GC Troubleshooting - Some Common Problems

1•If the source of the problem is not obvious, a systematic approach will help to isolate it.•This flowchart gives some guidance on how to isolate where the problem is coming from.3•To rule out operating parameters as the source of the problem, consider how each can affect the chromatography.•Too high of a starting temp. can result in lighter analytes being lost under the solvent front.•If the splitter opens too soon, the response of heavier analytes can suffer. analytes.•For critical separations, the column flow should be set to get maximum efficiency from the column at that point in the run.•Having heated zones too cool:•injector: poor response•detector: poor response and tailing•MSD interface: tailing•Having heated zones too hot:•injector: sample decomposition•MSD interface: damage to column4•Now I will talk briefly about some common causes of baseline noise and drift.•You are probably all familiar with what column bleed looks like. It appears as a rise in the baseline from the beginning to the end of the run. I will touch on this in a little more detail shortly.•The symptoms of a dirty detector often include random spikes in the baseline.•Contaminants in the carrier gas can cause both an increase in the baseline background level and the drift.• Depending on the specific column and application, some contaminants can accumulate in the column while the oven is cool and start to elute off as the temperature is increased. (Such as the case of water contamination and a carboxen column.)•We shall see an example of the effect of carrier gas purity in the next slide.5•Now we will talk about column bleed.•Columns should be preconditioned before installing for the first time. Follow the instructions included with the column for starters, and modify for your specific application if necessary.• If columns are being reinstalled after sitting idle, they should be run through some sort of preconditioning prior to reuse.•Do NOT precondition columns into an MSD or an ECD. They could contaminate these detectors.•On an MSD, excessive column bleed can not only dirty the source, but also interfere with spectral identification.•This is an example of a typical bleed profile on a capillary GC column.•Bleed is measured as the difference between the baseline level at the end of the run and at the beginning. This measurement is usually done on an FID, and the bleed is recorded in picoamps (pA.)6•Now I would like to talk about column bleed and GC/MS.•MSD users have the unique advantage of being able to tell if baseline rise is actually being caused by column bleed or something else.•If an unacceptable rise is observed in the TIC, spectra should be taken at various points to determine if the column is the source.•A high interface temp. will not cause a classic bleed profile as we saw before, rather it will cause an increase in the baseline noise throughout the run.78•Now I would like to talk for a moment about gas purification.•Even high purity gas can become contaminated.•Changing cylinders•Dirty Gas lines•Off-gassing from the regulator•Dirty cylinder•Clean carrier gas will extend column life, reduce noise, and increase sensitivity. •Clean H2, air and detector makeup gases will result in better sensitivity.•Clean ECD makeup gas is essential to detector linearity. 9•Now I will touch on the causes of ghost peaks.•Ghost peaks may appear at the same point in every run, or may appear only every few runs.•Residue in the inlet can also include pieces of septa which can become lodged in the liner.•The wash solvents used on an autosampler can easily become contaminated. Change them daily.•Sometimes sample carryover does not necessarily appear as “ghost peaks”.•A sample containing a high level of a target compound could carryover into subsequent samples.•In certain situations, contaminants in carrier gas can condense on a column when the oven is cool and elute off as the oven is heated.•Some GC’s have internal traps for the carrier gas. If these become saturated, they can off-gas and cause ghost peaks in an analytical run.. •If an internal trap is packed with molecular sieve, it can off gas water after it becomes saturated.10•Now we will look at an example of how pieces of septa in the inlet can affect an analysis.•This is a mix containing a variety of active compounds. I use this mix to evaluate the condition of my system.Always do a blank injection on a newly installed liner prior to running samples. This goes for liners that you have “recycled” as well as brand new ones fresh out of the box. New liners will often produce extraneous peaks. These will go away after one run.1112•For cleaning liners, it is always better to replace rather than reuse.•Dual tapered and cup-style liners are difficult (if not impossible) to clean.•If mineral acid is used to clean liners, they will have to be deactivated again prior to use.•If you deactivate liners packed with glass wool, do so after the glass wool plug is in place.•Your time is valuable, keep a stock of new liners on hand rather than cleaning old ones out.•The bottom two chromatograms show the difference in inertness between liners packed with deactivated and untreated glass wool. We see that DDT almost totally breaks down into DDD and DDE when exposed to the untreated wool.13•We have all dealt with response problems at one time or another.•If a compound’s response was OK at one time, it is important to think about what could have changed in the system.•In day-to-day analysis (especially semivolatiles) sample decomposition is the most likely culprit of response problems.•As columns age and retention times shift, two peaks which were once resolved may now coelute.•When installing a brand new column of the same type, keep in mind that your retention times are probably going to be longer than on the column you just replaced. Extend your runtime on the first few injections to make sure everything comes off the column.•For methods development work, a test run should be done to determine if a compound is volatile enough to analyze by GC.•Since we have all installed capillary columns so many times, it is easy to overlook improper column installation as the source of the problem.•We will look at some examples of this a little later on.The smaller volume of the 2 mm ID liner increases the rate with which analytes reach the column – this provides for more efficient focusing. This really effects response for early eluting analytes.1415•Now for an example of how active sites in the column can cause response problems. •Nasty samples can quickly damage any column.•This is a chromatogram of a semivolatile standard on a PTE™-5 before injecting a particularly nasty sample.•Several active compounds are indicated.16•This is the same column from the last slide after making approximately three injections from a sample with a nasty matrix.•Notice the loss in response of the compounds indicated. •In the case of 2,4-dinitrophenol and 4-nitrophenol, no response was seen.•Pentachlorophenol is barely visible.•Sometimes, clipping a loop off the column can bring response back. In the case of this column, clipping did not help.•In this case, if a guard column had been used, the column may have been salvageable. 1718Column installation is something most of us who have been doing GC a while do not think about. Review the instructions from you GC manufacturer and make sure you are following them. As you can see here, column installation can affect peak shape – as we see in the top cgram. The column was installed too low in the inlet, resulting in peak tailing. Column installation can also affect response – in the bottom cgrams, we see the difference in response from having the column at different heights above the top of the ferrule. We get better response for the heavier compounds with a higher position.For GC-MS, make sure you have the column inserted into the source the proper distance. •It is also very important to use the right liner for your application.•Most semivolatile analysis requires splitless injection.•Increased sample contact time with the liner makes proper surface deactivation critical.•Voa analysis on MSDs without jet separators often requires a split injection.•Note: for this type of set-up, I have found the .75 mm ID SPME liner to work very well. It’s small internal volume aids in focusing the analytes on the column.2021222324•This is some suggested reading of materials available from Supelco which pertain to troubleshooting.•All of these bulletins are available on our website.•I have several copies of each here today for those who are interested.26