Introduction
to Radiation Testing
Quick links:
Radiation Environments
Space radiation environments include both natural and man made particles. The natural particles in space include electrons, protons, and heavy ions. The Van Allen belts that are generated by the magnetic field of the earth serve to trap ionizing radiation. These belts protect people on the earth from the radiation found in space, however, they create a problem for space electronics. Electrons and protons get trapped in these belts. The constant solar wind and the sporadic solar flares generated by our sun serve to intensify the radiation within these belts and also irradiate regions outside the belts, adding yet another variable in analyzing the radiation environment. Another source of radiation are Heavy ions, which originate from outside our solar system. These high-energy particles don't get trapped in the Van Allen belts and are generally impervious to shielding.
The number and energies of the electrons, protons and heavy ions vary depending on the altitude, inclination and time. The sun has an eleven-year cycle. For electrons, the worst-case environment is during solar maximum. For trapped protons, the worst-case environment is during solar minimum. For solar flare protons, the worst-case environment is during solar maximum.
Man made environments can also add different types of radiation to the environment. Some spacecraft have nuclear isotopes for propulsion or for use in scientific experiments. These isotopes could produce a wide variety of types of radiation including gamma and x-rays or beta particles. Nuclear weapon detonations on the ground, in the air, or in space, produce bursts of radiation that can adversely affect electronic systems performance. Systems with time-critical functions, such as military communications and weapons systems, are prime examples of applications that may be required to survive, operate through, or operate after such a detonation.
The primary weapon particles emitted are neutrons, electrons, alpha particles, fission fragments, gamma radiation, and x-rays. Electrons, alpha particles, fission fragments, and x-rays are almost totally attenuated by interaction with the atmosphere close to the burst point. Secondary weapon environments are blast and shock, thermal radiation, and electromagnetic pulse (EMP). These secondary environments are caused by the interaction of the primary weapons environments with the atmosphere. In addition, the energy spectrum of the neutrons is modified by interactions with the air and ground. These interactions produce additional gamma rays, which, along with fission-product gamma rays, are detonated as secondary or delayed gamma rays.
When to test The radiation tolerance of commercially available electronic parts is typically unknown. However, to be utilized in a space, electronic parts must meet the radiation requirement of the intended space mission. For commercial space, the typical radiation tolerance specifications that are required include Total ionizing Dose (TID) and Single Event Effects (SEE). Military environments have additional requirements. When this data for a part is unavailable or incomplete, then additional test(s) are required. Even when data is available, additional analysis is crucial to analyze the actual mission requirements and impending effects. Actual mission requirements depend on multiple factors, some of which include the radiation environment, amount of shielding provided by the spacecraft, the function of the part and what the part is doing.
Back to Top
Total Dose Testing
Total Ionizing Dose (TID) in electronics is similar to a sunburn to humans. Total dose is the cumulative ionizing radiation that an electronic device receives over a specified period of time. Like a sunburn to humans, the damage is dependant on the amount of radiation and how long it took to accumulate the total dose. Also like a sunburn, some parts will improve (or heal themselves) after removed from the irradiation source and left biased. This healing process is commonly referred to as "annealing".
Total Dose radiation is measured in rads (radiation absorbed dose). 1 rad(Si) equals 100 ergs of energy per gram when deposited in silicon. Total Dose describes the amount of ionizing radiation from all sources. When the semiconductor material absorbs ionizing radiation, electron-hole pairs are created. In bulk material, the charge can recombine or be swept out, but in impure silicon or dielectrics the mobility of the charge is changed. Charge can become trapped at interfaces or within the bulk material. Radiation-induced surface ionization can generate leakage currents. These excess leakage currents in bipolar devices can diminish the current gain. MOS technology is susceptible to the effects of total dose from excess charge getting trapped in the gate oxide material. Hole mobility is lower than electron mobility, therefore, the net trapped-charge has a positive polarity. This trapped positive charge results in a negative gate-to-source threshold shift in n-channel and p-channel devices. Factors contributing to the severity of the shift are oxide quality, electrical bias during irradiation, and the total ionizing dose level. Degradation of MOS devices depends on both the semiconductor process employed and the application.
Total Dose Effects: Total dose testing involves testing the silicon die for intrinsic parametric and functional degradation at set total dose levels to determine the total dose tolerance of the device. A C0-60 source is the preferred source for total dose testing because it penetrates all electronic packages and uniformly irradiates the die. C0-60 produces a 1.1 and 1.3 MeV Gamma pair. Maxwell tests to MIL-STD-1019.5, which governs the procedures and methodology of the test. The preferred test dose rates are as close to that of the mission's expected dose rate as possible, which frequently means low dose rate testing. For these tests, Maxwell uses a room irradiator which irradiates the parts over several days or weeks.
Maxwell's Total Dose Effects Simulator: Total dose testing is done at Maxwell with either of two Gammacell-220's, or its room Cobalt-60 Irradiator. All parts are tested with a bias board.
Back to Top
Single Event Effects: Satellites in an earth orbit or on an interplanetary mission are bombarded by heavy ions from galactic cosmic rays and Protons from solar flares and those trapped in Earths radiation belts (Van Allen Belts) for orbiting satellites. The natural cosmic ray, heavy ion flux is approximately 100 particles per square centimeter per day. Protons, depending on the location and orbit, can have fluences several orders of magnitude higher than Heavy Ions, although they have LET (described below) values considerably lower then heavy ions. This is important because SEU (described below) rates will vary dramatically depending upon orbit. Heavy Ions or high-energy protons can penetrate into a device leaving a trail of free charges. The amount of energy deposited in a device is called the LET (Linear Energy Transfer) and is measured in MeV/mg/cm2 (energy per mass density). This sudden generation of free charges can create several types of phenomena known as Single Event Effects (SEE) or Single Event Phenomena (SEP).
SEE can include several specific types of effects in semiconductor devices including Single Event Upset (SEU), Single Event Latchup (SEL), Single Event Transients (SET), Single Event Burn Out (SEBO), and Single Event Gate Rupture (SEGR).
Single event upset (SEU) refers to the temporary upset of a single transistor element or output, which is caused by excess free charge changing the digital voltage state. Proton induced nuclear reactions in silicon can result in heavy recoil nuclei capable of single event upsets. About 1 proton in 105 will undergo a nuclear reaction capable of SEU. Because of the low LET values required to upset most modern commercial memory devices, the more numerous protons are the major cause of most SEUs in orbits with trapped protons.
Single event latchup (SEL) is a major loss of device functionality due to the occurrence of a single radiation event, which induces high a current state in a portion of the device. This state may or may not cause permanent damage to the device, but requires the removal of power from the device in order to return to normal operation. Usually SELs are caused by heavy ions due to the high LET levels required to cause a part to go into a Latchup state.
Single event burnout (SEBO) refers to a single ion induced condition in power MOSFETs, which in turn results in device destruction due to the activation of a high current state.
Single event gate rupture (SEGR) refers to a single ion induced condition in power MOSFETs, which results in a conducting path through the gate oxide. A hard fault often called a hard error, is a permanent, unalterable change of operation that is typically associated with permanent damage to one or more of the device elements (e.g. gate oxide) comprising the affected device.
The end product of Single Event Testing for SEU and SEL, is a plot of the SEE cross section (the number of events per unit fluence) as a function of ion LET (Linear Energy Transfer, or ionizing energy deposited along the ion's path through the semiconductor). This plot should extend from threshold LET all the way to saturation (i.e., the saturated cross section). This data can be combined with the heavy ion or proton environment for the intended orbit to estimate the expected SEE rate of the device. For SEBO and SEGR, the end product is a detailed explanation of device response for each tested condition (ion species and energy, temperature, bias voltage, etc.). Ion species, energy and angle of incidence should be recorded for all tests, in addition to ion LET.
Single Event Effects (SEE) Testing
Single Event Testing uses terrestrial heavy ion and proton sources to simulate the environment in space. Testing is usually done with bare exposed die and the part is tested for Latchup over a range of LET values as well as SEUs. For SEL, the supply current is measured. If the die goes into a high current state, the part is turned off and the I/Os crowbarred to ground. For SEUs, depending on the complexity of the part, a signal is inputted into the devices and then the output signal is compared to the input signal to evaluate for errors. Testing in the single event environment is typically done with an ion accelerator such as a cyclotron or Van de Graaff. Preliminary screening is sometimes done using radioisotope sources such as Californium-252 or relativistic heavy Ion sources. Maxwell Technologies uses the Tandem Van de Graaff Accelerator Facility at the Brookhaven National Laboratory as one of its choices for performing SEE testing. These tandem accelerators routinely accelerate 40 different elements including 60 different isotopes over the full range of the periodic table. For Protons Indiana University and the University of California Davis Cyclotrons are used.
Back to Top
Neutron Effects: Neutron fluence causes permanent damage to electronics. The primary physical effect of neutrons bombarding semiconductor material is the formation of displacement defects within the crystal lattice structure. Such defects effectively decrease the mean free path, and thus shorten the recombination time in the semiconductor material. Exposure to neutron fluence, measured in neutrons per square centimeter, degrades the operating parameters of semiconductor devices. For example, current gain decreases, base-to-emitter voltage drops increase, and storage time increases. Minority-carrier-devices, like bipolar junction transistors, are the most sensitive to neutrons because the increase in recombination sites decreases the number of minority carriers for device operation. Majority carrier devices will show signs of degrading when the damage from neutrons is enough to reduce the number of majority carriers. For example, MOS field-effect-transistors (FETs) show increases in drain-to-source on-resistance. This parameter varies the most in those MOS devices with the highest voltage ratings.
Neutron Testing
Neutron testing usually consists of parametric measurements made before and after exposure to one or more cumulative levels of neutron fluence. To determine the effects of the neutron exposure(s), one simply compares the measured parameters at each of the radiation levels. For neutron testing, the parts to be irradiated are irradiated passively (i.e., no bias on the parts).
Neutron Effects Simulators: Neutron effects are usually simulated with Fast Burst Reactors (FBR) or TRIGA Reactors. Maxwell Technologies has a teaming arrangement with several FBR facilities. Fast Burst Reactors simulate, as closely as possible, the neutron radiation environment produced by a fission weapon. In some cases, Californium-252 sources are also used to simulate the neutron environment. Back to Top
Dose Rate Effects: The ionizing dose rate is the rate at which gamma rays and x-rays are deposited in semiconductor material. This deposition is measured in rads in silicon per second (rad(Si)/sec). When a nuclear weapon detonates, the highest dose rate is usually delivered within a narrow pulse width (in nanoseconds). This environment is also called the prompt pulse or prompt gamma environment. Ionizing dose rate radiation induces excess electron-hole pairs into semiconductor materials. A reverse-bias semiconductor junction will, by drift and diffusion, sweep the excess charge across the junction. A reverse leakage current, also called a photocurrent, will persist anywhere from a few nanoseconds to a few microseconds, depending on the dose rate and the junction characteristics. A junction with the largest current rating usually produces the most photocurrent. In discrete transistors, these photocurrents can be large enough to disrupt the operation of a device. As the dose rate increases the size of the photocurrents increase, and because of secondary effects, can burn out a transistor.
The burnout of discrete components due to photocurrents can be prevented by using resistors and inductors to current-limit all semiconductor paths to ground. Upset and burnout are even more complicated in ICs. With four-layer PNPN structures, the photocurrents can cause latchup similar to the "on" condition of a silicon-controlled rectifier (SCR). Once the parasitic SCR latches up, the device energy must be limited to prevent burnout. This is often done in an IC by removing power to the device before the excess current can damage semiconductor junctions.
Dose Rate Testing Dose rate testing is typically done with the device under test (DUT), biased such that the power supply and input bias levels are at the worst-case rated voltages and all unused inputs are either grounded or tied to the supply voltage. During the radiation pulse, transient changes in output voltage and current are recorded. The dose rate of the pulse, which causes a voltage transient equal to the specified value, is the upset level.
Dose Rate Simulator: Maxwell Technologies uses an electron linear accelerator (Linac) for dose rate testing. This L-band machine provides pulses of high-energy electrons, which can be varied in width from 20ns to 7us. Energy varies with beam loading, but is generally in the range of 12 to 17 MeV. Variation of dose rate is accomplished by controlling the number of electrons injected for each pulse. The direct electron beam is used for dose rate tests. A 1/8-inch aluminum scatter plate at the exit window of the Linac serves to spread and homogenize the beam to avoid "hot spots" in the radiation field.
Back to Top
Back to Radiation Testing
|
|
 |
|
|
