Introduction
All of the examples shown in this year's troubleshooting section could actually be caused by more
than one problem. The key to resolving these performance problems lies in the proper approach to
troubleshooting. Begin by accurately defining the problem and developing a list of probable causes,
then write out a logical sequence of steps to investigate the problem. This will eventually lead to
isolation of the cause. The following discussion of each exercise should give you some guidance on
how to direct your troubleshooting efforts in the future.
Exercise 1
Definition of the Problem -- The bleed level for this column begins to climb at approximately 180°C.
This is well below the maximum temperature for this column and a rise in the baseline at this temperature
is not to be expected. Also, the baseline at the maximum temperature in the analysis is jagged. In
addition to the bleed problem, the peak shape for several of the compounds in the run is tailing.
Probable Causes
- Maximum temperature of the column has been exceeded
- Column is contaminated from sample or matrix residue
- Samples contain a high amount of contamination
- Column operated at low temperature for repeated analytical cycles
- Solenoid valve left in the closed position during standby
- Stationary phase decomposition
Logical Steps -- Because there are problems with peak shape along with higher than expected bleed, column
contamination might be the first area to explore. Trimming the inlet side of the column is an easy and quick
way to remove contamination that has built up on it. Removing the first 0.5 meters of the column is usually
sufficient to remove most of the contamination that could cause problems like those shown. After trimming,
the column should be reinstalled and column performance should be checked by running a test mixture or
column performance check sample.
If trimming the column does not improve performance, the column can be thermally conditioned by maintaining
the oven temperature at the column's maximum recommended operating temperature for an extended period of time.
Evaluate column performance by running a test mix again. If bleed and peak shape problems persist at this point,
a back-up or replacement column should be installed.
Solvent rinsing to remove whatever contamination is causing the problem could restore the performance
of the old column.
Exercise 2
Definition of the Problem -- Ghost peaks are present in the chromatogram. Peaks are well-defined and have
reproducible retention times. Response is high enough that they interfere with any peaks that run in the
same area of the chromatogram.
Probable Causes
- Syringe contamination
- Syringe wash solvent contamination
- Inadequate sample vaporization
- Run time is too short
- Sample contamination
- Sample backflash during injection
- Septum bleed
Logical Steps -- The first step in isolating the cause of this problem is to separate the components of the
analytical system. Ghost peak problems related to the GC and the column can be investigated by performing
an analytical run without making an injection. If there are no peaks present, then the injection port and
the column can be ruled out as sources for the ghost peaks. If there are peaks present, components of the
system can be changed one at a time to isolate the cause. Septa, followed by injection port liners, followed
by capillary columns can be changed in sequence to try to isolate the source of the ghost peaks. Changes
should be made one step at a time, and the performance of the system should be evaluated with a test mix
after every change.
If the GC and the column have been ruled out as sources for the ghost peaks, the focus of troubleshooting
should shift to the sample introduction process. Substitute a new syringe or a fresh bottle of syringe wash
solvent, again in a step-by-step sequence. If a new syringe and a fresh supply of wash solvent do not correct
the problem, the ghost peaks are most likely related to the sample preparation process.
In this case, where the peaks are well-defined and exhibit reproducible retention times, clues seem to
indicate that the source of the ghost peaks is related to the sample introduction process, most likely
contamination in the syringe or the wash solvent.
Exercise 3
Definition of the Problem -- Peak shape for the early eluting compounds in this analysis is poor. Most of
the peaks are tailing and resolution between some of the peaks is not as good as it should be. In addition,
the solvent peak is very broad and also exhibits a large amount of tailing.
Probable Causes
- Poor sample vaporization
- Too much sample injected
- Splitless injection parameters are not optimized
- Column is not installed properly
- Injection port liner is dirty
- Carrier gas leaks at the injection port
- Solenoid valve left in the closed position
Logical Steps -- The combination of a broad and tailing solvent peak along with poor peak shape and resolution
for early eluting components usually indicates that splitless injections are not being performed properly.
Sample size can be investigated by decreasing the amount injected and looking for improved performance. Sample
vaporization can be checked by increasing the injection port temperature in several steps and looking for
improved peak shapes.
However, the most likely cause of the problems shown is a temperature program that does not allow for solvent
focusing or cold trapping of the injection at the head of the column. Peak shape and resolution should be
improved in this example by lowering the initial temperature in the oven program to 40°C, holding at
40°C during the time the solenoid valve is closed, then rapidly programming the oven temperature to
150°C. After reaching 150°C, the program rate can be slowed to the original 4°C/minute rate
in order to get the best resolution between all of the compounds.
|
;)
;)
;)
