Splitless headspace gas chromatography with capillary column using automatic electropneumatic injection

Abstract

Splitless sample injection plays an important role in kinetic studies and sometimes also in trace analysis since this technique enables one to avoid the withdrawal of large amounts of substances from the sample, a quantity perhaps 10-50 times more than that needed for the investigation.Perkin Elmer instruments (F 42, F 45, H S 6) are widely used with capillary columns and splitting for headspace analysis. The electropneumatic dosing system of these instruments [5] works in such a way that the amount of samples administered to the column depends linearly on injection time; this is indeed the case with packed columns and capillary columns if injection with splitting is used. However, if splitless injection is applied for capillary columns at the usual carrier gas velocities, i. e.0.5-3 cm3/ min, a maximum curve is obtained (Figure l., curves 1 and 2) which indicates the system to be unstable. With the increase of carrier gas velocity in the control system the dependence approaches linearity (Figure 1 ., curves 3 und 4).The instability can be understood in terms of "swinging" oscillation of the system, a process triggered by opening and closing of the electromagnetic valve. This "swinging" renders the amount of sample introduced susceptible to the parameters of injection, i. e. duration of pressure balance, input pressure, free gas volume of the sample vessel, etc.The assumption of instability of the pneumatic injector if carrier gas flow rates are low is supported by the appearance of double or multiple peaks with one-component samples; the effect depends upon injection time (Figure 2). This indicates pulsing injection.Splitless injection requires the stabilization of the electropneumatic dosing system. This is effected by the insertion of a damping component (pneumatic resistor) (Figure 3) between the pressure regulator and the electromagnetic valve.On use of this modification of an electropneumatic dosing system, no double or multiple peaks were observed and a linear relationship is obtained between injection time and sample volume (Figure 4.). Sample amount versus Injection t h e as Influenced by carrier gas flow rate.Average of ten detetminations, relatlve empirical mean deviatlon 3.6%, substance: n-propyl alcohol 12 mg/cm3].

Towards properly controlled analyticalmeasurement methods

“…The calibration shown in figure has an S-shape. Having increased the signal-to-noise ratio 5-fold (the method used was described by Leisztner et al [10]), we obtained a linea relationship as demonstrated in figure 2. The signal-to.…”