High performance capillary electrophoresis - T.E.A.M.

Instrumentation/Operation
If sensitivity is not limiting, the smallest injection lengths
possible should be used. However, injection reproducibility
is usually diminished with short injection times due to
instrumental limitations. This is especially true when short
and/or wide-bore capillaries are employed or when concentrated
samples are used. Reproducibility can be improved
significantly by use of an integrated pressure/time profile
with active feedback control to compensate for system
risetime effects and variations in the applied pressure.
From an instrumental standpoint, injection reproducibility
can be better than 1 to 2 % RSD. Reproducibility in peak
area, however, can be reduced by other phenomena, including
sample interaction with the variations in capillary
temperature, capillary walls, integration of peaks with low
signal-to-noise ratios, and so on (see table 19).
Precise temperature control (± 0.1°C) of the capillary is
necessary to maintain constant injection volume. As with
migration time, viscosity of the buffer in the capillary and
thus the injected quantity varies 2 to 3 % per °C. Note that
sample viscosity does not significantly affect injection
volume since the sample plug is only a very small volume
relative to the total liquid volume in the capillary.
Finally, to avoid unwanted sample injection by siphoning it
is important that the duration of the injection be as short as
possible. In addition, the liquid levels of the sample and
buffer reservoirs should be equal. Siphoning can cause poor
peak area reproducibility and even overloading. It has also
been found that simply placing the capillary in a sample
reservoir will cause an injection due to capillary action. This
phenomenon has been called a zero-injection effect. While
often insignificant, with concentrated samples the injected
amount can be quantifiable and should be considered during
quantitative analysis.
86