Only monatomic carrier gases can be employed in the HeD and ArD, as only these gases form metastable atomic states. The character of the carrier gas determines the energy of the metastable state and thus decides whether the eluted substance will be ionized. The impurity content in the carrier gas significantly affects the maghitude and character of the HeD response, as discussed above. For this reason, many workers have dealt with the purity of the carrier gas [2-4, 9, 10, 12, 19, 31, 38, 46, 471. We feel that the most effective purification line is that discussed by Bourke et al. [3, 41.
Most of the H,O, CO, and easily condensed gases are removed by molecular sieve 5 A at room temperature and at the temperature of liquid nitrogen. Nitrogen is removed by passing the carrier gas over hopcalite (a mixture of copper, manganese and their oxides) at 350 "C. The last stage again contains a molecular sieve at the temperature of liquid nitrogen, to remove the remaining H,O and CO,. In this way,
the concentrations of impurities were decreased to the following values (ppm):
COz < 0.02; CH, < 0.05; CO < 0.06; O 2 < 0.07; H,O < 0.1; N1 < 0.2; H , < 0.5;
Ar < 0.5; Ne about 1 to 10.
When the effect of impurities on the HeD and ArD response is evaluated, the max- imum concentrations of impurities that do not significantly affect the SHeD value are given [9, 19, 471. The lowest concentrations are required for electronegative impuri- ties, e. g., oxygen and water [45], which significantly decrease the HeD response due to electron capture.
The HeD background current changes with changing carrier gas flow-rate [lo, 12, 461. Experiments with very pure helium [3] showed a decrease in the background current by 10% for a change in flow-rate from 50 to 120 cm3/min. When the carrier gas contains an impurity, the background current first increases with increasing flow-rate and then slowly decreases [lo]. The response then becomes negative.
Independence of the ArD signal of the carrier gas flow-rate from 20 to 100 cm3/min has also been observed [24].
8.3.2 Construction of the helium and argon detectors
The geometry of the HeD detection space strongly influences the measurement results and is indirectly reflected in the detector nomenc!ature. The literature contains refer- ences to macro-, micro- and triode-detectors, the schemes of which are depicted in Fig. 8.5. It should be remembered that the free path of /?-particles is very short in
3
,- F
,I 4
13 t4 14
FIG. 8.5. Scheme of the argon detector; (a) micro-detector with homogeneous electric field; (b) macro-detector with heterogeneous electric field; (c) triode argon detector;
1 - radioactive source and cathode, 2 - anode, 3 - column, 4 - rinsing gas, 5 - gas outlet, 6 - triode [23].
HeD and ArD. About 90:/, of non-elastic collisions occur within 0.4 mm of the surface of the /J-source [15]. It is thus obvious that, in a macro-detector, the electron released on ionization must pass through large distances in order to reach the collecting electrode. During this transport, the probabilities of recombination, capture, etc., increase and the signal is lower than that of a micro-detector [39].
142
The electric field intensity exerts a decisive effect on the detector performance.
It has already been mentioned in the introduction that the HeD and ArD are the only ionization detectors that operate in the multiplication region of the volt-ampere curve. The electric field intensity reaches values as high as 4000 V/cm [13, 301 and secondary electrons are thus accelerated to energies capable of directly ionizing eluted substances (equation (8.5)). When an argon micro-detector is employed, intense elec- tric fields are attained at lower voltages and the detector sensitivity is thus increased.
Either homogeneous electric fields (see, for example, ref. 12) or, more frequently, inhomogeneous fields are employed. I n order to suppress the space charge effect and to decrease the probability of recombination in inhomogeneous electric fields, designs containing a third electrode with a potential more negative than that of the anode have been proposed. The collection of cations in the effective detector space is thus improved and the background current and noise are decreased.
6
7’
1+-
- L l -
50 100 150 200 250 300
PCI
FIG. 8.6. The dependence of SAID on the detector temperature at various voltages.
As sources of primary particles, b-sources are most frequently employed, although an a-source, 226Ra, has also been used. At present, 63Ni and 3H are used most frequently. These cources have activities of about 100 mCi; as follows from equation (8.4), the HeD response increases with increasing activity. The use of radioactive sources is discussed in Chapter 3. Metastable atoms are also generated by an electric discharge in neon [26].
The detector temperature affects the number of particles emitted and thus changes the HeD signal. Experimental results with the ArD are shown in Fig. 8.6.