2.3 Effect of experimental parameters on the magnitude.and shape of the TCD signal
2.3.2.4 Temperatures of the sensor and the cell walls
From the principle of the thermal conductivity detector, it follows that T, must generally be larger than T, (when a heated filament is used), so that (T, - T,) 2
I
- 200 "C. The detector signal is directly proportional to the difference between the temperatures of the heated filament and the cell walls, from which follow important conclusions concerning the adjustment of the experimental conditions.
At a selected detection block temperature, SrCD is proportional t o the temperature of the heated filament, i.e., to the heating intensity (Fig. 2.8); on the other hand, at a given temperature of the heated filament, i.e., with constant heating, STCD is inversely proportional t o the detection block temperature. In practice, it is most advantageous to maintain the temperature of the heated filament as high as possible (taking care not to burn the filament) and the detection block temperature as low as possible (avoiding condensation of eluted substances in the cell).
When working with columns with programmed temperatures, the possibility of a decrease in the TCD signal must be borne in mind if the detection block is connected to the column thermostat. During an increase in the temperature of the column thermostat, the temperature difference (T, - T,) decreases and consequently STCD also decreases.
The above rules also hold for thermistor sensors. It has been found experimentally that the highest measuring sensitivity is attained for a small difference between the sensor and cell temperatures, (T, - T,) = 35 - 50 " C , while the sensor temperature should not exceed 100 "C. STCD is strongly dependent on the detection block tempera- ture:
Temperatures of the sensor and the cell walls
STcD 8 1 . 5 R 1 l 2 T - 2 f c (2.21)
Therefore, it is preferable t o maintain the detection block temperature as low as possible.
2.3.2.5
The magnitude of the time constant depends on the effective volume of the detector.
The requirement that the time constant of the measuring device should be as low as possible, leading to the smallest possible distortion of the elution curves, is partially in opposition to the requirements regarding the magnitude of the detector signal concerning, for example, the length of the heated filament or a large thermistor radius.
Time constant of the TCD
The cell time constant is expressed by the relationships
vdet = ' * 'dst
and
7 = 0.632td,,
These relationships are valid, however, only for cells in which transport occurs
exclusively through convection. A number of thermal conductivity designs have been proposed in which both diffusion processes and mass convection towards the sensor surface participate. According to this criterion, cells are classified as flow-through, semi-diffusion and diffusion. The shapes of these cell types are depicted in Fig. 2.9.
FIG. 2.9. Various shapes of thermal conductivity cells; a - flow-through, b - diffusion, c - semi-diffusion.
The expression for the time constant of flow-through cells reflects the significant dependence of their signal on the gas flow-rate. Therefore, manostats are placed before the cells in order to stabilize the flow-rate and the pressure [12]. The smallest distortion of the shape of the elution curve is achieved with a low time constant;
designers of thermal conductivity cells thus attempt to make the flow-through cell volume as small as possible. Cells with volumes of 20 pI [21], 2.6 ,d [32] and even 1 pL1[23] have been described.
TABLE 2.5
TIME CONSTANTS OF VARIOUS CELL TYPES I N THERMAL CONDUCTIVITY DETECTORS
Cell 7 [sec] Notes
Flow-through 0. I - 1 most frequently used, most sensitive to all changes
Diffusion 10-20 unsuitable for modern
measuring requirements
Semi-diffusion o1 / r 2 > 1 up to 10 properties of the flow-through cell
C I / U 2 J . 1 up to 20 properties of the diffusion cell
54
It is obvious that the time constant will increase with increasing participation of diffusion in the transport process. The time constant of a semi-diffusion cell can be expressed by the relationship
z = 0.632t,,, - 1 2 - - V
111
where trl is the gas flow-rate through the measuring branch (see Fig. 2.9). The time constants of the cell types discussed are listed in Table 2.5. In addition to the time constant of the measuring cell, the sensor time constant, determined by its mass and heat capacity, must also be considered. Thus the sensor requires a certain time to record a change. The time constant of a heated filament is given by the relationship
As the heat capacity of the filament is proportional to its volume, i.e., mC, - $1, the
requirement of high T C D sensitivity (see equations (2.9) and (2.19)) leads to an increase in the sensor time constant. The heat capacity of the thermistor approxima- tely equals the material constant, 6, and varies around 1 sec for most sensors of this type. In a provisional arrangement, the transistor sensor had a large time constant of 10 sec [31], which the authors felt could be decreased to 0.6 sec.
In some designs, the heated filament is sealed in glass in order to decrease the catalytic action of the heated filament on the thermal decomposition of substances and to prevent corrosion of the sensor. Similarly, materials other than glass have also been employed, e.g., fluorinated plastic [ 5 ] . These modifications lead to an increase in the lifetime of the sensor and make its use at higher temperatures possible, but the
TABLE 2.6
THE TIME CONSTANT OF THE THERMAL CONDUCTIVITY DETECTOR Sensor
Cell
heated filament
filament in thermistor transistor
Length of cell 2 10 2 0.2 0.2
Diameter of cell 0.5 0.002 0.1 0.2 0.2
Gas flow-rate [ml/min] 30
Time constant [sec] 0.45 0.01 4 1 9.6
- sensor [cm]
- sensor [cm]
_- .I
U
f f
time constant of the sensor and consequently that of the whole detector are increased considerably due to an increase in the heat capacity of the measuring element. Cell time constants and those of individual sensor types are given in Table 2.6, from which it follows that a heated filament has the lowest time constant. It should be used in combination with a flow-through cell whenever theoretical or quantitative empirical conclusions are to be made on the basis of the measured results.