Instrument Intelligence Fundamentals
For chromatography and mass spectrometry
Workflow intelligence in LC/MS systems
The latest generation of LC/TQ and LC/(Q)-TOF instruments offers a suite of intelligent automation features to reduce the burden of manual intervention and keep the instruments at peak performance. Software predictive tools not only help to reduce the time users need for troubleshooting, but also help prevent costly downtime, and unexpected maintenance issues, and prolong the lifespan of the equipment. New workflow intelligence capabilities, such as intelligent reflex workflows are a feature of MassHunter software and are designed to save time while also reducing the burden on analysts. These workflows aim to reduce manual intervention, increase lab productivity, and avoid sample reruns.
Agilent’s Emma Rennie talks about the promise of instrument intelligence.
Intelligent reflex workflows in LC/Q-TOF and LC/TQ
Intelligent reflex is a smart algorithm that automatically reinjects samples based on answers provided by quantitative mass spectrometry data analysis software. This algorithm enhances data quality in real time and dramatically increases throughput through several logic-based workflows. These sample reinjection protocols boost efficiency in high-throughput and routine laboratory scenarios for LC/TQ and LC/(Q)-TOF users. Users can also improve identification confidence with two LC/Q-TOF-only intelligent reflex workflows. All workflows assume that there is a normal, routine worklist structure with QCs, blanks, and samples. The workflows use MassHunter Quantitative Analysis methods, pre-established calibration curves, and outlier thresholds combined with hardcoded rules to prevent carryover and allow users to follow their laboratory standard operating procedures. Reinjections added as a result of intelligent reflex workflows are visible at a glance.
Five intelligent reflex workflows are available. Check out the workflows below or in this technical note.
Carryover detection workflows
The intelligent reflex carryover detection workflow prevents carryover from contaminating multiple samples in a worklist. If carryover is detected, the system will inject up to a user-defined maximum number of blanks to attempt to overcome the contamination. The user can choose to pause the worklist if carryover is still detected after the maximum number of blank reinjections have been acquired. When carryover is no longer detected or not present, the worklist continues as scheduled.
Shown in the screenshot is the insertion of blanks when carryover is detected during ongoing analysis.
Above calibration range workflows
If a target analyte is detected above a predetermined calibration curve range, the intelligent reflex above calibration range workflow can be used to estimate a calculated concentration. When an analyte exceeds the calibration range, first a blank is injected to ensure that there is no carryover. Then, the sample is reinjected with a reduced injection volume to bring it back within the range of the calibration curve.
The screenshot shows how the software is appending a reinjection with lower injection volume due to original measurement reporting above the upper limit of quantitation.
Fast screening workflows
Intelligent reflex fast screening workflows are used to confirm a presumptive positive. These workflows boost productivity and save time for worklists where there is a low expected number of positive samples. For example, in forensic toxicology samples, there is likely an expectation that most samples will be clean, with a low percentage of samples returning positive results. Fast screening workflows typically use two different acquisition and quantitative methods. In this scenario, the first-tier method is fast, often a ballistic gradient. This fast method is used to rapidly screen all samples. If a run positively detects a target analyte, that sample is re-analyzed using a longer method for true confirmation. Learn more about applying the fast screening workflow.
The screenshot shows how the worklist automatically inserts a confirmation method after detection of a presumptive positive.
Targeted MS/MS confirmation workflows
The targeted MS/MS confirmation workflow is available for LC/Q-TOF instruments and confirms compounds that were screened as questionable or present during untargeted screening. Two different acquisition and quantitative methods are used for this intelligent reflex workflow. The first-tier method uses All Ions (DIA) data, while the second-tier method uses MS/MS data for confirmation.
The screenshot shows how the targeted reinjection was appended to the running worklist along with a QC and blank injection. The originally created worklist was completed, followed immediately by the targeted reinjection without manual input by the scientist. Learn more about this workflow used in pesticide screening in food matrices.
Iterative MS/MS workflows
Intelligent reflex iterative MS/MS workflows offer thorough characterization of a sample or pooled QC by LC/Q-TOF. Users can boost identification confidence by iteratively excluding higher abundant precursors. For example, the first analysis detects precursors A, B, and C. The second analysis uses the same method but will exclude A, B, and C. If the second analysis detects precursors D and E, then the third analysis will exclude A, B, C, D, and E—and so on, until the user-defined number of iterations is run. The user can start with matrix blank iterations to ensure that high-abundant matrix precursors are not selected during sample runs.
Read more about intelligent reflex workflows.
Look at different types of intelligent reflex workflows
Start and end of the workflow
Method development and data analysis bookend any analytical workflow.
Method optimization
MassHunter Optimizer software is an intelligent software algorithm for end-to-end LC/MS/MS method development, optimization and QA/QC deployment. The Optimizer tool is part of MassHunter 12 and allows users to optimize multiple reaction monitoring (MRMs) transitions and ion source parameters in a fully-automated or semi-automated fashion. Using a modular end-to-end workflow approach, users may input chemical formula information which will result in optimized MRM transitions for each compound and optimal ion source parameters for the overall method.
The intelligent optimizer can be used to create a new method from scratch, add new compounds to an existing method, or fine-tune or verify parameters of an existing method. Optimization results can be reviewed in a report format by the user after the workflow is completed. Any changes to the method are saved in an auditable fashion in accordance with 21 CFR Part 11 compliance.
| Compound Parameters | Source Parameters |
|---|---|
| Collision energies | Capillary and nozzle voltages |
| Fragmentor voltages | Nebulizer pressure |
| Identification of precursor and product ions | Gas temperatures Gas flows |
Learn more about MassHunter Method Optimizer software in this step-by-step video and the 6475 ebook.
As an example, the image shows a fully optimized seven-analyte standard mixture chromatogram created from scratch. Chemical formulas of the neutral analytes were entered into the method development interface to automatically calculate the potential [M+H]+ or [M-H]- precursor ions. The optimization workflow was performed unattended in two major phases. First, MRM optimization included precursor fragmentor voltage, RT determination (optional) product ion selection, and MRM collision energy voltage. Then, ion source optimization included drying gas temperature, sheath gas temperature, capillary voltage, nebulizer pressure, drying gas flow, sheath gas flow and nozzle voltage.
Read this poster about MassHunter Optimizer to learn more about the end-to-end workflow for LC/MS/MS method development, optimization, and QA/QC deployment.
Book a training course that covers MassHunter Optimizer and source optimizer tools at Agilent University.
AI tools for peak integration
AI Peak Integration for MassHunter software automates manual peak detection and integration during the data analysis process. A machine learning model is custom trained during a user’s normal data analysis workflow by observing manual integration events. It then replaces manual peak integration with adaptive AI-assisted peak detection and integration.
Available for AI Peak Integration for GC/MS Analysis of Phthalates.