HP 5890 Series II Plus Reference Manual page 208

Chromatographic Troubleshooting
Retention time symptoms
2. Reproducibility is good early in the run but not toward the end:
3. Reproducibility is good later in the run but not for the first few peaks:
4. Retention time changes with amount of sample:
208
This may occur in temperature•programminga very densely
packed column; as column contents expand with heating,
resistance to flow may be so great that a mass flow controller
cannot maintain constant flow. Try increasing carrier source
pressure.
If this is the cause, the problem will either vanish or its onset will
move later in the run.
This may indicate too low a starting temperature; many stationary
phases have a minimum operating temperature, usually the
melting point of the material. Below this temperature, gas•solid
chromatography is performed; above it, gas•liquidchromatography
is performed. If an oven temperature program passes through this
range, results can be very erratic.
When earliest peaks elute very rapidly, they may not have had
time to achieve chromatographic equilibrium with the stationary
phase; they act like solvent peaks and are blown straight through
the column.
A useful rule is that peaks of interest should require at least four
times as long to appear as an unretained solvent or air peak. If
this problem is suspected, try reducing oven temperature 30 C to
approximately double the retention times.
Sample is overloading the column: When there is more sample
than the stationary phase can handle, peaks will be deformed and
shifted from correct retention times.
In gas•liquidchromatography an overloaded peak shifts to a
longer retention time and tails on its trailing edge. With gas•solid
chromatography the effect is the opposite. Try diluting the sample
or injecting less.