Elimination of Hexavalent Selenium Interference in Dairy Product Analysis

Accurate selenium determination in dairy products depends critically on eliminating interference from hexavalent selenium (Se(VI)), which otherwise causes significant underestimation of total selenium during hydride generation and mass spectrometric analysis. This technical article explains the chemical origins of Se(VI) interference across milk, milk powder, yogurt, cheese, and infant formula matrices, and reviews validated digestion, pre-reduction, and instrument-based mitigation strategies consistent with GB 5009.93, GB 5009.268, and internationally recognized ICP-MS practices. Laboratory managers, quality assurance personnel, and food safety professionals will find practical guidance on method selection, interference diagnostics, and quality control for selenium testing in dairy.

Technical Overview

Selenium is an essential trace element with a narrow margin between nutritional adequacy and excess, and dairy products are a meaningful dietary source, particularly for infants and young children. In dairy matrices, selenium occurs as inorganic species, mainly tetravalent selenite Se(IV) and hexavalent selenate Se(VI), and as organic species, predominantly selenomethionine and selenocysteine derived from fortified selenized yeast or feed-derived transfer. Because regulatory and nutritional decisions are based on total selenium, all species must be measured quantitatively and reproducibly.

The core analytical problem is that hexavalent selenium behaves as a direct interferent in the most widely used determination techniques. In hydride generation atomic fluorescence spectrometry (HG-AFS), potassium borohydride reduces Se(IV) efficiently to volatile selenium hydride, but Se(VI) is kinetically inert under normal carrier-acid conditions and does not form hydride. Any Se(VI) that escapes pre-reduction is therefore invisible to the detector, producing a negative bias that may reach 20 to 50 percent depending on the Se(VI)/Se(IV) ratio established during digestion and fortification. In inductively coupled plasma mass spectrometry (ICP-MS), selenium isotopes suffer intense argon dimer polyatomic interferences, notably 40Ar40Ar+ on 78Se and 80Se, while residual hydrochloric acid from reduction steps introduces additional chloride-based polyatomic overlaps at m/z 74 and 77. The combined effect of incomplete Se(VI) reduction, spectral overlap, and dairy matrix suppression can compromise data quality unless a systematic interference elimination protocol is applied.

Dairy matrices further complicate the analysis because of their high protein, fat, lactose, and calcium phosphate content. Incomplete destruction of organic matter releases organoselenium that must be converted to inorganic selenium, and variable residual carbon in digests alters selenium ionization efficiency in ICP-MS through the well-documented carbon enhancement effect. Consequently, interference elimination in dairy testing involves three coordinated layers: complete matrix digestion, quantitative reduction of Se(VI) to Se(IV) where hydride generation is used, and instrumental or mathematical correction of spectral interferences where ICP-MS is used. This article addresses each layer in the context of the applicable Chinese national food safety standards, including GB 5009.93 for selenium determination and GB 5009.268 for multi-element ICP-MS analysis, under quality systems conforming to ISO/IEC 17025.

Testing Samples

Liquid Dairy Products

Raw milk, pasteurized milk, UHT milk, modified milk, and milk-based beverages require thorough homogenization before subsampling because selenium partitions between the aqueous phase and the milk fat globule membrane. Representative sampling typically involves warming to 35 to 40 degrees Celsius, gentle inversion, and immediate weighing of the homogeneous aliquot into digestion vessels. Flavored and fortified liquid products may contain added selenite or selenate from mineral premixes, which raises the importance of Se(VI) reduction completeness.

Milk Powders and Infant Formula

Whole milk powder, skim milk powder, infant formula, and follow-on formula present high solids and, when fortified with selenized yeast, a significant fraction of selenomethionine-bound selenium. Powder samples are homogenized by tumbling or grinding, and moisture is determined in parallel so results can be reported on both as-is and dry-matter bases. Infant formula testing carries the strictest stakes, since selenium levels fall within tightly controlled fortification windows defined by GB 14880 and corresponding international compositional requirements.

Fermented Dairy Products and Cheese

Yogurt, fermented milk drinks, and natural or processed cheese contain elevated protein and fat levels, and cheese additionally introduces high sodium chloride. Salt increases the risk of chloride polyatomic interferences in ICP-MS and of elevated blank contribution, so digestion protocols for cheese emphasize complete organic destruction and careful acid volume control. Hard cheeses may require particle-size reduction below one millimeter before digestion to ensure representative aliquots.

Whey and Milk Protein Ingredients

Whey protein concentrates and isolates, caseinates, and milk protein concentrates exhibit protein contents of 35 to 90 percent. These matrices demand extended acid volumes or repeated peroxide additions during digestion, and the concentrated mineral premixes sometimes co-formulated with them can deliver selenium almost entirely as selenate, making pre-reduction verification especially important for hydride-based methods.

Cream, Butter, and High-Fat Dairy Products

High-fat samples are prone to incomplete oxidation and to fat layering during subsampling. Practical measures include reduced sample mass, pre-extraction or absorption of free fat on an inert support where validated, and microwave digestion programs with staged ramping to prevent rapid pressure excursions from lipid oxidation.

Testing Items

Interference elimination studies for dairy selenium typically encompass the following measurable deliverables, each reported with full method validation data:

• Total selenium content, expressed in mg/kg or μg/100 g, on as-received and dry-matter bases.
• Inorganic selenium speciation, quantifying Se(IV) and Se(VI) separately by HPLC-ICP-MS to characterize the interference source.
• Organic selenium fractions, principally selenomethionine in yeast-fortified formulas, following enzymatic extraction.
• Pre-reduction efficiency verification, comparing hydride response of a digested sample before and after Se(VI)-to-Se(IV) conversion.
• Method performance characteristics, including limit of detection, limit of quantification, repeatability and reproducibility precision, spike recovery, linearity, and measurement uncertainty budgets.
• Matrix interference diagnostics, such as carbon enhancement assessment, chloride carryover, and transition-metal suppression tests for hydride generation.

Testing Methods

Homogenization and Digestion

Total selenium methods begin with closed-vessel microwave digestion using nitric acid and hydrogen peroxide, which is the reference approach in GB 5009.93 and GB 5009.268. Typical programs ramp to 180 to 200 degrees Celsius with a hold of 15 to 30 minutes, ensuring complete destruction of proteins and fats so that organoselenium is converted to inorganic selenium species. Residual nitric acid is either allowed to persist for ICP-MS work or gently evaporated and replaced with hydrochloric acid medium when the digest proceeds to hydride generation. Digest blanks, duplicate digestions, and certified reference materials such as skim milk powder and infant formula reference materials from national metrology institutes accompany every batch.

Pre-Reduction of Se(VI) to Se(IV)

This step is the heart of hexavalent selenium interference elimination for hydride-based determination. Several validated strategies exist, and selection depends on throughput, matrix, and validation history:

• Thermal reduction with hydrochloric acid: the digest is adjusted to approximately 6 mol/L HCl and heated at 90 to 100 degrees Celsius for 10 to 30 minutes, quantitatively converting Se(VI) to Se(IV). This is the principal approach embedded in the national standard method for selenium in foods.
• Mixed-reducer systems: thiourea combined with ascorbic acid in hydrochloric acid medium reduces Se(VI) at room temperature over 20 to 30 minutes and simultaneously masks transition metals such as nickel, copper, and cobalt that otherwise suppress hydride formation.
• Bromate-bromide reduction: potassium bromide and potassium bromate in hydrochloric acid generate elemental bromine in situ, an aggressive classical reducer, followed by boiling to expel excess bromine before hydride generation.
• Microwave-assisted or ultraviolet-assisted reduction: closed-vessel or UV photolytic reduction shortens the conversion to a few minutes, minimizes selenium loss, and improves reproducibility for high-throughput dairy laboratories.
• Over-reduction safeguards: reagent concentrations and heating times must be validated to avoid reducing Se(IV) further to colloidal elemental selenium, which produces signal loss and poor precision; verification with mixed Se(IV)/Se(VI) spikes at several ratios is recommended.

Hydride Generation Atomic Fluorescence Spectrometry

After pre-reduction, the sample is carried in dilute hydrochloric acid into a gas-liquid separator where 0.5 to 2 percent potassium borohydride in dilute alkali generates hydrogen selenide, which is swept by argon into an argon-hydrogen flame for atomic fluorescence detection with a selenium hollow cathode or electrodeless discharge lamp. Calibration uses matrix-matched or standard-addition curves spanning the regulatory range, and quality checks include periodic re-reads of a mid-level standard and continuous blank monitoring. Typical method detection limits reach the 0.1 to 0.5 μg/kg range for dairy, with spike recoveries normally falling between 80 and 120 percent when reduction is complete.

ICP-MS with Collision and Reaction Cell Correction

For multi-element workflows under GB 5009.268, selenium is measured by ICP-MS with deliberate management of argon dimer interferences. Kinetic energy discrimination with helium collision gas attenuates 40Ar40Ar+ and 40Ar38Ar+ sufficiently for 78Se and 80Se, while hydrogen or methane reaction gas chemistry converts argon dimers into non-interfering species and offers the highest sensitivity at mass 80. Isotope selection balances abundance against interference: 80Se provides the strongest signal with effective cell correction, 82Se offers the cleanest background where bromine is negligible, and 74Se is avoided because of chloride dimer overlap. Germanium or tellurium internal standards compensate for matrix suppression from calcium, phosphorus, and potassium. Residual carbon in digests enhances selenium response, so laboratories either matrix-match calibration with a controlled carbon content or add a small fixed proportion of alcohol to all standards and samples. Residual hydrochloric acid is minimized before ICP-MS measurement to limit chlorine-based polyatomics, and hydrogen peroxide is fully decomposed to prevent carbon deposition on the interface cones.

Speciation Analysis by HPLC-ICP-MS

When the interference source itself must be characterized, species are extracted without changing oxidation state. Enzymatic hydrolysis with protease and lipase at near-neutral pH, sometimes assisted by ultrasonic energy, releases selenomethionine, Se(IV), and Se(VI) from dairy protein, and the extract is separated on an anion-exchange column coupled directly to the ICP-MS. Chromatographic resolution allows each species to be quantified at its retention time, converting the interference question into a direct speciation measurement and providing the evidence base for selecting the appropriate reduction chemistry.

Quality Control and Data Integrity

Every interference elimination protocol must be supported by documented control measures executed within an ISO/IEC 17025-conformant system:

• Reduction completeness checks using independent Se(VI) spikes that must recover within the validated acceptance window.
• Certified milk powder and infant formula reference materials analyzed in each batch with results inside certified uncertainty ranges.
• Digestion blanks and reagent blanks to control borohydride, acid, and peroxide selenium contribution.
• Duplicate preparations with relative deviation limits, typically below 10 percent at levels near the limit of quantification.
• Control charts for internal standard recovery, calibration verification, and collision cell performance to detect drift that could mask interference.
• Participation in interlaboratory proficiency testing schemes covering selenium in dairy matrices.

Applications

Interference-free selenium data from dairy products support a broad set of regulatory, commercial, and scientific applications:

• Compliance verification against fortification limits and permitted selenium compounds established by GB 14880 for dairy categories, including infant and follow-on formula.
• Confirmation that selenium levels remain within safe upper boundaries in raw milk and processed dairy, historically framed by GB 2762 contaminant provisions and their successive revisions.
• Nutrition labeling support, ensuring declared selenium values meet rounding and tolerance requirements for domestic and export markets.
• Incoming material qualification of selenized yeast and mineral premixes, where distinguishing selenomethionine from selenite and selenate forms is a contractual specification.
• Import and export conformity assessment, with testing performed by accredited third-party laboratories offering internationally traceable results.
• Process research on selenium retention through evaporation, spray drying, and heat treatment, and nutritional studies on selenium bioaccessibility from fermented dairy.
• Forensic and complaint investigations where anomalous selenium findings require ruling out analytical interference as the underlying cause.

FAQ

Why does hexavalent selenium interfere with selenium determination in dairy?

Se(VI) does not react with borohydride to form volatile hydride under the acidic conditions used in hydride generation, so unreduced Se(VI) is not measured, causing total selenium to be underestimated. In ICP-MS, the selenium region is additionally crowded by argon dimer and chloride polyatomic ions that must be removed by collision or reaction chemistry.

How large is the negative bias if Se(VI) reduction is incomplete?

The bias equals the unreduced Se(VI) fraction, which in dairy can be substantial when selenate-containing mineral premixes are used. Practical validation commonly shows biases of 20 to 50 percent for untreated digests, which is why pre-reduction is mandatory rather than optional in standard methods.

Which pre-reduction reagent is preferred for dairy samples?

Thermal reduction in approximately 6 mol/L hydrochloric acid at near-boiling temperature for 10 to 30 minutes is the reference approach in the national standard method. Thiourea-ascorbic acid mixtures are a convenient room-temperature alternative that also masks transition metals, while microwave-assisted reduction suits high-throughput laboratories.

How is complete Se(VI)-to-Se(IV) conversion verified?

Laboratories spike digests with known Se(VI) amounts and confirm that recoveries fall within validated limits, compare hydride response before and after reduction, and periodically perform chromatographic speciation to confirm that no measurable Se(VI) remains in the treated solution.

Does hydrochloric acid used for reduction cause problems in ICP-MS?

Chloride introduces polyatomic overlaps on selenium isotopes at m/z 74 and 77 and can elevate argon chloride backgrounds. For ICP-MS workflows, hydrochloric acid is either evaporated after reduction and the residue reconstituted in nitric acid, or cell gas and isotope selection are adjusted to control chlorine-based interferences.

What recovery and precision criteria apply to dairy selenium testing?

Spike recoveries of 80 to 120 percent are widely accepted across the regulatory range, with tighter targets at fortified levels, and repeatability relative standard deviations typically below 10 percent. Results must be supported by certified reference materials, blanks, duplicates, and documented measurement uncertainty.

How are high-fat matrices such as cheese handled without introducing error?

Representativeness is protected by fine grinding and thorough homogenization, sample mass is reduced to keep the acid-to-organic ratio high, and digestion programs are staged to manage lipid oxidation. Validation with a high-fat certified reference material confirms that neither selenium loss nor incomplete digestion compromises the result.