To localize LC-MS/MS carryover, first establish whether a clean blank acquires analyte-associated response after a high sample. Then compare an injected blank with a true no-injection run, test column involvement and isolate fluidic sections one at a time. A blank peak alone does not identify the contaminated component.
Mobile-phase contamination, vial contamination, interference and late elution can resemble carryover.1,2 In this guide, IS means internal standard and MS means mass spectrometry. It assumes you can retain the original chromatograms and use a separate diagnostic method without overwriting the production method. Confirm the instrument’s permissible connections, solvents and operating procedures before changing the flow path. The comparisons below are diagnostic reasoning, not a validated universal localization protocol; carryover is one of the causes of distorted response covered in the LC-MS ion suppression and matrix effects guide.
Is the blank peak actually carryover?
Carryover is residual sample material that contributes response to a subsequent run.1 Establish sequence dependence before assigning that label. Run a fresh blank before the challenge, the relevant high sample and fresh blanks afterward. Include a low-level control after the challenge when evaluating quantitative impact. Use high material within the method and instrument’s suitable loading conditions; an extreme overload can create a problem that is not representative of routine use.
Check the expected retention region, quantifier and qualifier behavior where available, integration and the unprocessed trace. A matching transition alone is not definitive analyte identification. A blank that is already positive before the high sample requires investigation of baseline contamination or interference as well as residual material from earlier sequences.
Use a newly prepared blank in a separate vial or well. Repeatedly sampling one contaminated blank confounds the challenge. Match relevant blank solvent and preparation conditions and document whether the blank contains IS. An analyte-free blank without IS can help reveal response associated with IS addition, but it does not replace all method-required blanks.
Table 1 defines the diagnostic controls so that their names are not mistaken for interchangeable operations.
| Control | What it tests | Important limit |
|---|---|---|
| Fresh solvent blank injection | Diluent, vial and normal injection-path exposure | Does not separate the injector from downstream components |
| Processed blank matrix | Preparation-related background in representative matrix | Endogenous analyte or interference may make it positive |
| Zero-volume injection | The instrument’s programmed zero-volume operation | Valve switching and needle contact may still occur |
| True no-injection gradient | Background with no sample aspiration or injection cycle | Flow may still traverse injector surfaces |
| Clean independent source feed | Background after removing the LC path from the comparison | Feed solvent, syringe, line and probe must also be considered |
LC-MS PK Peak Integrity Engine
Interrogate carryover, matrix effects, ion suppression, and integration integrity before your PK data goes to QC.
What is the localization workflow?
Figure 1 uses three phases. Its middle phase contains independent comparisons selected by the observed pattern, so no arrows connect those tests.
Confirm sequence dependence: establish clean controls
- Run a fresh blank before the high sample.
- Run the high sample and fresh blanks afterward.
Compare fluidic paths: independent tests selected by the pattern
- Compare an injected blank with a true no-injection run.
- Compare the suspect column with a clean compatible path.
- Isolate tubing and the divert valve one section at a time.
Check and verify: resolve the supported contribution
- Check background with a clean independent source feed.
- Correct the supported cause and repeat the challenge.
These comparisons localize contributions rather than proving a unique component. A no-injection run does not necessarily bypass injector surfaces. A column change alters retention and matrix delivery. A clean source-feed comparison requires clean solvent, syringe, line and probe. Preserve acquisition coverage when changing the flow path, and use instrument-approved connections and cleaning procedures.
What does an injected blank versus a no-injection run tell me?
Use the same gradient and acquisition coverage while changing whether the injection cycle occurs. Verify from the instrument documentation or event log that the selected run genuinely omits the injection operation. Waters documents a MassLynx configuration for acquiring a gradient blank without operating the injector; that configuration applies to the named platforms, not every autosampler.3
If the injected blank is positive and the no-injection run is clean, sample introduction becomes a stronger candidate. Check the fresh diluent, vial and closure before replacing autosampler parts. A clean no-injection run does not by itself establish contamination specifically on the needle: the difference includes sample aspiration, valve operation and contact with the container.
If both runs are positive, residual material may be eluting from the column or another wetted component, or contamination may arrive continuously with the mobile phase. On some instruments, solvent still passes through injector surfaces even without an injection cycle. Trace the actual flow path before saying the autosampler has been excluded.
Compare paths under a repeatable exposure history. If you simply run several blanks and then change the test, natural washout can make the changed condition look corrective. Re-establish the challenge where appropriate and document the run order.
Figure 2 compares conceptual faster and slower declines across fresh blanks. High-sample area is normalized to 1 and first-blank area is 0.1; each later blank is a fixed fraction q of the one before:
Bn = B1 × qn−1
q = 0.2: 0.1, 0.1 × 0.2 = 0.02, 0.1 × 0.2² = 0.004 q = 0.8: 0.1, 0.1 × 0.8 = 0.08, 0.1 × 0.8² = 0.064
Every peak has the same Gaussian width and position. These choices illustrate contrasting shapes, not measured carryover or an acceptance threshold. A clean pre-challenge blank is assumed. Neither shape identifies a contaminated component, and natural washout can confound a later intervention.
How do I test column involvement without drawing a false conclusion?
A compatible clean replacement column preserves separation better than a union, although differences in chemistry, dimensions or condition still matter. A column-bypass test is useful when permitted by the system, but analyte may reach the MS near the void region or as a broad signal. Widen acquisition and review the entire relevant interval; disappearance at the original retention time is not evidence of disappearance from the system.
Use a diagnostic flow path that preserves permitted source flow and operating conditions. Removing a column changes pressure and can change pump delivery, mixing and ionization conditions, so do not compare raw areas as though those conditions were unchanged. Avoid sending a concentrated sample plug or incompatible cleaning solvent into the MS. Stop flow and release pressure before reconnecting components under the manufacturer’s procedure.
For a gradient method, a double-gradient test can probe whether material remains or accumulates for a later gradient. Waters and SCIEX describe related tests, but their interpretation must include mobile-phase contamination as well as column retention.2,4 Repeated peaks are not unique proof that the column is contaminated. Check clean mobile phase and containers and retain the method’s required equilibration between gradients.
Can tubing, fittings or the divert valve cause apparent source carryover?
Yes. Poorly seated fittings can create spaces that retain sample.1 Residual material can also remain in valves or lines. Waters reports an MS carryover case in which bypassing the divert valve produced clean blank chromatograms; the contamination was in that valve.5
Map the connections before the experiment. Where the instrument permits, compare one section with a clean compatible replacement or approved bypass. Keep the other sections and exposure conditions as comparable as possible. Repeat the challenge to confirm improvement instead of attributing ordinary washout to the replacement.
If the response changes, the altered section is implicated, but altered flow, dead volume and solvent history still need consideration. Do not dismantle a pressurized line or energized source, and do not transfer a model-specific valve-rebuild procedure to another instrument.
Table 2 links useful evidence to the next controlled action.
| Evidence | Contribution to investigate | Controlled next action |
|---|---|---|
| Fresh blank positive before challenge | Existing contamination or interference | Compare fresh solvent, container and preparation controls |
| Injected blank positive; true no-injection run clean | Introduction-path or blank-associated contribution | Verify blank cleanliness and wash-path operation |
| Signal changes with clean column comparison | Column-associated contribution or altered delivery | Repeat with acquisition covering the changed elution interval |
| Signal changes after one tubing or valve comparison | Altered section or fluidic conditions | Repeat the challenge with controlled flow and solvent history |
| Background persists with clean independent source feed | Remaining source-side or feed-path contamination | Confirm feed cleanliness; follow the manufacturer’s diagnostic procedure |
| Background falls after removing the LC feed | LC-associated contribution | Restore sections systematically to localize the contribution |
When should I suspect the source?
Suspect a source-side contribution after the upstream comparisons support it, or when a manufacturer-approved check indicates it. A clean independent solvent feed removes the LC contribution from that comparison. If background falls, the LC path is implicated. If it persists, examine the remaining feed path, probe and source-associated components.6
This check compares background under its stated conditions, not chromatographic peak area at the original retention time. The solvent and source conditions influence ionization; a negative result does not exclude every species or operating condition. Confirm the independent feed is clean before interpreting persistence as source contamination. Internal instrument maintenance beyond authorized user procedures belongs with trained service personnel.
How do I select a wash or cleaning correction?
For the specific introduction-path variables, see autosampler carryover and needle-wash optimization. Use the localization evidence here to decide which fluidic contribution to address.
Match the correction to the supported location. For an introduction-path contribution, verify that the intended wash reaches the relevant surface and that solvent actually flows: changing wash strength will not repair an obstructed wash line. Waters’ location-specific guidance covers wash time, needle-wash choice, obstructed washes and affected injector components for its named systems.7
Choose solvents that dissolve the residual material and are compatible with seals, metals, tubing, column and MS. There is no universally strongest or safest wash. Prevent precipitation when changing between buffered aqueous solvent and strong organic solvent. Divert cleaning waste as required, restore the production mobile phase and equilibrate before checking performance.
Avoid changing the needle wash, column wash and source maintenance together. A combined intervention can reduce carryover while leaving its origin unresolved.
How do I verify carryover is controlled?
Repeat the relevant high-sample/blank challenge after the correction and evaluate a subsequent low-level control. Record blank response, sequence position, chromatography and quantitative impact. Include the matrix and operating conditions that caused the original failure. Apply the assay’s established acceptance criteria to analyte and IS where applicable.
A carryover fraction relative to the high sample is not interchangeable with response relative to the assay’s low end, as Figure 2 (panel C) shows:
Carryover fraction (%) = blank area / preceding high-sample area × 100 in the Figure 2 model, 0.1 / 1 × 100 = 10%
Response relative to the low end (%) = blank response / low-level response × 100
State the denominator whenever reporting a percentage. A manufacturer specification is not automatically the acceptance limit for your assay. Do not subtract a preceding blank response from later samples as an unvalidated correction.
If carryover remains unexplained or quickly recurs, retain the challenge chromatograms, method, plumbing map, component history and wash conditions for technical support. If the main problem is declining response rather than residual peaks, use LC-MS sensitivity-loss troubleshooting.
Figure 3 summarizes correction and verification:
- Verify wash delivery to the relevant surface.
- Choose a compatible correction for the supported location and change one factor.
- Restore production mobile phase and equilibrate.
- Repeat the high-sample/fresh-blank challenge and low-level control.
- Apply established assay criteria to blank response and quantitative impact.
- Document the correction or continue localization if carryover persists.
Frequently asked questions
Does a peak at the analyte retention time prove column carryover?
No. Residual material introduced upstream, in the needle, injection port, valve or connecting tubing, can be retained on the column and eluted at the analyte’s retention time, so the retention time says where the material eluted, not where it was held. Mobile-phase contamination that accumulates on the column during equilibration can also appear there. Use the injected-versus-no-injection comparison, a clean compatible column and, for gradient methods, a double-gradient run to separate those contributions before replacing the column.
Is zero injection volume the same as no injection?
Not necessarily. A zero-volume injection runs the instrument’s programmed injection operation, which may still switch the valve, move the needle and contact the sample container. A true no-injection run omits the injection cycle entirely, and even then mobile phase may still pass through injector surfaces on some instruments. Confirm valve switching, needle movement and aspiration behavior on the actual instrument from its documentation or event log; the label in the sequence table is not a plumbing diagram.
Does disappearing signal after column bypass clear the autosampler?
No. Bypassing the column also removes retention, so analyte that was held and eluted as a peak may now reach the MS near the void region or as a broad, low signal outside the original acquisition window. The bypass also changes pressure, pump delivery, mixing and ionization conditions. Widen acquisition to cover the whole relevant interval, control the fluidic differences and compare like with like before concluding that the autosampler is not contributing.
Should I add blank injections to solve the problem?
Extra blanks can reduce exposure to residual material, but they do not establish the cause or guarantee low-level accuracy, and they can hide a contribution that will return when the sequence changes. Natural washout across several blanks can also make a later intervention look corrective. Localize the contribution, correct one supported factor, then verify the complete high-sample/blank challenge with a low-level control. Document any sequence-management rule under the laboratory’s procedure.
Can a contaminated blank vial look like carryover?
Yes. If a blank vial, closure or diluent is contaminated, every injection from it can show analyte-associated response, and repeatedly sampling the same vial after a high sample confounds the challenge. Use a newly prepared blank in a separate vial or well for each position, and compare fresh solvent, container and preparation controls when a blank is positive before the challenge. Waters suggests an injection-volume study: blank peaks that grow with injection volume point toward the vials.2
The takeaway
A blank peak after a high sample is the start of a carryover investigation, not its conclusion. Establish sequence dependence with fresh blanks, then use independent comparisons (injected versus no-injection runs, a clean compatible column, one tubing or valve section at a time and a clean independent source feed) to localize the contribution. Correct the supported location with a compatible wash or cleaning step, change one factor at a time, then repeat the challenge with a low-level control and report percentages with their denominators.
References
- Waters, “Reducing carryover,” Alliance iS HPLC Systems User Guide, topic LCI-USG-0094 (last updated 12 May 2026): carryover definition and tubing, fitting and vial-seal guidance.
- Waters, “What are some tips for troubleshooting carryover or ghost peaks on my LC column?,” WKB246120: double-gradient and injection-volume tests.
- Waters, “How to perform the gradient blank injection (for MassLynx)?,” WKB81913: procedure and named-system applicability.
- SCIEX, “Reference and troubleshooting guide,” page 2, carryover branch (accessed October 2026).
- Waters, “Carryover on MS instrument,” WKB126032: cause and fix sections.
- Waters, Controlling Contamination in LC/MS Systems: Best Practices, 715001307 version 09, pages 11–12: isolating the problem to the LC or MS system.
- Waters, “How to troubleshoot LC carryover (injector carryover) based on its location,” WKB64313: procedure and named-system applicability.
Further reading
- LabVeda, Autosampler carryover in HPLC and LC-MS: how to find it and wash it out.
- LabVeda, Sensitivity loss during an LC-MS/MS batch.
- LabVeda, Ion suppression and matrix effects in LC-MS/MS – the cluster pillar.
Reviewed against manufacturer sources. Every manufacturer statement on this page is checked against the cited source; manufacturer instructions remain platform-specific, and control selection and confounder analysis are LabVeda’s diagnostic synthesis. Numerical examples are illustrative calculations from the equations stated and are not acceptance criteria. For validated or compendial methods, the applicable procedure and regulatory framework take precedence over the general rules of thumb given here. Evidence review: October 2026.