The utilization of pulsed-laser for single-event effect testing has increased over recent years. While aerospace/defense programs still rely on heavy ion testing to screen and qualify microelectronics, laser testing can be a very useful complementary source. Figure 1 shows an image of a typical laser SEE test setup . A test board is mounted onto a stage, and a lense is directly over the test sample.
Why should I consider laser testing?
Pulsed-laser testing offers advantages that broad beam heavy ion testing cannot accommodate. The laser spot size can be controlled by the experimenter. The spot size can be focused to a diameter of less than 1 μm, which allows the experimenter to selectively irradiate transistors on a die , . This feature makes laser testing a valuable tool for radhard chip designers to identify radiation sensitive transistors and/or functional blocks.
In part due to its ability to target transistors and areas on a die, and due to the relative ease to adjust the energy level, laser testing has also been a useful tool for the characterization of single-event transients (SET) or upsets (SEU). One can scan the laser across the die to identify sensitive locations. Then those sensitive areas can be evaluated in more detail by adjusting the laser energy to estimate the relative sensitivity. Pulsed-laser can also induce single-event latchup (SEL), given sufficient charge collection in the sensitive node. Therefore, laser testing can be used as a SEL screening tool for integrated circuits (IC) with CMOS processes.
A crucial advantage of laser testing is that it can often be a much cheaper alternative to heavy-ion testing. A laser system is much cheaper and easier to develop. There are also complete systems available for purchase. The cost for a one day access at a laser SEE testing facility can be 2 to 4 times cheaper than access at a cyclotron facility.
The setup at a laser SEE testing facility is also less complicated than that at a cyclotron facility. The laser system may consists of an optical table or come in a self-contained workstation. There is no need to route long cables from the user area through to the beam chamber, since laser systems will not emit harmful ionizing radiation. The test equipment and test boards are next to each other, similar to any bench setup. This allows for a easier time to set up features such as adding heating elements to evaluate temperature effects. Experiments to evaluate the effects of elevated temperature on analog SETs have been performed with laser systems .
Which type of laser should I use, SPA or TPA?
Much of the physics and principles of pulsed-laser irradiation have been thoroughly described in publications and presentations , , , . Practically, laser SEE testing can be carried out with either a single-photon absorption (SPA) or a two-photon absorption (TPA) laser. SPA systems vary in wavelength. For example, the SPA SEE setup at the Naval Research Laboratory (NRL) in Washington, DC has a wavelength of 590 nm . The SPA system at the European Aeronautic Defence and Space Company (EADS) has a wavelength of ~1060 nm . The penetration depth is a function of wavelength. So, the 1/e penetration depth is 1.7 μm and 676 μm for a 590 nm and 1060 nm wavelength laser, respectively. Typically, SEE testing is accomplished from the chip's topside. The chip must be delidded first to expose the die surface. In some cases, particularly for highly scaled ICs, the die topside is covered with multiple layers of metalization, which prevents sufficient laser beam to reach the sensitive nodes. This is also a problem for flip-chips, where the die is mounted upside down inside the package. These cases necessitate backside irradiation with a SPA laser of sufficient penetration depth or a TPA laser.
Figure 2 shows the charge density profiles for two cases of SPA lasers, and a TPA laser. Notably, the charge density is highest near the silicon surface for SPA, and attenuates with increasing penetration depth. On the other hand, the charge density is at maximum near the focal point of the TPA laser, some distance beneath the surface. The lack of charge attenuation, relative to the SPA laser, and the ability to reach sensitive nodes without interference from metalization overlayers makes TPA laser testing particularly efficient at SEE characterization that requires precise location and controlled energy deposition.
Limitations of laser testing:
We’ve discussed the many benefits and advantages of laser SEE testing relative to heavy-ion testing. An important drawback is the difficulty to consistently and accurately quantify the equivalent linear energy transfer (LET) from laser charge deposition. The track profiles differ significantly between a laser beam and a heavy ion. The charge track generated by a heavy ion is relatively narrow, and its diameter remains relatively unchanged throughout the length of the ion track. The laser-generated charge track is much more spread out, and its shape varies through the track. Therefore, attempts to match the laser SEE response with that of heavy ion have been inconsistent and not reliable enough. However, recent studies from NRL using axicon lense have shown promising results of laser-induced SEE that mimics heavy ions results . The new system produces a similarly shaped charge track profile as heavy ions. More work needs to be done to realize the results of that study into test ready systems.
An important note is that there are still fundamental differences between charge deposition by laser vs. heavy ion. Heavy ions can cause ionization effects in insulators unlike laser. Therefore, laser cannot cause certain effects like single-event gate rupture or deposit total-ionizing dose. Furthermore, heavy ions can cause nuclear reactions, whereas laser cannot. Heavy ions can collide with the nucleus of elements inside the die and within the package – i.e. silicon, metallization materials, and package plating materials – and generate secondary recoil ions, which can cause additional SEE. Thus, laser testing will not be able to reveal a part’s susceptibility to SEE due to nuclear interactions.
Laser SEE testing can become a valuable resource for a project's radiation testing program. The radiation effects engineer should understand its characteristics to be able to judge when and how to use laser testing.
D. McMorrow, V. Pouget, P. Fouillat, and S. Buchner, "Fundamentals of the Pulsed-Laser Technique for Single-Event Effects Testing," presented at the 2015 International School on the Effects of Radiation on Embedded Systems for Space Applications, Puebla, Mexico.
S. P. Buchner, F. Miller, V. Pouget, and D. McMorrow, “Pulsed-laser testing for single-event effects investigation,” IEEE Trans. Nucl. Sci., vol. 60, no. 3, June 2013, pp. 1852–1875.
D. Chen, S. Buchner, A. Phan, H. S. Kim, A. Sternberg, D. McMorrow, K. LaBel, "The effects of elevated temperature on pulsed-laser-induced single event transients in analog devices", IEEE Trans. Nucl. Sci., vol. 56, no. 6, Dec. 2009, pp. 3138–3144.
V. Pouget, P. Fouillat, D. Lewis, H. Lapuyade, L. Sarger, and F. Roche, “An overview of the applications of the pulsed laser for SEU testing,” in Poc. 6th Int. On-Line Testing Workshop, 2000, pp. 52–57.
D. McMorrow, W. Lotshaw, J. Mellinger, S. Buchner, and R. Pease, “Subbandgap laser-induced single event effects: Carrier generation via two-photon absorption,” IEEE Trans. Nucl. Sci., vol. 49, no. 6, Dec. 2002, pp. 3002–3008.
J. M. Hales, A. Khachatrian, S. P. Buchner, J. Warner, A. Ildefonso, G. N. Tzintzarov, D. Nergui, D. M. Monahan, S. D. LaLumondiere, B. Lotshaw, J. D. Cressler, and D. McMorrow, “New approach for pulsed-laser testing that mimics heavy-ion charge deposition profiles,” to be published in 2020 January issue of IEEE Trans. Nucl. Sci.
Figure 1. Photograph of a pulsed-laser SEE testing setup (After ).
Figure 2. Charge density profiles for SPA lasers on the left, and a TPA laser on the right. (After ).