Trace-organic screening in water reuse asks which low-concentration molecules or signals are present and how confidently they can be identified. Nontarget discovery is useful, but a detected mass-spectrometry feature is not automatically a confirmed chemical or proof of risk.
The gold standard separates targeted quantitation, suspect screening, unknown-feature discovery, reference confirmation and complete water-quality verification, preserving both sample provenance and analytical uncertainty.
eduKateSG · Water Reuse Trace Organic Contaminant Screening
Find your next learning step
Move from sampling to analytical features, worked measurements and reliable identification, then assess the scope of any water-reuse conclusion.
Full chapter index · Practice and transfer · Education root · How X Works
Full chapter index
Define the chemical question · Chapters 1–2
Calculate analytical evidence · Chapters 3–4
Evaluate unknown features · Chapters 5–7
Validate reuse conclusions · Chapters 8–10
Continue connected learning · Chapters 11–12
CHAPTER 1 OF 12 · Define the chemical question
1. Water reuse needs chemical evidence beyond one aggregate number
Water reuse produces useful water through a chain of treatment processes whose effectiveness depends on the original water, membrane barriers, oxidation and operational control. Organic trace contaminants may occur at concentrations far below those of the ordinary dissolved organic background.
Total organic carbon (TOC) is an aggregate carbon measurement rather than a list of chemical identities. A small amount of a compound relevant to a risk assessment could contribute very little to TOC, even when the TOC measurement is accurate. Conversely a relatively high TOC value need not indicate a specific hazardous compound.
Trace-organic screening helps investigate which compounds are present, whether they change across treatment and which results warrant confirmation or further assessment. The gold standard separates measured detector features, plausible identities, confirmed chemical identities and quantified concentrations.
Singapore PUB describes NEWater as a high-grade reclaimed-water supply with a broad monitoring programme across physical, organic, inorganic and microbiological parameters. That documented context shows why research screening complements rather than replaces validated public-water quality assurance.
Do not treat a newly discovered analytical feature as proof of danger or as evidence of regulatory noncompliance. A reliable claim needs validated identification, quantitative exposure information when relevant and the applicable scientific or regulatory interpretation.
CHAPTER 2 OF 12 · Define the chemical question
2. Targeted, suspect and nontarget analyses answer different questions
Targeted chemical analysis begins with a defined list of compounds and established analytical characteristics. It can use calibrated standards, internal controls and specific reporting limits to determine whether those target analytes were detected and, where validated, at what concentrations.
Suspect screening starts with a list of plausible compounds but may not have a reference standard for every candidate. High-resolution mass spectrometry (HRMS) can provide accurate mass, isotopic patterns, chromatographic retention and fragmentation clues that help evaluate candidate chemical identities.
Nontarget analysis attempts to detect a wider range of features without beginning with a complete predetermined list. It often produces many mass-to-charge and retention-time combinations, only a fraction of which can be assigned confidently to specific compounds.
NIST’s Method Assessment for Non-Targeted Analyses programme describes the importance of quality-control materials, benchmarking and interlaboratory comparability because an unknown-signal discovery workflow is more uncertain than a well-established targeted test.
An identified molecular formula is not necessarily a uniquely identified structure. Isomers can share the same formula and accurate mass, making independent standards or further fragmentation, retention and other evidence important.
Report the method’s chemical coverage and omissions. One LC-HRMS method cannot reliably claim to measure every volatile, nonionizing, highly polar or insoluble compound in a water sample.
CHAPTER 3 OF 12 · Calculate analytical evidence
3. Sampling, blanks and chain of custody come before identification
Field sampling can introduce trace contaminants through bottles, tubing, personal-care products, handling equipment or transport. A sophisticated mass spectrometer cannot distinguish a field-introduced compound from the original water without suitable blank and sampling controls.
Field blanks, laboratory procedural blanks, extraction blanks, calibration checks and selected matrix spikes each test a different possible error source. Their role should be described before interpreting a newly detected peak.
Choose an appropriate preservative and container under the specific validated analytical method. Chemical changes during storage may create transformation products or cause disappearance of unstable compounds, so time and handling conditions matter.
Samples taken before and after a membrane or oxidation process must represent comparable water parcels or appropriately integrated operating periods. A large transit-time mismatch can make apparent removal or formation misleading.
Record water source, exact collection point, time, flow, treatment state, sample method, container, storage, laboratory method and reporting limit. A spectacular discovery claim without these details is difficult to reproduce.
The EPA non-targeted analysis research programme emphasizes that broader chemical coverage also introduces greater uncertainty. Source traceability is one way to keep that uncertainty visible.
CHAPTER 4 OF 12 · Calculate analytical evidence
4. Worked example: dilution and a quantified targeted result
Imagine a fictional validated targeted assay for a single organic compound. A sample is diluted by mixing 1.0 mL of water sample with 4.0 mL of suitable solvent, creating 5.0 mL total mixture and a dilution factor of five.
Suppose the correctly calibrated instrument reports 3.0 ng/mL in the diluted extract under the stated method. If sample preparation introduced no additional recovery or volume changes and the calibration applies, the original water concentration is 3.0 × 5 = 15 ng/mL.
Because 1 ng/mL equals 1 µg/L in a water concentration context, the hypothetical result is 15 µg/L. That unit conversion requires volume-based equivalence; it does not identify or evaluate the toxicity of the compound.
Do not apply a dilution factor twice when the laboratory software already reports original-sample concentration. The method record should state which concentration basis is used.
The observed analytical concentration is different from the real mass emitted by a source. To estimate a mass load, the flow and sample timing must be appropriate for the same water stream.
All numbers here are teaching data, not public-water measurements, chemical safety limits or results of a named laboratory.
CHAPTER 5 OF 12 · Evaluate unknown features
5. Worked example: spike recovery and quantitative bias
Suppose a hypothetical sample is spiked with a known additional amount corresponding to 10 ng/L of a reference analyte. After subtracting the unspiked sample response, the procedure recovers a measured equivalent of 8 ng/L.
Apparent spike recovery is 8/10 × 100 = 80%. The result describes one sample and method; it does not automatically show all other compounds have the same extraction or matrix response.
If a separate validated method explicitly supports recovery correction for the compound and sample, an observed uncorrected concentration of 4 ng/L under a constant 80% recovery assumption would correspond mathematically to 4/0.8 = 5 ng/L. Such correction is not always appropriate or permitted.
Actual quantitation may use isotope-labelled internal standards, matrix-matched calibration and method-specific correction approaches. An instrument report may already include those adjustments.
Record the spiking point, reference material, blank, recovery uncertainty and analytical batches. A single 80% recovery does not certify repeatability or invalidate every observation in a study.
NIST reference material work emphasizes that well-characterized controls are important for comparability of trace-organic measurements.
CHAPTER 6 OF 12 · Evaluate unknown features
6. Worked example: accurate mass does not uniquely name a molecule
High-resolution mass spectrometers often measure mass-to-charge ratios very precisely. Mass error can be expressed in parts per million as ppm = (measured m/z − theoretical m/z)/(theoretical m/z) × 1,000,000.
Take a fictional candidate ion with calculated m/z 200.0000 and measured m/z 200.0010. Difference is 0.0010, giving 0.0010/200 × 1,000,000 = 5 ppm.
Five parts per million can be useful for filtering candidate elemental compositions under a stated instrument and calibration method, but it cannot distinguish every structural isomer or prove the proposed chemical name.
An observed ion may be an adduct, isotope, in-source fragment, contaminant or multiple charged species. Incorrect assumptions about charge or adducts can create an apparently compelling but false molecular mass match.
Fragmentation spectra, retention behaviour, sample blanks and available reference spectra help discriminate possibilities. NIST’s DIMSpec programme emphasizes preserving method and spectral metadata because even known molecules can produce differing fragmentation patterns under different instruments.
The mass calculation supports candidate prioritization, not a validated concentration, toxicity determination or confirmed molecular identity.
CHAPTER 7 OF 12 · Evaluate unknown features
7. A feature is not automatically a unique compound
A feature in a chromatographic HRMS dataset may combine an m/z value, retention time, intensity and other attributes. One molecule can produce several features through isotopes, adducts, fragments or ionization states.
Conversely two different molecules can co-elute or have very similar accurate masses, especially in complex water matrices. Data processing may merge, split or misalign features when signal quality is poor.
Nontarget screening therefore benefits from blank filtering, isotope and adduct grouping, replicate agreement, retention-time alignment, peak-shape checks and sensible candidate grouping.
An increase from 500 to 700 detected software features does not automatically mean 200 new chemicals entered a water supply. Algorithm settings, dilution, contamination and sensitivity can change the number of reported peaks.
A peak intensity in arbitrary counts is not generally interchangeable with a concentration in ng/L across different compounds. Ionization efficiency and matrix suppression differ markedly.
EPA explains that non-targeted analysis often provides qualitative evidence, and quantification can require further analytical work. The final report must distinguish feature detection from molecular identity and concentration.
CHAPTER 8 OF 12 · Validate reuse conclusions
8. Identification confidence has a hierarchy
A molecular identity backed by an authentic reference standard measured using a compatible method can be reported with higher confidence than a name proposed solely from an exact-mass match.
Suspect screening may add reference-library fragmentation evidence, predicted chemistry and chromatographic behaviour. Even a strong library match may leave isomeric alternatives if the analytical evidence is not discriminating.
A NIST study on PFAS identification proposed a structured confidence scale, with highest levels involving reference-standard evidence and lower levels for more tentative suspect matches or exact-mass features. That framework is specific to its chemical domain and should not be generalized without qualification.
Record the candidate list and version, accepted formula, adduct assignment, fragment match evidence, retention data and whether an authentic standard was actually analysed.
Do not convert a confidence score into a public safety probability without a separate validated probability model. A ‘level 2’ chemical identification means something different from a ‘2% risk’.
Where a finding has potential health or regulatory consequences, independent targeted confirmation and competent interpretation are especially important.
CHAPTER 9 OF 12 · Validate reuse conclusions
9. Nondetects, detection limits and false positives
A laboratory nondetect means the result did not satisfy the method’s identification or reporting criteria. It does not demonstrate that the compound’s true concentration is exactly zero or that the method had sensitivity to every chemical.
Suppose a fictional validated reporting limit is 5 ng/L for one known analyte. A result reported below that limit cannot be replaced by 0 ng/L in a mass-balance calculation without an explicit statistical assumption.
If one investigation adopts an illustrative censored-data interval of 0 to below 5 ng/L, the value may be represented as an upper-bound constraint rather than a measured concentration. Real lower or upper bounds depend on method and blank evidence.
Conversely a candidate feature can be a false positive because of laboratory contamination, background peaks or incorrect molecular assignment. A wide feature list needs criteria for prioritizing compounds and controlling spurious identities.
EPA and NIST both describe method benchmarking and quality-control work as important because expanded screening capabilities carry uncertainty. The screening output should be ranked by the strength of evidence, not by the dramatic appearance of a compound name.
Comparison before and after treatment needs compatible reporting limits. A later nondetect from a less sensitive method is not necessarily proof of removal.
CHAPTER 10 OF 12 · Validate reuse conclusions
10. Chemical detection and risk assessment are different decisions
A detected compound must be interpreted in terms of confirmed identity, concentration, exposure pathway, hazard information and relevant local requirements. Mere detection does not establish a health risk.
Similarly, the disappearance of one parent molecule through advanced oxidation does not prove complete destruction of its carbon or that all transformation products are innocuous. The treatment and analytical model should be explicit.
PUB reports a comprehensive NEWater quality programme with hundreds of monitored parameters across the supply. Exploratory screening can inform prioritization within such a broader context but should not be misrepresented as the whole quality decision.
Where reference toxicological information is limited, further research may be needed before making a meaningful risk comparison. The absence of an existing standard does not justify inventing one.
NIST and EPA research on nontarget methods highlights the value of discovering previously overlooked chemicals and the challenges of identification and quantitative exposure assessment.
A responsible communication distinguishes ‘feature observed’, ‘candidate substance’, ‘confirmed substance’, ‘concentration measured’ and ‘risk assessed’. Each step requires new evidence.
CHAPTER 11 OF 12 · Continue connected learning
11. Practice, FAQs and changed-water transfer
Practice A: A fivefold-diluted sample measures 3 ng/mL in an appropriate targeted method. Convert to original concentration 15 ng/mL and state the assumptions about sample preparation.
Practice B: A 10-ng/L spike returns an 8-ng/L increment. Calculate 80% recovery, and explain why a recovery correction is method dependent.
Practice C: For theoretical m/z 200.0000 and measured 200.0010, calculate 5 ppm mass error. Why can that not prove the exact molecule?
FAQ — Does nontarget analysis quantify everything detected? No. Many findings are qualitative features or tentative identities.
FAQ — Does nondetect mean no compound exists? No. Method sensitivity, analytical coverage and reporting criteria matter.
FAQ — Does a new trace-organic detection prove unsafe reclaimed water? Not by itself. Verified identity, exposure, hazard and regulation are separate questions.
Transfer the same screening workflow from treated membrane permeate to complex untreated used water. Reassess matrix effects, detection limits, sample preparation, feature prioritization and independent confirmation.
Verified eduKateSG reading routes for the distinct analytical and treatment questions:
- The Gold Standard of Water Reuse Total Organic Carbon Monitoring — aggregate organic carbon versus molecular identity
- The Gold Standard of Water Reuse Advanced Oxidation Validation — chemical transformations and emerging products
- The Gold Standard of Water Desalination Membrane Performance — physical barriers and membrane rejection
- The Gold Standard of Water Quality Sensor Calibration — measurement standards and QA
- The Gold Standard of Water Treatment Process Control — risk-based operating evidence
- The Gold Standard of Spectroscopy — molecular measurement and spectral inference
- The Gold Standard of Statistics — uncertainty and multiple comparisons
CHAPTER 12 OF 12 · Continue connected learning
12. Connected water-treatment routes and primary references
The trace-organic screening topic connects water reuse, analytical chemistry, total organic carbon monitoring, membrane performance and advanced oxidation. These are different questions in a common evidence chain.
An audit-ready workflow preserves sample chain of custody, instrument data, reference materials, blank screening, feature list, tentative assignments, confidence tiers, quantitative confirmation and appropriately scoped interpretation.
All concentrations, recoveries and accurate masses above are fictional learning examples. This article does not assess a particular water supply, certify any chemical’s safety or replace qualified laboratory and regulatory procedures.
Primary official references for screening, analytical confidence and water quality:
- US EPA — Non-Targeted Analysis Research — scope and uncertainty of chemical discovery
- US EPA — Drinking Water Filter NTA Study — qualitative features and constraints
- NIST — Method Assessment for Non-Targeted Analyses — quality controls and comparisons
- NIST — DIMSpec Mass Spectrometry Data Infrastructure — fragmentation spectral provenance
- NIST — Identification Confidence With HRMS — confidence levels and standards
- PUB Singapore — NEWater Quality — whole-supply analytical monitoring
- NIST — Standard Reference Materials for Trace Organics (2025) — reference materials and traceability
The immutable Clementi Secondary 1 Mathematics teaching reference illustrates instructional sequencing, not chemical water testing authority.