Correct Answer
c. LC pump check valves need cleaning or replacement
Investigation
Looking at peaks in the quantitative data review option, where one analyte at a time with matching internal standard is displayed, was confusing as some peaks were normal, some were asymmetric and different peaks were missing in different samples. The first step to solve the problem was to use the qualitative data review option in the LC-MSMS software to create overlays of total ion chromatograms (TICs) and the LC pressure traces for a normal sample and a problem sample (Figure 1, Figure 2, Figure 3, and Figure 4).
Two things became apparent with this view of the data. Missing peaks occurred in the time frame where the volume % of mobile phase A (water + formic acid) to mobile phase B (acetonitrile + formic acid) generated the maximum pressure (>9,500 psi – see Figure 2 and Figure 5. Figure 5 shows overlay of normal TIC, pressure trace and LC gradient shown as change in % mobile phase B.).
At this higher pressure – abnormal low pressure spikes at regular intervals were apparent in the pressure trace for problem samples. The downward spikes did NOT appear at the beginning and end of the LC gradient when pressure was <9,500 psi or in normal samples.
A quick way to distinguish between an instrument problem (leak, check valves, bad mobile phase) and a sample preparation problem (mistake adding internal standard) is to inject a system suitability test (SST) sample (neat standard in pure injection matrix with no sample preparation) or an extracted sample from a previous batch with known good results.
Diagnosis
The SST sample also had missing peaks in the middle of the run. So this is an instrument problem rather than a sample preparation problem.
Short term, regular, repetitive loss of pressure as seen in Figure 2 occurs when there is an air bubble in the LC pump head or when the LC pump check valves become “sticky” and mobile phases no longer move in and out of the pump head with normal timing. As mobile phase delivery to the column becomes irregular, Rt shifts are a common symptom of dirty check valves. It is less common in our experience to see normal chromatography at the beginning and end of the LC gradient and Rt shifts only in the middle of the LC gradient.
It is important to understand that all compounds are still eluting from the LC column in the expected order. But as the MSMS method is designed to acquire each MRM within a narrow time window (+/- 0.2 min around the expected Rt) – peaks may be incompletely defined (bizarre shapes) or missing if the observed Rt shifts too far away from the expected Rt and to the boundary or outside the MSMS acquisition time window for that MRM.
Corrective Actions
Priming or purging the LC pumps will typically remove air bubbles. A prolonged purge of the LC pumps didn’t correct the problem. Dirty inlet and/or outlet check valves was the next most likely source of the problem. The check valves were replaced and that solved the problem.
The hypothesis is that the check valves functioned abnormally only at the higher pressure (>9,500 psi) that occurs in the middle of the LC gradient. Therefore – early (<2 min) and late (>4 min) eluting peaks had normal Rt in many samples.
For information on replacing check valves – consult your instrument manual and/or vendor Technical Support. For information on how LC pump check values work – see http://www.chromacademy.com/troubleshooter-hplc/resources/15021997.pdf.
Keeping spare inlet and outlet check valves in stock and knowing when and how to change them can return the instrument to normal performance very quickly – without having to wait for a service representative to arrive.