Deriving Mass Spectrometry Screening Thresholds for Polyester Migrants
Deriving mass spectrometry screening thresholds for polyester migrants requires lowering toxicological limits by real surface-to-volume and response uncertainty factors.
Intake
Analytical Screening Limits in Finished Articles
Analytical laboratories screen food-contact polyethylene terephthalate and related copolyesters by coupling gas chromatography or liquid chromatography to high-resolution mass spectrometry. The target of this screening is the identification and semi-quantification of non-intentionally added substances alongside unreacted monomers, catalyst residues, and processing aids. The screening limit defines the analytical concentration floor above which an unknown chromatographic peak demands structural identification, toxicological classification, and toxicological evaluation against migration thresholds under Regulation (EC) 10/2011 and Regulation (EC) 1935/2004.
Direct quantification requires pure analytical standards. Non-intentionally added substances, by definition, include reaction by-products, thermal breakdown fragments, and oligomeric cycles lacking commercial reference standards. The analytical screening limit translates a dietary exposure threshold, typically the default toxicological threshold of 0.01 milligrams per kilogram of food for unlisted substances lacking genotoxicity alerts, into an instrument-detectable response factor in an extract or migration simulant.
Derivation workflows govern whether a finished container passes inspection or faces customs impoundment under Article 3 of Regulation (EC) 1935/2004.
Under Commission Regulation (EU) 10/2011, an unlisted substance without demonstrated genotoxicity demands verification below the specific migration limit of 0.01 mg/kg in food simulant.
Analytical chemists derive screening thresholds from toxicological hazard limits down to instrument detection limits through surface-to-volume geometry, simulant concentration factors, and response-factor uncertainty. The process protects product launches by establishing defensive documentation before shipments enter customs clearance.
The standard European packaging model assumes a cube of six square decimetres surface area housing one kilogram of food simulant. Food simulants include simulant A (10% ethanol), simulant B (3% acetic acid), simulant C (20% ethanol), simulant D1 (50% ethanol), simulant D2 (vegetable oil), and simulant E (poly(2,6-diphenyl-p-phenylene oxide), known commercially as Tenax). When testing articles designed for volumes outside this ratio, real surface-to-volume adjustments govern the arithmetic.
Small-format pharmaceutical blisters, infant feeding bottles, and caps introduce geometric concentration effects that force analytical screening limits to drop by an order of magnitude.
Omission of matrix adjustments creates blind spots where migrating species remain uncharacterized below an arbitrary detection line while exceeding toxicological exposure thresholds in the intended filling volume. The failure invalidates the Declaration of Conformity.
Split
Linearity and Ion Suppression across Matrices
Mass spectrometers do not measure mass directly. They measure ion currents generated by ionization interfaces. In gas chromatography-mass spectrometry (GC-MS) operated under standard electron ionization at 70 electronvolts, fragmentation patterns remain reproducible across instruments, permitting library matching against NIST or Wiley spectra.
Liquid chromatography-mass spectrometry (LC-MS) utilizes electrospray ionization (ESI) or atmospheric pressure chemical ionization (APCI), where ionization efficiency depends upon mobile phase composition, proton affinity, analyte polarity, and concurrent matrix eluates.
Polyester migrants encompass distinct chemical spaces. Volatile breakdown products like acetaldehyde, 2-methyl-1,3-dioxolane, and residual ethylene glycol respond well to headspace GC-MS. Non-volatile migrants demand LC-MS.
These non-volatile migrants include cyclic poly(ethylene terephthalate) oligomers, designated as the cyclic trimer, cyclic tetramer, and cyclic pentamer, alongside linear oligomers terminated with carboxylic acid or hydroxyl end groups.
| Migrant Class | Representative Analyte | Ionization Interface | Ionization Sensitivity Factor |
|---|---|---|---|
| Volatile breakdown product | Acetaldehyde | Headspace GC-EI-MS | 1.0 (Baseline Reference) |
| Thermal degradation cyclic | 2-Methyl-1,3-dioxolane | GC-EI-MS | 1.2 to 1.4 |
| Plasticizer residual | Diethylene glycol dibenzoate | LC-ESI(+) / APCI(+) | 3.5 to 4.8 |
| Cyclic polyester trimer | First cyclic oligomer (m/z 577) | LC-ESI(+) as ammonium adduct | 0.3 to 0.6 |
| Cyclic polyester tetramer | Second cyclic oligomer (m/z 769) | LC-ESI(+) as ammonium adduct | 0.15 to 0.4 |
| Catalyst coordination complex | Antimony glycolate complexes | LC-ICP-MS / ESI(-) | 0.05 to 0.2 |
The cyclic oligomer series presents a classic quantification challenge. Pure standards for the cyclic trimer (triethylene terephthalate, molecular weight 576.5 Daltons) exist only through custom synthesis or preparative isolation. Standards for the cyclic tetramer, pentamer, and mixed oligomers containing isophthalic acid, diethylene glycol, or 1,4-cyclohexanedimethanol units remain scarce.
Laboratories quantify these peaks using surrogate calibration standards. The primary surrogate used for polyester screening by LC-MS is diphenyl phthalate, bisphenol A, or 2-(2-hydroxy-5-methylphenyl)benzotriazole (Tinuvin P). The assumption that an unknown cyclic oligomer responds with the same instrument sensitivity as a surrogate standard introduces response factor uncertainty.
Response Factor Distributions in Screening
The response factor represents the peak area per unit concentration. In electrospray LC-MS, response factors across diverse chemical classes span three to four orders of magnitude. A substance with poor proton affinity can elute at high concentration while yielding a tiny chromatographic peak.
A quaternary amine or highly conjugated base produces an enormous peak at trace levels.
Deriving a screening threshold requires accounting for this response factor spread. Setting the screening threshold at the response corresponding to 0.01 mg/kg for an average surrogate calibrant ensures that fifty percent of uncharacterized compounds with below-average ionization efficiency will migrate undetected above toxicological thresholds. The laboratory applies a response factor uncertainty margin.
A safety factor applied to the surrogate response lowers the operational peak area cutoff. Applying an uncertainty factor of 10 lowers the analytical threshold from 0.01 mg/kg to 0.001 mg/kg in the extract. This lower threshold captures poorly ionizable migrants, but drives instrument sensitivity demands into baseline noise regions.
Low thresholds increase identification workloads by unmasking trivial matrix artifacts.
Taper
Mathematical Derivation of the Screening Limit
Deriving the screening threshold requires a systematic series of equations moving from the toxicological limit in food down to the target concentration in the chromatography vial. Toxicological evaluation typically applies the Threshold of Toxicological Concern (TTC) concept developed by Cramer and expanded for food contact non-intentionally added substances.
Three toxicological tiers govern unlisted migrants. Genotoxic substances carrying structural alerts possess an exposure threshold of 0.00015 milligrams per person per day, translating to 0.00005 mg/kg food for an infant consuming 1 kg food per day or 0.0001 mg/kg food under adult exposure assumptions. Cramer Class III substances (high toxicity potential) possess a threshold of 0.09 mg/person/day (0.015 mg/kg food).
Cramer Class I substances (low toxicity potential) carry a threshold of 1.8 mg/person/day (0.3 mg/kg food). In the absence of toxicological hazard data, European risk assessments enforce the default cutoff of 0.01 mg/kg food as the specific migration limit baseline.
The analytical screening limit (ASL), expressed in milligrams per kilogram of food simulant, is derived through this direct relationship:
ASL = (TTC_limit V_pack) / (A_pack CF_geom UF_RF)
In this expression, TTC_limit represents the toxicological threshold in milligrams per kilogram of food. V_pack represents the packaged food volume or mass in kilograms. A_pack represents the package food-contact surface area in square decimetres.
CF_geom normalizes to standard regulatory geometry (6 dm² per 1 kg food). UF_RF is the response factor uncertainty factor, set between 1 and 10 depending on ionization interface robustness.
Extract preparation introduces volumetric concentration or dilution. When testing overall migration residues or volatile fractions, laboratories evaporate simulant extracts to dryness and reconstitute them into smaller solvent volumes before mass spectrometry injection. The concentration factor (CF_prep) shifts the concentration seen by the instrument detector:
C_detector = ASL CF_prep
CF_prep is defined as V_simulant / V_reconstituted. If two hundred millilitres of 10% ethanol migration simulant is extracted with dichloromethane, evaporated, and reconstituted in two millilitres of acetonitrile for LC-MS analysis, CF_prep equals 100. An ASL of 0.01 mg/kg in the simulant becomes 1.0 mg/L in the injection vial.
A concentration step yielding a factor of one hundred converts a sub-trace food limit into an easily measurable analytical signal on a single quadrupole detector.
Worked Calculation for Hot-Fill Trays
Take an amorphous polyethylene terephthalate (APET) thermoformed tray intended for hot-fill ready meals containing two hundred grams (0.2 kg) of acidic sauce. The food-contact surface area measures 3.5 square decimetres. The food simulant specified under Regulation (EU) 10/2011 Annex III for acidic foods is simulant B (3% acetic acid).
The testing condition selected under Annex V for hot-fill followed by extended storage is condition OM3 (two hours at 70 degrees Celsius or two hours at 100 degrees Celsius followed by 10 days at 40 degrees Celsius).
Assume an uncharacterized non-intentionally added substance without structural genotoxicity alerts, requiring compliance with the 0.01 mg/kg food threshold. First calculate the real surface-to-volume ratio: 3.5 dm² for 0.2 kg represents 17.5 dm²/kg. The regulatory standard assumption is 6 dm²/kg.
The tray presents a surface-to-volume ratio nearly triple the European default.
The permissible migration into the simulant per unit area is calculated: 0.01 mg/kg divided by 17.5 dm²/kg yields 0.00057 mg/dm². Simulating the tray using 200 mL (0.2 L) of simulant B over 3.5 dm² gives a simulant concentration of 0.01 mg/kg.
Next apply the response factor uncertainty factor for LC-ESI-MS screening using Tinuvin P as an external surrogate standard. Because electrospray ionization efficiencies for polyester decomposition fragments vary widely, assign an uncertainty factor of 5. The analytical screening limit in simulant drops to 0.01 divided by 5, equaling 0.002 mg/kg (2 micrograms per kilogram).
The testing protocol applies solid phase extraction to the 200 mL simulant volume, eluting and reconstituting into 1.0 mL of methanol-water (50:50 volume ratio). The CF_prep equals 200 mL / 1.0 mL = 200. The target analytical concentration in the chromatography vial calculates as: 0.002 mg/L multiplied by 200, which yields 0.40 mg/L (400 micrograms per litre).
The instrument quantification limit for the surrogate standard Tinuvin P must remain below 0.40 mg/L to detect uncharacterized migrants at the required toxicological threshold.
Clash
Surrogate Standards versus Authentic Response Factors
Setting screening thresholds involves structural disagreements between packaging converters, raw material resin manufacturers, and brand owners. Resin manufacturers favor higher screening limits (0.05 mg/kg to 0.01 mg/kg based directly on surrogate peak areas with an uncertainty factor of 1). They point out that modern solid-state polycondensation clean-up stages reduce volatile contaminants below toxicological concern.
Converters defend their baseline pellets, noting that cyclic oligomers represent intrinsic polymer constituents rather than impurities.
Brand testing desks, facing product recalls and customs enforcement at European ports under Rapid Alert System for Food and Feed (RASFF) alerts, demand uncertainty factors of 5 to 10. They cite documented instances where cyclic polyester oligomers, particularly the cyclic dimer containing diethylene glycol (DEG-PET dimer, molecular weight 426 Daltons), ionize poorly in electrospray positive mode compared to phthalate or UV-absorber surrogates.
The disagreement intensifies over semi-quantification reporting. When an uncharacterized chromatographic peak displays a peak area equivalent to 0.008 mg/kg using a surrogate calibration line, resin suppliers declare the substance compliant below the 0.01 mg/kg threshold. Importer quality teams reject the dossier, applying an uncertainty factor of 2 that places the estimated migration at 0.016 mg/kg, requiring formal identification and genotoxicity assessment.
The dispute resolves only through the acquisition of true standards or isolation and structural confirmation using high-resolution tandem mass spectrometry (Q-TOF or Orbitrap) and nuclear magnetic resonance spectroscopy. These structural determinations cost real money and delay production schedules by months.
Contrasting Screening Frameworks across Jurisdictions
Testing requirements vary significantly between European Union and United States jurisdictions. The table below outlines structural differences between screening frameworks for polyester articles.
| Parameter | European Union (Regulation EU 10/2011) | United States (FDA 21 CFR / Packaging Guidance) |
|---|---|---|
| Unlisted migrant baseline limit | 0.01 mg/kg food simulant (Article 11) | 0.05 mg/kg (0.5 ppb daily diet threshold of regulation) |
| Toxicological tier framework | TTC approach applied to all NIAS | Threshold of Regulation (TOR) per 21 CFR 170.39 |
| Genotoxicity alert threshold | 0.00015 mg/person/day (0.0001 mg/kg) | Excludes known carcinogens; 0.5 ppb diet limit |
| Surrogate uncertainty factor | Commonly 5 to 10 in commercial screening protocols | Typically 1 to 2; focus rests on migration chemistry |
| Cyclic oligomer compliance status | Evaluated as NIAS under Article 19 risk assessments | Tolerated as inherent polymer constituents up to TOR |
The European regulatory apparatus places legal liability directly upon the entity introducing the finished packaging into commerce. Article 19 of Regulation (EU) 10/2011 tasks the manufacturer with conducting a formal risk assessment for all non-intentionally added substances. In contrast, United States food packaging clearances rely heavily on pre-market clearances, Generally Recognized as Safe (GRAS) determinations, or food contact notifications submitted to the Food and Drug Administration.
Sieve
Workflow for Target and Non-Target Screening
Establishing analytical screening limits requires a defined laboratory sequence. The progression moves from material preparation through sample cleanup, high-resolution screening, threshold subtraction, and toxicological evaluation.
Uncertainty accumulates at each operational step. Laboratories follow disciplined laboratory protocols to prevent cross-contamination from laboratory plasticware, plasticized tubing, and analytical solvents.
- Material contact execution exposes the test article to official food simulants under controlled temperature-time regimes specified in EN 1186, using single-side contact migration cells to isolate the true contact surface.
- Extract clean-up and concentration removes non-volatile matrix interferences or concentrates trace migrants via solid phase extraction on polymeric sorbents, recording volumetric recoveries for all target controls.
- High-resolution mass spectrometry screening analyzes the extract simultaneously through GC-Q-Orbitrap and LC-ESI-Q-TOF platforms, capturing full-scan spectra across 50 to 1200 mass-to-charge ratios alongside data-dependent fragmentation spectra.
- Surrogate calibration and response normalization constructs multi-point calibration curves using reference standards spanning varied physicochemical properties, generating instrument detection limits and quantification limits.
- Threshold filtering and background subtraction eliminates procedural blank peaks and filters remaining chromatographic entities against the derived analytical screening threshold.
- Compound identification and in silico profiling annotates peaks exceeding the screening threshold using mass accuracy below 3 parts per million, isotopic matching, and spectral libraries, followed by Derek Nexus or Toxtree mutagenicity screening.
Each phase requires documented controls. If extraction recovery falls below 70 percent for reference surrogates, the analytical screening limit calculation must be adjusted upward to account for material losses during sample preparation.
Are Cyclic Oligomers Considered Additives or Non-Intentionally Added Substances?
Discussions among packaging chemists often center on the legal classification of cyclic poly(ethylene terephthalate) oligomers. Monomer starting substances like terephthalic acid and ethylene glycol appear on the Union List in Annex I of Regulation (EU) 10/2011. Cyclic oligomers form unavoidably through back-biting reactions during condensation polymerisation in the melt stage.
Because cyclic oligomers lack individual authorizations on the Union List, enforcement authorities treat them as non-intentionally added substances governed by Article 19 risk assessments. European food safety authorities have reviewed migration data for polyester oligomers, proposing indicative toxicological limits for the sum of cyclic oligomers up to 5.0 milligrams per kilogram of food based on 90-day oral toxicity studies. Toxicological consensus remains divided regarding whether small cyclic dimers carrying ether linkages from diethylene glycol demand lower specific thresholds than the standard cyclic trimer.
If an analytical laboratory groups all oligomeric peaks into an aggregate sum, the screening limit derivation applies to total peak area. Treating each oligomer as an independent unknown forces the application of the stringent 0.01 mg/kg default threshold to each isolated peak. This analytical choice dictates whether an article passes or fails compliance testing.
Stamp
Documenting Threshold Derivations in the Conformity Dossier
Test reports lacking complete operational parameters fail regulatory audits. An inspector reviewing a technical dossier examines the mathematical lineage linking the measured peak area to the declared food contact application. The Declaration of Conformity relies on the supporting technical file required by Article 16 of Regulation (EU) 10/2011.
The technical dossier must contain clear data items to withstand scrutiny by commercial buyers or national health authorities:
- Target food types identifying whether the packaging contacts fatty foods, acidic aqueous liquids, alcoholic solutions, or dry bulk foodstuffs, referencing official simulant allocations.
- Contact time and temperature justifying conditions like OM2 (10 days at 40°C) or OM5 (2 hours at 100°C) based on real-world supply chain limits, filling processes, and maximum end-use shelf life.
- Surface-to-volume ratio derivations documenting precise surface areas and filling masses, replacing the generic 6 dm²/kg default with calculated geometries when handling non-standard formats.
- Surrogate calibrant response data reporting calibration curve equations, linear regression coefficients, instrument detection limits, and justified response factor uncertainty adjustments.
- Recovery percentages providing measured extraction efficiencies across procedural blanks and fortified matrix samples.
Omission of these data parameters prevents the verification of reported compliance limits. When an importer signs a declaration based on incomplete laboratory summaries, the importer assumes direct legal responsibility for uncharacterized migration products.
A laboratory report lacking recovery figures and ionization uncertainty factors provides no legal defense during a port-of-entry audit.
The financial impact of inaccurate threshold derivations is direct and substantial. Commercial supply contracts across the European Union incorporate strict non-compliance indemnification provisions. If national border inspectors issue a RASFF alert due to unauthorized polyester migrants exceeding toxicological screening thresholds, the commercial buyer initiates product withdrawals across distribution networks.
Quarantine costs, testing re-examinations, inventory write-offs, and hazardous waste disposal quickly exceed hundreds of thousands of euros on high-volume product launches. Supply agreements routinely assign these losses to converters when supporting technical dossiers fail to demonstrate traceable screening derivations. A properly structured conformity dossier protects capital, validates the manufacturing supply chain, and maintains market access across international borders.
Under standard commercial packaging contracts, a delivery lot remains unaccepted until the technical file demonstrates analytical detection limits calibrated to the specific fill volume of the production article.