Determination of Silicon Content in Recycled Aluminum: Methods, Standards, and Applications

Silicon is the single most influential alloying element in recycled aluminum, and its accurate determination is essential for scrap sorting, melt adjustment, ingot certification, and product conformity. This article explains the sample types, testing scope, and internationally recognized analytical methods for measuring silicon in recycled aluminum, including gravimetry, molybdenum blue spectrophotometry, spark OES, ICP-OES, and XRF. Applicable standards such as ISO 797, ISO 808, ASTM E34, ASTM E1251, GB/T 20975, and EN 1676 are also discussed, together with typical applications and frequently asked questions.

Technical Overview

Recycled (secondary) aluminum is produced by remelting aluminum scrap such as used beverage cans, end-of-life vehicle castings, extrusion and rolling process scrap, cables, and building products. Silicon is the principal alloying element in the cast Al-Si family that dominates the scrap stream, and because secondary smelters inevitably mix low-silicon wrought grades (1xxx, 3xxx, 5xxx, 6xxx) with high-silicon die-casting alloys and hypereutectic piston alloys, the silicon content of recycled metal can range from a few hundredths of a percent to more than 20 percent by mass.

Silicon level directly controls fluidity and castability, hot-tear resistance, feeding and shrinkage behavior, melting range, hardness, wear resistance, machinability, and the effectiveness of eutectic modification with strontium or sodium. In secondary smelting, the certified silicon value drives alloy sorting, dilution and compounding calculations, and grade assignment against specifications such as EN 1676 for alloyed remelting ingots, GB/T 8644 for recycled aluminium ingots, and ASTM B179 for remelting ingots. Because traded ingots are typically settled commercially on the certified chemical analysis, accurate silicon determination is both a quality requirement and a commercial necessity.

• Low-silicon wrought remelt stock (1xxx/5xxx types): typically below 0.5% Si, controlled as an impurity
• Extrusion and rolling alloys (6xxx types): approximately 0.2% to 1.2% Si
• General die-casting alloys (A380-type and EN AC-46000 AlSi9Cu3 types): approximately 7.5% to 11% Si
• AlSi12 family (A413-type and EN AC-47100 types): approximately 10.5% to 13.5% Si
• Hypereutectic piston alloys (390-type): approximately 16% to 20% Si and above

Testing Samples

Recycled Ingots, Sows, and T-Bars

Representative material is obtained by drilling or milling across the cross-section of the cast unit at multiple locations, following sampling practices such as ASTM E716 and corresponding GB sampling standards. Clean, degreased drillings free of oxides and lubricants are collected, typically 5 g to 20 g per unit, to capture both the bulk composition and any localized heterogeneity introduced by the remelting process.

Aluminum Scrap and Shredded Material

Scrap is the most heterogeneous input to secondary smelters. Incremental sampling and coning-and-quartering, or melting a composite portion into a chill-cast button before analysis, are used to obtain a statistically meaningful sample. The analytical result applies strictly to the submitted sample, so documented sampling protocols at accredited third-party laboratories are essential for settlement-grade results.

Molten Metal Samples from Furnace or Launder

Chill-cast discs or pucks are taken from the melt using standardized ring molds to produce a fine, rapidly solidified, homogeneous microstructure suitable for spark OES. Sampling at the start, middle, and end of a pour allows the smelter to track composition drift during casting.

Finished Cast and Wrought Products

Die castings, extrusions, forgings, and rolled products are sampled by sectioning, drilling, or remelting small coupons, with the sampling location chosen away from segregations, runners, and surface coatings wherever possible.

Testing Items

The primary testing item is the silicon content, reported as mass fraction in weight percent, with a typical reportable range of approximately 0.001% to above 25% depending on the selected method. In practice, silicon in recycled aluminum is determined as part of a full compositional package, because secondary alloy specifications are inherently multi-element.

• Silicon (Si) as the key controlled element
• Major and minor alloying elements: iron, copper, magnesium, manganese, zinc, nickel, titanium, chromium
• Trace and tramp elements: lead, tin, vanadium, calcium, sodium, strontium, and others relevant to scrap origin
• Optional melt-quality parameters where required, such as hydrogen content and inclusion assessment

Testing Methods

Gravimetric Method (Silicic Acid Dehydration)

The classical referee technique, standardized in ISO 797, ASTM E34, and the corresponding parts of GB/T 20975. The sample is dissolved in sodium hydroxide, often with hydrogen peroxide, in nickel or PTFE vessels; the solution is acidified with perchloric or sulfuric acid and evaporated to fuming to dehydrate the silicic acid. The precipitated silica is filtered, ignited, and weighed, then volatilized with hydrofluoric acid so that non-volatile impurities can be corrected gravimetrically. The method is applicable from roughly 0.3% Si upward, offers the lowest bias, and is used for certification, calibration, and dispute resolution, at the cost of longer turnaround.

Molybdenum Blue Spectrophotometry

Standardized in ISO 808, GB/T 20975.5, and related JIS H methods, this technique converts reactive silicate into yellow silicomolybdic acid at controlled acidity and reduces it to molybdenum blue, whose absorbance is measured near 810 nm. The working range is typically 0.005% to 0.5% Si, making it ideal for low-silicon wrought grades and purity verification of remelt stock. Hot alkaline dissolution must be performed in nickel or plastic vessels, since sodium hydroxide leaches silicon from glassware and would bias results.

Spark Source Optical Emission Spectrometry (Spark OES)

The workhorse of melt and product control, covered by ASTM E1251, ASTM E607, and GB/T 7999. A lathe-turned, chill-cast surface is sparked in argon, and silicon is measured on sensitive emission lines such as 288.16 nm, 251.61 nm, or 390.5 nm using calibration curves built from matrix-matched certified reference materials. A full determination takes only seconds, and precision in the 5% to 13% die-casting range is excellent. For recycled hypereutectic alloys containing coarse primary silicon, multiple burn positions and appropriate reference materials are required to address micro-segregation.

ICP-OES (Inductively Coupled Plasma Optical Emission Spectrometry)

The sample is dissolved via sodium hydroxide and peroxide followed by acidification, or by closed-vessel microwave digestion with hydrofluoric acid that is subsequently complexed with boric acid. Matrix-matched standards, internal standardization, and carefully selected silicon wavelengths (for example 212.412 nm, 251.611 nm, or 288.158 nm) yield detection limits around 0.001% while simultaneously determining the full element package. Methods such as GB/T 20975.25 and ASTM E3061 provide the standardized basis, and ICP-OES is preferred when silicon and trace elements must be reported from a single digest.

X-Ray Fluorescence (XRF) Screening

Benchtop and handheld XRF instruments provide rapid, non-destructive screening of ingots, billets, and sorted scrap, and fused-bead preparation can improve accuracy in laboratory systems. Because surface condition and matrix effects limit accuracy relative to OES and ICP, XRF is generally used for sorting support and pre-screening rather than certification.

Sample Preparation and Quality Control

Surfaces must never be prepared with silicon carbide abrasive paper before silicon determination; lathe turning with clean steel or carbide tools, or pure alumina abrasives, must be used instead. Glassware is avoided during alkaline dissolution steps, reagent blanks are controlled, certified reference materials bracket the calibration range, and replicate determinations at multiple locations are used to quantify the inherent heterogeneity of recycled metal. Reports from accredited third-party laboratories state the method, the standard applied, and the measurement uncertainty.

Applications

Silicon determination in recycled aluminum underpins virtually every stage of the secondary aluminum value chain, from scrap purchasing to final product release.

• Scrap sorting, valuation, and settlement analysis for secondary smelters and traders
• Charge calculation, dilution, and melt adjustment during remelting and alloying
• Certification of remelt ingots to specifications such as EN 1676, GB/T 8644, and ASTM B179
• Production control of secondary AlSi die-casting alloys for automotive powertrain, body, and structural components
• Quality assurance of hypereutectic piston and wear-resistant alloys
• Recycled-content declarations and carbon-footprint reporting in circular-economy programs
• Incoming inspection and failure analysis of castings and wrought products made from secondary metal

FAQ

What silicon range can be measured in recycled aluminum?

Depending on the method, silicon can be quantified from approximately 0.001% by spectrophotometry, ICP-OES, and spark OES up to more than 20% by gravimetry and OES. The method is selected according to the expected concentration, the alloy matrix, and the required turnaround.

Which method is the most accurate?

The gravimetric silicic acid dehydration method per ISO 797 and ASTM E34 is recognized as the referee method with the lowest systematic bias. Spark OES is preferred for routine speed, and ICP-OES offers the best combination of trace capability and multi-element coverage.

Why do spark OES and wet-chemical results sometimes differ on recycled metal?

Recycled aluminum can contain coarse primary silicon particles and undissolved hard spots. Spark OES analyzes only a micro-volume per burn, whereas wet chemistry integrates grams of sample. Chill-cast melt sampling, multiple burn positions, and averaging bring the two techniques into agreement.

How is representativeness ensured for scrap samples?

Through incremental sampling, coning-and-quartering, or melting a composite portion into a chill-cast button before analysis. Accredited third-party laboratories apply documented sampling protocols and report sampling uncertainty along with the analytical result.

How much sample is needed and how long does testing take?

Typically 20 g to 50 g of clean drillings or one chill-cast disc is sufficient. Spark OES results can be issued the same day, while gravimetric and ICP determinations usually require three to five working days.

Are there contamination risks during preparation?

Yes. Silicon carbide grinding media and glassware exposed to hot sodium hydroxide both introduce positive silicon bias. Standardized procedures eliminate both sources, and reagent blanks verify the control.