Diagnosing signal instability when processing data in Analyst®, MultiQuant™, or SCIEX OS software

For research use only. Not for use in diagnostic procedures.

Answer

Signal instability typically results in fluctuating responses from samples with the same level of analytes, causing significant variability in measured outcomes. An example of signal instability would be the internal standard peak areas varying significantly from sample to sample in the same batch. Due to random fluctuations in peak area, signal instability could also affect linearity and recovery. If signal instability is observed while processing data in Analyst®, MultiQuant™, or SCIEX OS software, users need to investigate the following possible influences on their signal:

1. Sample and material preparation

The complexity of the sample prep/extraction workflow and the quality of chemicals and materials used can impact method performance. Some examples include:

• Compromised columns (contaminated, expired, or exceeding the recommended number injections)
• Contaminated mobile phase
• Discrimination between samples during extraction (e.g., variable loss of analyte during solvent evaporation step)

2. LCMS method instability

An unsuitable LCMS method could cause signal instability. For example, higher than necessary source temperatures can promote the decomposition of specific analytes, producing a low signal vulnerable to minor temperature fluctuations.

Sometimes, methods do not show signs of instability, and other changes in the system will suddenly trigger the problem. For example, methods that don’t flush the LC column properly might not have any issues until the column is exposed to high levels of matrix or until the column has aged a few weeks.

3. Instrument issues

Another primary source of signal instability stems from issues related to hardware performance; contamination of the mass spectrometer, physical issues with source spray and nebulization, and faulty autosampler performance are examples of instrumental issues resulting in signal instability.

To determine the source of signal irreproducibility, the following experiment is suggested:

1. Evaluate the method and make necessary changes to ensure its robustness and safety. Create a test method under a different name for system diagnostic purposes. A simple, unscheduled MRM method with 20-30 transitions representing the problematic analytes is preferred.
2. Prepare a medium-level standard (1-1.5 mL, ideally neat and matrix-free) for repeat injections (STND). Aim for a concentration level that produces a signal (peak height) of about 2.0-5.0e5.
3. The standard should be prepared in 100% mobile phase A or the same solvent composition as t=0 in the LC method. This assumes a reversed-phase LC method (e.g., using a C18 column) and the solubility of analytes in such a composition.
4. Prepare a blank containing internal standard (BLNK).
5. Prepare a double-blank (DB) sample with no internal standard. Note: For blanks, use the same solvent source and composition as for the medium standards in step 2.
6. Submit a batch with samples in the following order: BLNK, DB, DB, BLNK, STND, DB, BLNK, STND, STND (10-20x), BLNK, DB. The objective is to inject a medium standard from the same vial 10-20 times.
7. After the batch is run, ensure the blank and double-blank samples do not show any sign of carryover or contamination.

A metric plot of the peak area for all injections of STND from the same vial and their standard deviation could illuminate the source of the problem. If these injections show poor reproducibility (RSD above 10-15%), the problem may be instrumental; otherwise, sample prep and materials used are most likely the source of instability.