Screen 120 Additives in One DIA Run: Additive Screening for Formulators

For broad discovery work, LC/GC-HRMS non-targeted workflows identify the widest range of unknown or suspect additives in a single run. When regulatory compliance or trace-level quantitation is the goal, targeted LC-MS/MS or GC-MS methods in MRM mode deliver the sensitivity that non-targeted screening cannot match. Direct-analysis techniques and high-throughput or Bayesian optimization approaches serve triage and formulation-selection tasks where speed outweighs comprehensive sensitivity, provided every workflow is validated against matrix-matched standards with documented limits of detection, limits of quantitation, and recovery data.

---

TL;DR:

• Non-targeted HRMS can screen up to 120 additives in a single run but generally has detection limits in the 0.02 to 20 ppm range, risking missed trace contaminants.
• Targeted MRM methods achieve low parts-per-billion sensitivity but only detect pre-listed compounds, making them unsuitable for discovering unknowns.
• Sample preparation choices like QuEChERS, SPE, and headspace analysis are critical, with proper validation and use of internal standards reducing false negatives.
• Instrument selection depends on analyte chemistry; LC-HRMS suits polar, non-volatile additives, while GC-HRMS is better for volatiles and pyrolysis fragments.
• Combining non-targeted HRMS data, Bayesian optimization, and targeted confirmatory work creates efficient, adaptable screening workflows for regulatory and formulation needs.

---

• Where additive screening is heading next
• Sources
• FAQ

Targeted, non-targeted, and rapid direct-analysis: how the methods differ

Additive screening splits into three functional categories, and choosing the wrong one wastes instrument time without answering the analytical question. Targeted methods search for a defined list of known compounds using multiple reaction monitoring (MRM), delivering high sensitivity and reliable quantitation but no information beyond the target list. Non-targeted screening (NTS), by contrast, captures full-spectrum data so unknowns and suspects can be identified retrospectively, without rerunning the sample.

High-resolution mass spectrometry (HRMS) platforms acquire data in several modes that determine what a lab can extract from a single run. Full-scan acquisition records all ions above a threshold, forming the foundation of non-targeted work. Data-independent acquisition (DIA) fragments broad mass ranges systematically, which is what allows a DIA-based workflow to screen 120 additives across colorants, sweeteners, preservatives, and antioxidants in one analytical sequence. Data-dependent acquisition (DDA) selects ions for fragmentation based on real-time intensity, useful for suspect confirmation but less exhaustive than DIA. Parallel reaction monitoring (PRM) sits between the two, targeting a curated list with HRMS resolution.

Sensitivity differs sharply between these approaches. The same 120-additive DIA study reported LODs typically in the 0.02 to 20 ppm range for colorants and 0.1 to 10 ppm for other additive classes, a full order of magnitude or more above what optimized targeted MRM can reach for equivalent analytes. That gap defines the operational trade-off:

• Non-targeted HRMS suits broad surveys, supplier screening, and detecting unexpected compounds, but its ppm-level sensitivity may miss trace contaminants relevant to strict regulatory thresholds.
• Targeted MRM reaches low-ppb sensitivity for a defined analyte list, at the cost of blindness to anything not on that list.
• DIA data can be reinterrogated retrospectively as new suspect lists emerge, avoiding repeat sampling.
• Instrument and data-processing burden rises with acquisition breadth: full-scan and DIA runs generate larger datasets that demand more processing time per sample than a lean MRM method.

Sample preparation and extraction strategies by matrix

Extraction choice often determines whether an additive is detected at all, regardless of instrument quality. Matrix complexity, analyte polarity, and required sensitivity dictate which prep method fits.

1. QuEChERS remains the standard first choice for food matrices, pairing rapid extraction with dispersive cleanup to remove lipids and pigments before LC-HRMS analysis; validated protocols have paired QuEChERS extraction with fast UHPLC-HRMS runs of around fifteen minutes to detect multiple additive classes in dairy and other complex foods.
2. Solid-phase extraction (SPE) concentrates trace analytes from liquid matrices and removes interferences ahead of targeted quantitation, particularly for polar additives that QuEChERS cleanup steps may not fully resolve.
3. Solid-phase microextraction (SPME) offers solvent-free concentration for volatile and semi-volatile compounds, well suited to headspace sampling of coatings and packaging materials.
4. Headspace and thermal desorption techniques isolate volatile organic compounds (VOCs) and semi-volatile organic compounds (SVOCs) directly from solid or liquid samples without solvent extraction, reducing background interference in GC-based analysis.
5. Pyrolysis-GC/MS breaks down polymer-bound additives that resist solvent extraction, generating characteristic thermal decomposition fragments; a 2026 review of polymer additive analysis identifies pyrolysis and thermal desorption coupled to GC-HRMS as effective for additives that are poorly soluble or strongly bound within the polymer matrix.
6. Migration experiments for food contact materials require simulant selection matched to the food type and contact conditions, following established protocols so results reflect realistic exposure rather than an artifact of an overly aggressive solvent.

Coelution and matrix suppression remain the most common causes of false negatives in screening work. Staggering chromatographic gradients, adjusting cleanup selectivity, and diluting extracts to reduce matrix load all help, but the more reliable fix is building in internal standards and surrogate standards from the start, since they flag suppression and recovery losses that would otherwise go unnoticed until a result fails validation.

Pro Tip: Run a matrix-matched blank alongside every new sample type before trusting a screening result, since matrix effects vary enough between batches that a single validation run rarely covers future production lots.

Instrument selection guide by additive class and matrix

No single instrument covers every additive class, and matching technique to analyte chemistry avoids wasted runs.

• LC-HRMS (Orbitrap or time-of-flight) handles polar, non-volatile additives and is the workhorse for non-targeted and retrospective screening; polarity switching within a single run captures both acidic and basic species, though cycle time constraints mean wide isolation windows must be balanced against the number of points collected across each narrow UHPLC peak.
• GC-HRMS and GC-MS suit volatile and semi-volatile compounds, including pyrolysis fragments from polymer-bound additives, and benefit from extensive spectral library matching under electron ionization (EI), though EI's fixed fragmentation pattern limits structural flexibility compared to soft ionization sources.
• FTIR and Raman spectroscopy provide fast identity checks and surface-level analysis at higher concentrations, useful for confirming an additive class in a coating film or plastic surface, but neither technique reaches the sensitivity needed for trace quantitation.
• NMR confirms molecular structure and resolves complex mixtures when analyte concentration is sufficient, serving as a complementary technique rather than a primary screening tool.
• Thermal and elemental methods, including TGA-MS and ICP, infer additive classes from decomposition behavior or detect inorganic additives such as metal-based stabilizers and flame retardants that organic-focused chromatography would miss.
• Direct-analysis techniques such as DART-MS and AP-MALDI trade sensitivity for speed, making them suited to rapid triage rather than confirmatory work.

A comprehensive review of polymer additive detection recommends combining spectroscopic, chromatographic, elemental, and thermal methods rather than relying on any single platform, noting that low additive concentrations and strong matrix interactions are the recurring analytical challenges across polymer systems. A systematic review of non-targeted screening for food contact materials similarly identifies LC-HRMS and GC-HRMS as the dominant platforms for that application, with Orbitrap instruments generally offering better mass accuracy and stability for retrospective, repeated interrogation of stored data than time-of-flight systems in many lab settings.

Speeding up screening with high-throughput and Bayesian methods

Exhaustive high-throughput experimentation (HTE) tests large additive libraries across plate-based formats, generating broad coverage but at proportional cost in reagents, instrument time, and data handling. Every additional condition tested adds real expense, which is why labs increasingly pair HTE with computational methods that narrow the search space before committing plate space to it.

Bayesian optimization (BO) offers a data-driven alternative. Rather than testing every candidate, BO builds a surrogate model of expected performance, uses an acquisition function to select the next most informative experiment, and updates its model after each result. The approach depends heavily on how additives are represented numerically and how the search is initialized.

• Chemical representations that capture structural similarity, rather than simple one-hot encoding, generally improve BO performance on additive screening tasks.
• Clustering-based initialization strategies help BO explore diverse chemical space early rather than converging prematurely on a local optimum.
• Adaptive kernel choices in the surrogate model improve predictions as more data accumulates across screening rounds.

One study applying BO to 720 additives across four reactions found that BO configurations using clustering initialization and structure-aware representations outperformed both random search and standard one-hot encoding approaches, reaching high-performing additives with far fewer experimental runs than exhaustive testing would require. Pairing BO's narrowed candidate list with a rapid direct-analysis assay creates a practical triage pipeline: BO proposes promising candidates, and a fast screening technique confirms them before committing to full formulation trials.

Choosing the right screening strategy for your project

Matching a workflow to a project starts with four questions: what matrix are you working with, how volatile or polar is the target analyte class, what limit of detection does the application require, and what throughput and budget constraints apply. A regulatory submission demanding low-ppb confirmation calls for a different method than a supplier qualification screen looking for unexpected additive classes.

1. Triage with a rapid or non-targeted method to survey what is present without committing to expensive targeted runs.
2. Suspect identification narrows the field using library matching, retrospective HRMS data, or BO-guided candidate ranking.
3. Targeted confirmation applies MRM or PRM methods to quantify the specific compounds flagged during triage.
4. Validation locks in the method with calibration curves, recovery studies, and matrix-matched standards before results are used for decisions.

Before trusting any result, confirm the minimum validation checklist: calibration across the expected concentration range, spike-and-recovery data, matrix-matched standards rather than solvent-only calibration, documented LOD and LOQ, and repeat-injection precision. Inter-lab comparison, where feasible, adds confidence that a method performs consistently outside a single lab's conditions. Standards bodies such as ASTM publish validated extraction and separation procedures for polyolefin additives, reporting LODs around 2 ppm for phenolic antioxidants under optimal conditions, a useful benchmark when qualifying an in-house method against an established one.

Pro Tip: Treat inconsistent recovery across replicate extractions as a red flag before treating it as a result, since it usually points to matrix suppression or coelution rather than genuine analyte variability.

How screening data supports traceability workflows for industrial materials

Formulation labs rely on screening results well beyond initial detection work. Batch-to-batch consistency checks, migration testing, and compositional verification all feed into decisions about which additive suppliers earn a long-term qualification. A dispersant or defoamer that performs well in a single batch but shows compositional drift across shipments creates downstream defects that only show up after a customer's production run, which is why consistent screening data matters as much for procurement as for regulatory compliance.

Technical support work with formulators typically centers on documented batch consistency and qualification data, so a supplier's screening results are traceable to specific production lots rather than a single sample. Readers evaluating a switch or a new qualification path can review practical procurement checklists in how to qualify additive suppliers before the first order and the QC metrics that distinguish a reliable coating additive producer, both of which translate screening data into supplier-selection criteria.

Where additive screening is heading next

Non-targeted screening has moved from a specialist research tool to something formulation labs of moderate size can realistically adopt, though the adoption curve remains uneven: instrument cost and data-processing expertise still gate entry for smaller operations. Over the next few years, expect tighter integration between machine learning and both non-targeted HRMS data and Bayesian optimization, since the bottleneck in most labs is no longer generating spectra but interpreting them fast enough to act on.

Smaller labs considering rapid screening or BO approaches do not need to build the most sophisticated pipeline first. Starting with a well-validated targeted method for known priority analytes, then layering in DIA-based non-targeted runs for retrospective flexibility, builds capability without overcommitting budget to infrastructure the team is not yet ready to use.

— Astra R&D Team

Sources

Readers validating a new method or benchmarking against established protocols can consult the DIA-based screening study of 120 food additives, the non-targeted screening review for food contact materials, and the polymer additive detection roadmap. For rapid ambient ionization technique, the practitioner review of DART and AP-MALDI methods covers throughput and sensitivity trade-offs in detail, and ASTM's D6042 standard for polyolefin additive separation offers a validated benchmark method.

• Comprehensive detection of 120 additives in food using nontargeted MS data acquisition
• Detection, identification, and quantification of polymer additives: a review of techniques, approaches, challenges, and a possible roadmap in analysis

FAQ

What are the main types of additives found in products?

Additives fall into functional classes such as colorants, preservatives, antioxidants, sweeteners, plasticizers, flame retardants, and stabilizers, depending on the product category. In food, the DIA screening study covering colorants, sweeteners, preservatives, and antioxidants illustrates the typical range of classes screened in a single non-targeted run.

What does chemical screening involve?

Chemical screening analyzes a sample to detect and identify compounds present, using either a targeted method that searches for a known list or a non-targeted method that captures broader spectral data for later interrogation. The choice depends on whether the goal is confirming known substances or discovering unexpected ones.

How is food additive testing typically performed?

Food additive testing commonly pairs extraction methods like QuEChERS with LC-HRMS or LC-MS/MS analysis, using non-targeted acquisition for broad surveys and targeted MRM for regulatory quantitation. Validated approaches have combined QuEChERS extraction with fast UHPLC-HRMS runs to screen multiple additive classes in a single fifteen-minute analysis.

What additive classes are most commonly screened in food products?

Colorants, sweeteners, preservatives, and antioxidants represent the additive classes most frequently included in non-targeted food screening panels, reflecting both regulatory attention and detection feasibility. The 120-additive DIA study breaks this down specifically into 79 colorants, 13 sweeteners, 12 preservatives, and 7 antioxidants within its screening panel, giving a practical sense of typical panel composition.

Recommended

• Coating Additives Guide: Selection, Chemistry & Formulation (2026)
• Formulators: 3 QC Metrics to Qualify a Coating Additive Producer
• Stop Scale Up Failures: Five Procurement Tests for Formulation Support
• How to Qualify Additive Suppliers Before the First Order