Unraveling the Case: Reasons Behind Poor LC RSD
Table of contents
Introduction
Liquid chromatography (LC) is an indispensable tool in analytical laboratories, offering high precision in quantitative analysis. However, when faced with unexpected variations in peak area reproducibility, troubleshooting can be a complex process.
This article explores a real-world case involving poor relative standard deviation (RSD) in an HPLC system and provides a systematic approach to diagnosing and resolving such issues.
The Case: A Troublesome RSD
Dr. Johnson, an analytical chemist from Welch Materials’ technical support team, received a support ticket from a food testing laboratory. When quantifying sugar compounds using Welch’s HPLC system equipped with a refractive index detector (RID), although retention times remained stable, the laboratory observed a peak area RSD exceeding the method’s acceptable limit of 0.3%, reaching an unsatisfactory 0.8%.
| Chromatographic Conditions | |
|---|---|
| Detector | RID |
| Mobile Phase | 100% pure water |
| Flow Rate | 1 mL/min |
| Column | Sugar column |
Troubleshooting the Issue
Understanding that peak area variability can stem from multiple factors, Dr. Johnson methodically examined potential sources of error.
1. System Leakage Check
The first step was to ensure the integrity of the LC system. Dr. Johnson inspected all tubing connections and observed the system under running conditions to detect any leaks. Since no liquid seepage was found and the pump flow rate was stable, system leakage was ruled out as the cause.
2. Assessing the Autosampler
Given that injection precision is a common culprit in peak area variability, the autosampler was the next component under scrutiny. The following key checks were performed:
-
Presence of air bubbles in the syringe: No bubbles were observed, eliminating injection volume inconsistency as a possible factor.
- Wash cycle effectiveness: During the wash cycle, liquid was seen leaking from the injection port—an abnormality suggesting improper sealing at the injection site.
-
Injection depth alignment: A software-based depth check confirmed that the needle was penetrating to the correct position.
- Injection port sealing: Upon closer examination, the septum was found to be worn and was promptly replaced.
Following these adjustments, a repeat analysis showed a slight improvement in RSD, reducing it to 0.7824%, but still falling short of the required threshold.
A more thorough inspection revealed that small fragments of the vial septum had accumulated at the injection port, interfering with sample uptake. These obstructions were also removed.
Hidden Culprit: Incorrect Equilibration
While the worn septum of the injection port and the fragments of the vial septum all contribute to the increase of RSD, the actual culprit of the case is insufficient equilibration of the column. After the system was allowed to equilibrate for 12 hours as per the column manual, a final test run demonstrated a significant improvement—RSD was now a satisfactory 0.2047%.
Lessons Learned: Why Peak Area RSD Can Be High
Peak area reproducibility issues can arise from various sources. Here are the primary factors to consider:
1. Instrument-Related Issues
- Sampler Malfunctions:
The sampler is always the first to check. Typically, bubbles formed during sample volume measurement are a primary source of error. For different types of samplers, issues and solutions vary.
If using a manual sampler, ensure no air bubbles in the syringe; for fixed-loop samplers, make sure no air enters the flow path. One specific issue to note is when the septum surrounding the sample vial's injection port is sealed, creating a negative pressure during high-volume sample aspiration. This can draw air bubbles into the flow path, resulting in inconsistent sample volumes between injections. Using pre-pierced septa can mitigate this problem.
For autosamplers, here are some common issues and their solutions:
| Issue | Solution |
|---|---|
| Rotor or needle seat seal wear | Replace the rotor or needle seat seals. |
| Bent or clogged injection needle | Ultrasonically clean the needle (ensure the needle tip does not touch the vial bottom); replace the needle if cleaning is ineffective. |
| Injection needle not in "Inject" position before sampling | Set the needle to "Inject" position before starting sample injection. |
| Bubbles in the injector syringe | Perform the syringe priming command before injections to remove any air bubbles. Manual degassing may be necessary. |
| Injection speed too fast | Reduce the aspiration speed to at least 3-5 seconds for proper sampling. |
| Leaky syringe or worn injector valve | Replace the syringe or injector valve. |
- Detector Issues:
Poor grounding of the detector can lead to fluctuations in the peak area. The degree of integration depends on the signal-to-noise ratio. Flow rate fluctuations leading to peak area changes are rare but can occur.
If the composition of the mobile phase changes, it may affect retention time or generate ghost peaks or negative peaks, which will interfere with integration. To avoid such issues, dissolve the sample in the mobile phase or use a ghost-buster column to trap unwanted peaks.
2. Sample-Related Issues
- Sample carryovers: If the analyte concentration varies significantly, and the peak area does not stabilize until after multiple injections, it could be resulted from sample carryover. Ensure the sampler is thoroughly cleaned after each injection, and the solvent used for cleaning is miscible with the sample.
- Temperature fluctuations: Although the effect is usually minimal, temperature changes can still impact peak area. For example, when a sample is taken from the refrigerator and placed into the autosampler tray, it will gradually warm to the ambient temperature. As the solvent expands, the volume injected initially may differ from subsequent injections. To prevent this, allow the sample to stabilize at room temperature before injection.
-
Sample-specific issues: For example, some proteins in reversed-phase chromatography may not fully elute during the initial gradient phase but may elute during subsequent blank gradients. In such cases, a blank gradient should be run between analyses.
Additionally, some samples may adsorb irreversibly onto the column's active sites. When using a brand-new column, peak areas might slowly increase after repeated injections as the sample adsorbs onto these sites. Once the active sites are saturated, consistent peak areas will be achieved. To reduce specific adsorption, use a diamond-like carbon column or introduce other samples to saturate the active sites.
3. Flow Rate and Pump Performance
Random fluctuations in flow rate can also cause peak area variation, usually due to issues such as valve malfunctions or air in the pump head. These issues are typically accompanied by retention time variation, and can be diagnosed by monitoring the backpressure.
Another potential cause is leakage on the high-pressure side of the system. Most modern liquid chromatography pumps are equipped with leak sensors, making it easier to identify this issue.
4. Data Processing and Integration Errors
Integration problems are harder to address. While significant changes in peak integration can be detected visually, finer issues are more challenging to identify. When the signal-to-noise ratio is 100 or lower, tailing peaks can cause noticeable integration errors. Baseline noise limits the accuracy of integration, and issues such as peak fronting or tailing can exacerbate the problem.
To resolve this, try adjusting the integration parameters and reprocessing the data to see if the issue can be corrected.
Conclusion
This case study highlights the importance of a structured approach to troubleshooting peak area reproducibility issues in HPLC analysis. By systematically ruling out potential sources of error, Dr. Johnson successfully identified and rectified the problem, achieving the required analytical precision.
Understanding key factors in instruments, sample, pump, and data processing can help chromatographers maintain robust, reliable results in their own laboratories.