# 8 Consumables That Solve the Most Common GC Problems
Most GC problems don’t start in the column. They start in parts that cost less than a coffee per run. A lab I know replaced a “dead” column that was actually choking on septum fragments, and the new column failed the same way within a week. Eight inexpensive consumables prevent most of that.
Eight consumables solve most GC problems: deactivated inlet liners, low-bleed septa, O-rings and inlet seals, ferrules, deactivated vials, syringe filters, gas traps, and crimped caps. Replace them on schedule and most tailing, leaks, drift, and carryover vanish before the column is blamed.
Here’s the short version of each, with the troubleshooting logic behind them. Read on, because item 4 alone can save you a column.
1. Deactivated Inlet Liners
Peak tailing is the classic sign that your liner has gone active. One week your pesticide recoveries look fine, the next the polar analytes shrink and their peaks lean to the right. Nothing changed in the method. The liner changed, one injection at a time.
What happens is simple. Every injection deposits a thin film of non-volatile residue on the glass surface. That residue exposes active silanol sites, and polar or reactive compounds adsorb there. Restek’s GC troubleshooting guide on tailing peaks explains that tailing usually comes down to a flow path disruption or reversible chemical adsorption, and that restoring a clean, inert flow path is the fix. A fresh liner is the fastest way to do that.
The chromatogram even tells you which failure you have. If every peak tails, think flow path: a cracked liner, a poorly cut column, or a leaky connection. If only a few peaks tail, usually your most polar ones, think adsorption on an active surface. That distinction is the difference between rebuilding the inlet and swapping to a better deactivated liner.
A good default for split injections of dirty extracts is a 4 mm single-taper liner deactivated with dimethylpolysiloxane. The taper directs sample onto the column and away from the hot metal base of the inlet, where breakdown happens. Glass wool adds surface area for vaporization and catches non-volatiles before they reach the column.
My rule: if your samples come from soil, food, or biological matrices, treat the liner like the filter it is. Swap it weekly, handle it with forceps and gloves, and never reuse one that has gone brown. It’s the cheapest peak-shape improvement you can buy.
2. Low-Bleed Septa
Septa cause two separate headaches. The first is a leak. Pierce a septum a few hundred times and the puncture channel stops resealing, so carrier gas escapes and air sneaks in. You see drifting retention times, falling head pressure, and noisier baselines. The second headache is coring, where the needle punches a plug of rubber that drops into the liner.
Agilent’s maintenance guidance for split/splitless and multimode inlets puts the replacement interval at 100 to 400 injections depending on the septa type, and sooner if you spot leak symptoms. Manual injections, blunt or burred needles, and thick-gauge syringes all shorten that interval.
I once chased a ghost peak for two days and it turned out to be the septa. A small core of rubber had fallen into the liner and was slowly bleeding siloxanes into every blank. The mass spec looked contaminated. The liner looked clean. The septum under a loupe had a visible hole in the middle. Ten seconds of inspection saved a third wasted day.
Buy septa with a PTFE face for low bleed, use a conical 23 to 26 gauge needle, and tighten the retainer nut until the C-ring sits about 1 mm above it. Overtightening actually drives coring, because the needle cuts deeper into compressed rubber.
3. Inlet O-Rings and Seals
O-rings and seals are the parts everyone forgets until the inlet won’t hold pressure. The O-ring seats the liner inside the inlet. The seal sits at the bottom of the inlet where it meets the column. Both live at 250 to 300 °C, and both degrade.
The consequences are quiet at first. Agilent’s inlet maintenance seminar notes that chromatographic symptoms such as unexplained extra peaks indicate the inlet base seal is dirty and needs replacement, and it recommends changing the seal whenever you loosen the reducing nut. That same seminar covers O-ring chemistry: standard Viton O-rings handle up to 350 °C, while graphite versions take the hotter jobs. See the details in Agilent’s GC inlet maintenance presentation.
Air ingress is the second cost. Every leak in the inlet path pulls oxygen toward the column, and oxygen at high temperature chews through stationary phase. Columns age fast, backgrounds rise, and in GC-MS the source gets dirtier with every run. A two-dollar O-ring stands between your column and all of that.
Here’s my inspection habit. Every time I pull a liner, I check the gold seal under a loupe for graphite dust and rubber fragments. If I see anything shiny where it should be flat gold, the seal gets replaced. A seal costs a few dollars. A day of baseline hunting costs the lab a whole sequence.
One warning: never reuse an O-ring that has been baked for weeks. It loses elasticity, seals unevenly, and can shed particles into the inlet. If you’re rebuilding the inlet, rebuild it completely.
4. Ferrules and Column Nuts
If retention times wander and the baseline hisses, suspect a leak before you suspect the method. Leaks at the column fittings are the most common pneumatic fault in any GC, and they’re almost always a ferrule problem: cracked graphite, an improperly cut column end, or a nut that was never tightened properly in the first place.
John Hinshaw’s column on fittings in LCGC’s “Making a Great Connection” is still the best practical summary. His key points: leaking fittings cause baseline noise and drift, unstable retention times, reduced column life, and inaccurate results. Retighten fused-silica connections about a quarter turn past finger-tight, re-check after the first oven heating cycle, and if a leak survives two or three tightening attempts, replace the ferrule.
Ferrule material matters too. Soft graphite seals well on fused silica and forgives imperfect cuts, but it creeps and sheds with heat cycling. Graphite-Composite blends hold up better for repeated thermal work. Match the ferrule to your instrument brand’s dimensions, because an 11 mm ferrule in a Shimadzu port or an Agilent one behaves differently even when it technically fits.
Never hunt leaks with soap solution. Surfactants contaminate the fitting and wick into the column, where they cause exactly the baseline garbage you were chasing. Use an electronic leak detector. It’s more sensitive than your eyes and it leaves nothing behind.
My habit: every new column install gets a new ferrule. A ferrule is a one-time compression seal, like a gasket. Reusing one is how a five-minute install becomes an afternoon of pressure instability. Log the leak check before the first sequence and you’ll thank yourself later.
5. Deactivated Autosampler Vials and Inserts
Carryover and vanishing peaks at low concentration often trace back to the vial, not the injector. Ordinary glass surfaces carry silanol groups and metal contaminants. Basic compounds stick to them. Hydrophobic compounds stick to them. Your 0.1 µg/mL standard loses half its response before the needle ever draws it up.
Contaminated vials cause bigger problems too. Agilent’s troubleshooting article on the sources of ghost and carryover peaks lists contaminated sample vials among the causes, alongside dirty syringes and septum bleed. Deposits from previous samples leach back into the chromatogram at random retention times. In GC-MS that shows up as background ions in your blanks, which is exactly how a clean method starts failing audits.
The fix is deactivated glass. Silanized vials cap the active sites, so analytes stay in solution where they belong. For volumes under about 100 µL, use a conical insert or a low-recovery vial so the needle actually reaches the liquid. Pair them with PTFE-faced septa and the whole container becomes inert.
One habit worth stealing: never refill and rerun a vial. The residue film from the first fill doesn’t rinse out with a quick solvent swirl, and the second injection inherits every sticky compound the first one left behind. Dispose, rinse outside the instrument if you must, and start fresh.
A word on economics. Deactivated vials cost maybe thirty percent more than standard ones. If your method reports trace-level results, that premium is trivial next to a reinjection queue. Match the vial to the sensitivity of the analysis, and keep one compatibility chart at the bench so nobody improvises.
6. Syringe Filters for Dirty Samples
Some samples are hostile. Soil extracts carry grit. Dissolved tablets carry binders. QuEChERS extracts carry plant waxes. Inject any of them unfiltered and that material goes straight into your liner, where it cakes onto the glass wool and the walls. Within a dozen injections your tailing is back, your responses sag, and you’re rebuilding the inlet again.
Filtration removes that burden before it enters the instrument. It’s the difference between protecting the system and periodically rescuing it. For most organic extracts, a 0.22 µm PTFE membrane is the workhorse: it handles strong solvents and catches fine particles. For aqueous samples, use regenerated cellulose or nylon instead, because PTFE is hydrophobic and resists wetting.
I keep a short mental checklist. Filter every sample that isn’t visibly crystal clear. Filter standards and samples through the same membrane type, since different membranes leach different extractables. And for really gritty matrices, put a glass fiber prefilter in front, so the fine membrane doesn’t blind after two milliliters. We cover the reasoning in more depth in our guide to filtering samples before injection.
The math is easy to love. A 13 mm filter disc costs cents. An inlet rebuild plus a trimmed column plus a rerun costs the afternoon. If your lab processes dozens of extracts a day, batch the filtration into one session with a small manifold and it adds seconds per sample, not minutes.
Match pore size to your detector too. Most GC methods are happy at 0.22 µm, and going finer rarely helps the chromatography while it slows the draw. The one exception is fine precipitates from saline or buffered matrices, where a glass fiber prefilter ahead of the membrane keeps things moving.
7. Gas Purification Traps
When did you last change your carrier gas traps? Most people can’t remember, and that’s a problem, because the carrier gas touches every part of the system all the time. Moisture and oxygen from cylinders and fittings destroy columns slowly and noisily: baseline rises under temperature programming, bleed creeps up, and sensitive detectors hiss.
The numbers matter here. Thermo Fisher’s GC installation guidance specifies water and oxygen content below 1 ppm in carrier gas, and it documents the correct trap order: the moisture trap sits closest to the gas supply, the hydrocarbon trap comes next, and the oxygen trap goes last in the series. Water in the lines damages the analytical column, and oxygen decomposes the stationary phase. You can read the full specifications in Thermo Fisher’s GC site preparation manual.
Indicating traps make this easy. They change color as they saturate, so a glance at the manifold tells you the state of your gas. Replace the cartridge at two-thirds exhaustion rather than waiting for the color to hit the end. A trap run to failure dumps everything it ever caught back into your line.
MS labs should treat this as non-negotiable. Oxygen and water in the carrier stream shorten ion source cleanliness intervals, drive up tune failures, and slowly cook the column’s stationary phase into siloxane background. A set of fresh traps costs less than one source cleaning, and vastly less than a column replaced a year early.
Buy traps with clear capacity ratings and dated labels. Then write the install date on the body with a marker, because your future self, staring at a rising baseline, will want to know.
8. Crimp Caps for Volatile Work
Sample evaporation is the invisible problem. A vial sits in the tray for fourteen hours, losing volatile analytes through an imperfect seal the whole time. Your early injections look strong, your late injections fade, and your internal standard recovery slides down the sequence like a slow leak in a tire.
Crimp caps win here for one reason: compression. An aluminum crimp squeezes the septum between the cap and the glass flange, giving the tightest seal available and the lowest vapor loss. Screw caps are more convenient and safer for your fingers, and a properly torqued screw cap works fine for routine work. Snap caps sit at the bottom of the list for volatiles; they’re convenient, but they rely on friction alone. For volatiles, headspace, and long unattended sequences, crimped seals are the conservative choice. Agilent’s own troubleshooting notes warn that loose vial caps cause selective loss of the lighter materials in a sample.
Two habits keep crimping honest. Crimp a test vial at the start of each new box and try to spin the cap; if it rotates, the seal is loose and the crimper needs adjustment. Then weigh a filled, capped vial before and after a long run. If you lose more than about one percent of mass, your seal is costing you data.
Fill volume belongs in the same conversation. A vial filled to the shoulder evaporates far less than one filled to the neck, because there’s simply less vapor space in play. Leave headspace for expansion and you protect both the seal and the autosampler needle.
We’ve written a full practical guide to preventing sample evaporation in autosampler vials, including fill-volume limits that matter just as much as the cap itself.
Conclusion
None of these eight consumables is glamorous. All of them are cheap. That combination is exactly why they get neglected, and why so many GC problems that look like column failures, method failures, or instrument failures turn out to be a worn septum or an exhausted trap.
My advice: stop treating these parts as emergencies and put them on a schedule. Septa every 100 to 400 injections. Liners weekly for dirty samples. O-rings and seals at every rebuild. Traps when the indicator says so. One quiet hour of preventive maintenance buys back days of chromatography.
If you want to go deeper on the two parts that fail most often, start with our guide to GC inlet liners and our walkthrough of the headspace vial mistakes that ruin GC data. Your columns will live longer, and your baselines will stay boring, which is exactly what a baseline should be.
Frequently Asked Questions
How often should I replace GC inlet liners and septa?
Replace septa every 100 to 400 injections depending on the septa type, and sooner if you see retention drift or pressure loss. Replace liners weekly when running dirty matrices, or any time you notice tailing, rising baseline, or visible residue. Many labs pair the two into one weekly inlet maintenance routine.
What causes unexplained peaks in GC blanks?
The usual suspects are septum bleed, a contaminated inlet seal or liner, dirty carrier gas, and residue leaching from sample vials or syringes. Agilent recommends a condensation test to separate gas-supply contamination from inlet contamination. If the peaks are broader than your sample peaks, look for semi-volatile carryover from a previous injection.
Are crimp caps really better than screw caps for GC?
For volatile analytes, headspace work, and long sequences, yes, because the compression seal loses less vapor. For routine liquid injections of non-volatile analytes, a properly torqued screw cap performs well and is faster to handle. The seal quality of the septa matters as much as the cap style, so match PTFE-faced septa to volatile work.
Which gas trap should be installed first in the line?
The moisture trap goes closest to the gas supply, followed by the hydrocarbon trap, with the oxygen trap last in the series. This order protects the oxygen trap from saturating with water first. Check the manufacturer’s rated capacity and replace indicating cartridges at about two-thirds exhaustion.
Do deactivated vials matter for GC-MS specifically?
They matter more as concentrations go lower and analytes get stickier. Basic and hydrophobic compounds adsorb to untreated glass, which distorts response at trace levels and encourages carryover in MS sources. Silanized or low-adsorption vials with PTFE-faced septa reduce both problems, and they cost very little compared with a reinjection queue.
