition when weighed and the method used to determine its composition.
In a hypothetical example, 5 kg of dry residue containing 60% SiO₂ represents 3 kg of SiO₂ in that residue. Directly subtracting all 5 kg from the initial silica contribution overstates the amount of silica not transferred. Conversely, measuring only the liquid without examining the solid leaves part of the balance unresolved.
This assessment still does not account for every loss. Samples, material retained in equipment or uncollected outputs may remain to be quantified. The objective is to make discrepancies visible rather than force every value into an equality.
15. Adapt the dossier to the Belgian recipient
Belgian context should help identify an interlocutor and their requirements rather than give a universal specification a local appearance. A manufacturer may request preparation, reporting formats or characteristics particular to its production line.
For initial contact, send a concise outline covering intended application, proposed material, available analyses, planned preparation and unresolved technical questions. The detailed dossier can subsequently include curves, methods and comparative results.
The Minerallum silica sand page provides an entry point for discussing a mineral-feedstock requirement. This project’s dossier should not confuse that discussion with an offer of liquid silicate or approval granted by a cited manufacturer.
16. Decide from the whole chain
The final decision should answer three separate questions: is the received feedstock sufficiently defined, is its behaviour compatible with the tested process, and does the resulting product meet the manufacturer’s criteria?
A favourable answer requires agreement between these levels. High SiO₂ content does not compensate for a poorly documented trial; an acceptable solution obtained after additional preparation does not remove the need to assess that preparation.
A useful dossier therefore retains unfavourable results, comparison conditions and the changes actually studied. It lets the manufacturer decide on identified feedstock with a reproducible specification and a comprehensible balance.