The family of radiation detectors that are based upon collection of the ion pairs produced in a gaseous medium may be characterized by the simple circuit shown. The electrodes may be in flat, cylindrical, or other geometry, and the chamber can be filled with almost any gas, with Argon, Helium, or air being widely used. Small amounts of other gases are often added to optimize counter performance.
The basic operating principle of the device is simple: ionizing radiation (alpha, beta, or gamma) entering the chamber causes ionization of the gas, and the resultant electron/ion flow to the electrodes causes a small surge of current (a pulse) which may be read by an appropriate meter circuit. If the circuit integrates the surges, a time-averaged current indication results. If the circuit differentiates, the pulse rate is indicated.
Looking at the operation in more detail, we see that the behavior of an ion pair formed in the chamber depends on the charge applied to the plates:
![[Top of Page]](../../Buttons/top.gif)
If the charge is low, the charges will move slowly toward the plate of opposite sign. Briefly accelerated, they lose most of their energy in collision with atoms in the gas. The slow movement ("drift") permits recombination of the plus and minus components, so that less charge arrives at the plates than produced in the gas volume.
Increasing the voltage across the plates produces a higher acceleration of the charged particles, greater particle velocity, and lower probability for recombination. The increase in the charge collected on the plates is indicated in region 1 of the performance map shown.
As the voltage is increased further, the charge arriving at the plates increases until it equals the charge formed in the gas.
Increasing the voltage modestly beyond this level only serves to get the charged particles to the plates faster: there is no increase in the amount of charge arriving. Devices operating in this region are called ionization chambers. The impressed voltage is in the order of 300 volts, and the pulse produced is typically less than 1 mv.
The charge produced is proportional to the energy lost by the incoming radiation. Two curves are shown. They indicate that any one detector has a range of responses which depend upon the incident radiation:
The major effect is a function of the type of radiation entering the chamber, and is related to its "specific ionization". The specific ionization for alphas is approximately 3 orders of magnitude greater than that for betas...which means that an alpha particle will produce 1000 times more ion pairs per unit distance. More ion pairs produced generates a larger current or pulse. Gamma rays, on the other hand, have very much larger mean free paths, providing a much lower specific ionization than betas, and generate a smaller pulse. Note also that the radiation having a longer mean free path is likely to scatter out of the chamber before depositing all the energy it carried into the sensitive volume.
Also, as the energy of the incident radiation increases, the energy deposited in the chamber increases, increasing the charge reaching the plates.
Ionization chambers are often used to measure neutrons (neutrons are not "ionizing radiation": they interact with the nucleus of the atom, not the surrounding atoms). By adding Boron, specifically 10B, into the chamber in either gaseous or plate form, the interaction 105B(neutron,alpha)63Li occurs during neutron bombardment. It is well suited to measure thermal neutrons with good efficiency. The 10B cross section for the interaction is over 5,000 barns, and the Q-value for the process is about 2.3 MeV. Thus, the strong pulses from this interaction are easily seen and other, smaller signals originating from other beta and gamma radiation may be filtered out by a discriminator circuit.
Increasing the applied voltage further accelerates the primary electrons produced to an energy which permits them to ionize other atoms of the gas. These secondary ion pairs will then contribute to the pulse. This "gas amplification" increases with increasing H.V. After passing through a transition region (2), the instrument operates in the proportional region. Here the pulse produced is proportional to the primary ionization caused by the incident radiation. The amplification easily reaches 106 and the proportionality assures continued separation of the alpha, beta, and gamma behavior.
Many ionization-type counters are constructed with a cylindrical geometry, having a central anode(+) wire and a concentric cylindrical cathode (-). In this geometry, the electric field varies as shown. It is evident that the electrons are subject to increasingly higher field strength as they move toward the center wire, hence ionization events occur more closely spaced, and we can conclude that the region near the center wire accounts for the major portion of the gas amplification.
Because alpha and beta radiation has an extremely short mean free path, transmission of these particles into the active chamber through its walls is impossible, and the sample is often placed directly into the sensitive volume.
By adjusting the gain, or establishing a window threshold in the signal processing circuit, a proportional counter may be used to distinguish between the various types of ionizing radiation.
Use of Boron trifluoride (BF3) in the detector provides a mechanism for detecting neutrons, and rejection of smaller signals from other ionizing radiation enables use of this type detector to measure neutrons without additional correction.
Continued increase of the H.V. beyond the proportion region is influenced by two main factors:
Note that low-energy photons (X-rays) are also produced when an electron recombines with an ion. That low-energy photon travels a longer distance in the chamber than the electrons, and it may interact with the fill gas, producing an ion-pair at another location not near the previous event.
The effect of these events begins to cause the collected charge to move away from proportional behavior, region (3), and further increase causes an electron avalanche which blankets the entire anode. Now, the entire wire collects electrons regardless of the number of initial ion-pairs formed. This is the Geiger region. When operated in this mode, the pulse produced is the maximum achievable, about volt, and independent of the radiation which initiates the pulse.
It is important to recognize that the size (duration) of the pulse is limited by the buildup of positive ions near the wire: remember that most ion-pairs are produced near the wire. The light electrons move rapidly to the wire, but the very heavy ions accelerate at least four orders of magnitude slower. Staying in that region longer, they reduce the electric field in that region, and electron multiplication stops.
These heavy particles, the ionized atoms of gas, continue to affect the behavior of the tube: depending on the tube dimensions they may reach the cathode about 100 micro-seconds later. By masking the charged surfaces , the electric fields cannot act on any subsequent ion-pairs, preventing electron multiplication and the following avalanche from occurring until the slowly-moving ions clear out. The detector is said to be insensitive for this time. As a result, Geiger detectors should only be used for reasonably low count-rate situations, although there are techniques available for correcting modestly higher count rates.
Finally, when the heavy ion slams into the cathode there is a great probability that a photon will be emitted from the surface, generating a secondary, or follow-on pulse as the photon ionizes the gas in the chamber. This event has no relation to the external radiation that is being measured, and lead to erroneous count rates. In order to combat this phenomena, a "quenching gas", many times alcohol or other organic gas, is added to the counting gas (generally less than 10%).
There is a sufficiently large population of alcohol molecules in the gas to ensure that the ions will collide with an alcohol atom. In that collision, a more loosely bound electron will move from the alcohol to the ion, and that now-complete gas atom will drift in the chamber. The resulting positively-charged alcohol molecule now moves to the cathode. When it strikes the cathode, the excess energy is not released as a photon, but goes into splitting (dissociating) the alcohol molecule. No second pulse (an afterpulse) is generated.
This technique does consume the quenching gas, as the dissociated components do not recombine, and the tube (or the gas) must be replaced after considerable use, about a billion pulses, depending on the particular tube geometry and fill gas components.
To permit low-energy charged particle entry into the sensitive volume a "thin window" is often provided. Many times constructed of Mylar, it allows entry of many (relatively) low-energy betas, but is susceptible to mechanical damage if handled.
Since the ionizing medium is a gas, a long mean-free path is presented for gammas, with resultant low interaction (initial ionization) rate, leading to a low overall efficiency.
Returning to the relatively long time that it takes for the very heavy positively charged particles to move to the cathode, the chamber is incapable of producing the next avalanche of electrons until their arrival at the cathode. In essence, the tube is "dead" for this period, and considerable time must elapse before another event can be recorded. This insensitive time, tau, is in the order of 150 to 250 microseconds. Thus, when n counts are recorded during a counting period, the observed count rate (r = n/t), must be corrected in order to find the true count rate (R = N/t). N is the number of counts that would have been recorded by a tube which has no insensitive time.
The true count rate (R) may be related to the measured count rate (r) by correcting for the insensitive (or "dead") time. When the counter receives a pulse, any initial ionizing event that follows within tau seconds cannot be recorded, as indicated on the diagram.
We see that r counts per minute were recorded in a period when the tube could accept pulses for a time (1 - rt) minutes. Normalizing the counts received to the time when the tube could receive them, we find that:
or
Note that at relatively low count rates, r and R are about equal and we can substitute:
or
Note also that both forms predict that at large count rates the observed count rate will continually decrease relative to the "true" count rate as the true count rate increases.
