Parameters Affecting Performance Characteristics

The sensitive area is important because it affects both efficiency and energy resolution. When a low-intensity radiation source is used or when an accurate particle count is required (within the count-rate limits of the system), a large-area detector is desirable. However, since detector capacitance and electronic noise are proportional to the area, smaller detectors give much better resolution. Selecting the right detector size requires a compromise between efficiency and resolution.

For energy spectrometry each output pulse must be generated with an amplitude proportional to the energy of the charged particle. Therefore, for these common applications the detector's sensitive depth must be sufficient to completely absorb all the particle energy (Fig. 3). As the sensitive depth increases, the detector capacitance C_D decreases, and this results in a decrease in preamplifier noise. However, the increase in sensitive depth also increases the sensitive volume of the detector, and this may increase the detector leakage-current- noise contribution. Minimum total system noise is obtained by matching the capacitance of the detector to the appropriate preamplifier.

For high-resolution timing applications, in which the rise time of the information pulse must be very short, the charge transit distances have to be kept as small as possible and large electric fields maintained. In such cases, the sensitive depth may be restricted by the need for very precise timing information and occasionally by the need to discriminate against unwanted background.

The major effects of detector capacitance are its influence on the noise contribution from the preamplifier and its deterioration of the preamplifier rise time. In applications that require low total noise, it is necessary to minimize the capacitance C_D by restricting the active area and/or by optimizing the sensitive depth. Stray capacitance from cables, connectors, etc., must be added to the detector capacitance to establish the total capacitive load that determines the preamplifier contribution to the noise, and therefore must be minimized.

The minimum electric-field strength required for complete charge collection (i.e., optimum resolution and response linearity) depends on the mass (specific ionization density) of the charged particle being analyzed, with the more massive particles requiring higher field strengths. For charged particles (alpha particles or lighter), this minimum field is attained by meeting the required resolution specifications. For heavy-ion (fission-fragment) detectors, however, and for very thin totally depleted detectors, in which the resolution cannot be routinely tested, the minimum specified electric-field strength has been established by experimental data obtained during actual use in the field. In applications requiring very high-resolution timing, it is desirable to keep the average field strength as large as possible, consistent with optimum noise and sensitive depth.

For a given resistivity material the breakdown voltage of the diode establishes an upper limit on the electric-field strength and on the depletion depth. ORTEC does not use breakdown voltage as a basic specification, because it is redundant if the sensitive depth, noise, resolution, and/or field strength are specified.

A large reverse-leakage current results in detector noise and excessive voltage drop across the bias supply resistor (R_b).* Since a quantitative relation between the detector leakage current and noise can be established only through a detailed knowledge of the origins of all current components, detector  noise performance has been selected by ORTEC as the basic performance specification. Detectors whose leakage currents would produce excessive voltage drops across R_b are rejected by our quality-control standards. All ORTEC detectors are furnished with detailed data on their original leakage current so that this information may be used for troubleshooting and for estimates of the drop anticipated across R_b.

*Resistor, usually located in the preamplifier, has a value of 10 MW.

Noise sources in the detector and the preamplifier introduce a dispersion that broadens a pulse-height spectrum of mono-energetic particles. Noise is customarily specified in terms of FWHM (full width half maximum) broadening of a mono-energetic peak. The detector and the preamplifier are separate and independent sources of noise, and the total system noise is equal to the square root of the sum of the squares of the individual noise contributions. Noise specifications for ORTEC detectors include the total noise width for the detector and standard ORTEC electronics at a temperature of 21 ±1ºC. These noise widths and actual resolutions therefore can be guaranteed only when the contribution from any other electronics does not exceed that from the appropriate ORTEC electronics.

The noise-broadening effect previously mentioned establishes a lower limit on the energy resolution (FWHM) of any given detector-preamplifier combination. However, factors such as statistical effects, imperfect charge collection, and variations in energy lost in the dead layer of the source and of the detector can cause additional broadening of the peak; their relative contribution is a strong function of the mass of the incident particle. For beta particles, the resolution is nearly always determined solely by the electronic noise broadening. For alpha particles, the ultimate resolution (with no significant contribution from noise) appears to be less than 10 keV. For very heavy ions such as ¹²⁷I, the typical resolution for nonchanneled particles is about 1 MeV.

The pulse rise time associated with any ionizing event is a complex function of the mass, energy, range, and orientation of the ionizing particle; the detector parameters (depletion depth, electric-field strength, diode series resistance, and sensitive area); and the characteristics of the associated electronics. Pulse rise times for typical ORTEC charged-particle detectors range from the order of one nanosecond to tens of nanoseconds. The charge collection time in silicon detectors at room temperature is ~100 ns/mm. In many experiments requiring nanosecond or subnanosecond time  resolution, good energy resolution is also desired, usually resulting in a need for compromises in detector parameters. Consequently, this high-resolution-time requirement, together with all other relevant experimental information, should be specified at time of first inquiry.

For some applications, such as (DE/Dx)(E) mass determinations and telescopic arrays, the energy range of the analyzed particles requires more depth than is provided by a single detector. Two or more detectors can then be combined so that the energy of the particle is totally absorbed in the detectors. The sum of the output pulses from the detectors will be proportional to the energy of the particle. For these applications the effective dead layer is the sum of the front and back dead layers (approximately equal to the electrode thickness) of all the detectors except the last one in the stack. For the last detector, only the front dead layer is considered. (Although all the detectors preceding the last one must be totally depleted, the last one need not be.) Quantitative, independent evaluation of this dead-layer thickness is supplied with each detector.

In applications that require unusually large areas of sensitive depths, it is desirable to connect several detectors in parallel to the same preamplifier. In these circumstances, the total noise contribution to the energy resolution broadening can be determined by the following procedure:

       (1)

The total capacitive load on the preamplifier is obtained by summing the detector capacitances and the stray capacitance:

             (2)

where C_t is the total load, C_d,i is the capacitance of the ith detector, and C_s is the total stray capacitance, including that from cables, connectors, interconnections, etc. The value of C_t and the appropriate curve for preamplifier noise versus input capacitance are used to determine the preamplifier contribution to the noise. The total noise broadening is then obtained from

                            (3)

where N_t is the total noise width and N_A is the preamplifier's contribution to the noise.

Often in low-level counting applications, multiple spectrometers are employed to keep up with large numbers of low-level samples to be counted. Because these are low- or ultra-low-level applications, count rates are extremely low. For this reason, it is possible to employ a multiplexed system, where a gated multiplexer-router is used to send pulses from multiple detectors to separate memory segments in an MCA system. This can lead to substantial cost savings. The more advanced of these systems provide for independent start, stop, and preset of the multiplexed inputs. Examples of such configurations from ORTEC are the 920E NIM MCB which provides 16 inputs and can connect directly to Ethernet, and the OCTÊTE-Plus, which is a complete integrated spectrometer with 8 complete chains including vacuum chamber, and the capacity to connect an additional eight existing vacuum chambers, making a very cost-effective system expansion.

Inadequate thickness uniformity of totally depleted DE detectors has undoubtedly been responsible for many disappointing experiments. A 10-MeV ⁴He particle incident on a 50-µm-thick silicon detector will lose approximately 5.9 MeV in traversing the detector. The rate of energy loss (dE/dx) of the exiting particle, however, will be about 160 keV/µm. This means that a detector thickness variation of 1 µm would cause an energy spread of 160 keV, which is many times greater than the detector resolution for particles that are completely absorbed in the detector. Considerations such as these show that precise control over the thickness uniformity of a device is highly desirable for many experiments. ORTEC uses an exclusive mechanical electrochemical wafer-polishing process that produces damage-free surfaces that are optically flat and parallel. By testing the wafers with optical interference techniques and by profiling the thickness of each wafer with an x-ray transmission technique, ORTEC ensures that each silicon wafer meets stringent thickness-uniformity specifications before being accepted as a planar totally depleted surface barrier detector (D Series). The measured mean detector thickness and uniformity are given on the Quality Assurance Data Sheet that accompanies each D Series detector.

The channeling of ions between crystal planes can produce significant differences in the rate of energy loss (and total range) between channeled and unchanneled ions. For very heavy ions this same effect can produce pronounced differences in the pulse-height linearity and energy resolution. Consequently, the silicon wafers for ORTEC totally-depleted and standard heavy-ion detectors are cut from the parent crystal at an angle that has been carefully selected to minimize channeling effects. Silicon charged-particle detectors that are cut at specific orientations are available on special order.