Search

By scanning the charge-deposition profile produced by a pulsed laser throughout a device, the spatially dependent charge-collection efficiency (CCE) can be determined. This is demonstrated by extracting the depth-dependent CCE in two photodiodes. The resulting collection efficiency curves are found to be consistent with their device structures and the expected charge-collection mechanisms. By applying the efficiency curves to the charge-deposition curves for both heavy ions and pulsed X-rays, the calculated collected charge values show very good agreement with experimental results. This suggests that the profiles extracted using the laser can help predict charge-collection data from other excitation sources and could improve modeling efforts by determining sensitive volumes. [ABSTRACT FROM AUTHOR]

Heavy-ion, focused X-ray, and pulsed laser single-event transient (SET) experiments are performed on a silicon epitaxial diode. Collected charge, transient rise times, and transient fall times are calculated and compared between different sources. The transient shape characteristics depend on the source (ion, X-ray, or laser), even when similar amounts of charge are generated. The observed differences are examined and explained in terms of basic charge collection mechanisms. [ABSTRACT FROM AUTHOR]

Optical single-event transients (OSETs) were measured for the first time in integrated silicon-photonic waveguides. A custom test fixture and novel experimental setup were used at the U.S. Naval Research Laboratory to induce a dense cloud of electron–hole pairs (EHPs) in a waveguide by the two-photon absorption process. Experimental data showed that the fractional (or percent) extinction of the optical power in the waveguide due to an OSET is dependent on the number of injected EHPs. This fractional extinction was also shown to be constant regardless of the optical power level in the waveguide. These OSETs are a direct result of a reduction in the optical transmission ratio due to transient free-carrier absorption. The peak of the largest measured OSET degraded the transmission ratio by 15%, which, according to simulations, corresponds to roughly $9\times 10^{19}$ cm−3 peak EHP generation. Simulation results suggest that the EHP density levels generated by heavy ions in space could potentially cause up to 100% fractional extinction of light in the waveguide, that is, total loss of the optical signal. A simulation technique to predict the OSET effects in any arbitrary photonic circuit is shown. Here, an electrooptic modulator was used as an example. The results of this study pose concerns for use of integrated silicon photonics for radiation-intensive applications. [ABSTRACT FROM AUTHOR]

A comparison of heavy-ion-induced single-event transients (SETs) in silicon–germanium heterojunction bipolar transistors (SiGe HBTs) fabricated on both bulk and silicon-on-insulator (SOI) substrates is presented. Experimental heavy-ion data show a reduction of sensitive volume of ≈ 350 × for the SiGe HBTs fabricated on SOI compared to those on a bulk substrate. Furthermore, the results indicate that the sensitive volume of the SOI SiGe HBTs is confined to the intrinsic device. These results have been confirmed using pulsed laser data and TCAD simulations. Limiting the charge collection depth of SiGe HBTs by fabricating them on SOI facilitates correlation of ion and laser results. The improved ease of correlation is a result of suppressing diffusive charge collection mechanisms that are difficult to replicate with the charge deposition profile generated by two-photon absorption using a tightly focused laser beam. However, a potential issue with pulsed laser testing is determining the loss of the pulse energy through the buried oxide, which requires detailed knowledge of the semiconductor platform. Overall, the reduction of sensitive volume for the SOI SiGe HBTs compared to those on a bulk substrate is a clear advantage from a single-event upset vulnerability perspective. [ABSTRACT FROM AUTHOR]

A novel approach for two-photon absorption (TPA) pulsed-laser testing produces extended charge deposition profiles that are analogous to those produced by heavy ions. In this approach, which utilizes an axicon rather than a spherical lens, the conventional Gaussian beam is replaced by a quasi-Bessel beam. The key feature of a quasi-Bessel beam, relevant to pulsed-laser single-event effects (PL-SEE) studies, is that its radial size (1~μm) remains constant over the axial length of the beam (several 100 s of μm) resulting in a charge-deposition profile that more closely mimics that produced by a heavy ion. This elongated charge distribution directly lends itself to a simple and intuitive description in terms of linear-energy transfer (LET), facilitating laser/ion correlation studies. Experimental results comparing a conventional TPA PL-SEE focusing geometry and the axicon focusing geometry are presented for three different test vehicles: a 45-nm silicon-on-insulator (SOI) nFET, a bulk silicon photodiode, and an LM124 operational amplifier. By using the axicon approach, strong laser/ion correlation is observed in the silicon photodiode for single-event transients (SET) across a wide range of LETs. The correlation is achieved by “tuning” the laser to the desired LET rather than relying on preexisting heavy-ion data, and therefore, this approach holds promise as a predictive tool. [ABSTRACT FROM AUTHOR]

A sensitive volume is developed using pulsed laser-induced collected charge for two bias conditions in an epitaxial silicon diode. These sensitive volumes show good agreement with the experimental two-photon absorption laser-induced collected charge at a variety of focal positions and pulse energies. When compared to ion-induced collected charge, the laser-based sensitive volume overpredicts the experimental collected charge at low bias and agrees at high bias. A sensitive volume based on ion-induced collected charge adequately describes the ion experimental results at both biases. Differences in the amount of potential modulation explain the differences between the ion- and laser-based sensitive volumes at the lower bias. Truncation of potential modulation by the highly doped substrate, at the higher bias, results in similar sensitive volumes. [ABSTRACT FROM AUTHOR]

Using a laser-equivalent linear energy transfer (LET) approach, strong correlation is found for collected charge following heavy-ion and pulsed-laser excitation in a bulk diode device. By significantly modifying the size of the charge distribution generated by one- and two-photon absorption compared to the depth of the sensitive region, charge deposition in a large variety of device technologies is mimicked. Nearly one-to-one correlation is found with heavy-ion charge collection, independent of the laser focusing geometry or magnitude of LET. Finally, this approach is applied to the previous studies where empirical laser-ion correlations on highly scaled technologies were proposed. These correlations are reproduced through direct calculation of the laser-equivalent LET, and outstanding observations are explained despite the sensitive areas of these devices being up to 1000 times smaller than the laser-deposited charge profile. In total, these studies suggest that this quantitative methodology for laser-ion correlation is broadly applicable for a range of complex device technologies. [ABSTRACT FROM AUTHOR]

This correspondence comments on the correlation between measured and simulated charge collection in two diode structures presented in [1]. In [1], results from pulsed-laser-induced charge deposition via two-photon absorption revealed reasonable correlation between experimental data and simulations in some aspects but showed lack of correlation in terms of the measured and predicted magnitudes of the collected charge. The source of this poor agreement has recently been determined, and is attributed to an incorrect pulse energy calibration used for those experiments. With the proper energy calibration applied, the agreement between simulation and experiment is nearly quantitative for all observables in both devices. These results significantly strengthen the validity of the nonlinear optical beam propagation method numerical approach for accurately predicting pulsed-laser-induced charge deposition. [ABSTRACT FROM PUBLISHER]

Simplified analytic expressions are presented for calculating pulsed-laser induced charge generation in devices with collection volumes of any size. The expressions for charge generation by single-photon absorption for various limiting cases are applied to charge-collection measurements for a bulk Si photodiode, and single-event upsets in a 90-nm CMOS, silicon on insulator (SOI), static random-access memory (SRAM). Strong quantitative agreement between the simplified expressions, the full numerical calculations, and the experimental data serves to validate the analytical expressions. Furthermore, despite the complexity of determining charge deposited via two-photon absorption, the analytic expressions exhibit excellent agreement with full charge-generation simulations utilizing nonlinear-optical beam propagation software. These expressions therefore, represent a convenient and accurate method for estimating pulsed-laser induced charge generation in specified volumes, and are particularly relevant for two-photon absorption, which previously has required the use of complex numerical approaches. [ABSTRACT FROM AUTHOR]

The impact of semiconductor process scaling on the overall transient response of SiGe BiCMOS platforms is investigated. Pulsed-laser two-photon absorption (TPA) and heavy-ion broad-beam testing of SiGe HBT device and digital test structures across several generations of SiGe technologies are utilized to investigate the potential impacts of semiconductor process scaling (e.g., lateral/vertical scaling, changes in doping, Ge content, etc.) on the overall transient shape, magnitude, and duration. Technology scaling is shown to increase the single-event effect (SEE) sensitivity of SiGe HBTs (i.e., elevated collected charge, where 1st Gen. QC <3 rd Gen. QC <4 th Gen. QC ). Modern third-generation and fourth-generation devices under a forward-active bias (i.e., forward-biased EB junction, reverse-biased CB junction) exhibit a large diffusive transient component between the emitter and collector terminals, driving an elevation in collected charge. 3-D TCAD modeling is utilized to understand the fundamental transient mechanisms and assess the primary scaling factors affecting SEE sensitivity. Ion-strike simulations show that bulk traps can substantially enhance the charge collection mechanisms within these devices. These results suggest that SiGe technology scaling may have a strong impact on the radiation-induced transient response of SiGe HBTs, with future SiGe technology generations potentially exhibiting increased sensitivities to single-event effects. [ABSTRACT FROM PUBLISHER]

Carrier-density distributions generated via two-photon absorption from pulsed laser excitation are simulated using nonlinear-optical beam propagation software. These simulated carrier-density distributions are used to calculate depth profiles of the integrated collected charge using a rectangular-parallel-piped approach for two silicon diodes of different structure. Using a set of proposed correlation metrics, the resulting simulated charge collection profiles are found to exhibit good agreement with measured transient charge-collection data for most, but not all of the metrics. The physical phenomena underlying the correlation metrics are discussed in detail. The remaining discrepancies that exist between the simulated and experimental results are addressed and their potential causes are detailed. [ABSTRACT FROM PUBLISHER]

Numerical simulation software is used to calculate quantitatively the two-photon absorption-induced carrier-density distributions generated under conditions that are experimentally relevant for single-event effects studies. The results provide valuable insight into how the magnitudes and shapes of the charge carrier distributions evolve over a large range of experimental conditions and the impacts this has for different device geometries. Furthermore, values of integrated charge are determined that can be more directly correlated with experimental observables. [ABSTRACT FROM AUTHOR]

Nonlinear beam propagation software is used to calculate quantitatively the two-photon absorption (TPA)-induced charge-density profiles generated in silicon by focused femtosecond laser pulses under conditions that are experimentally relevant for single-event effects studies. The described approach permits simulation and prediction of the impact of various optical nonlinearities on the beam propagation through, and generation of free carriers in silicon for F/#s approaching one. It is found that, even at moderate incident laser pulse energies, the nonlinear-optical processes of nonlinear refraction, free-carrier absorption, and free-carrier refraction all contribute, and must be considered in describing the TPA-induced charge generation in silicon. Free-carrier refraction is found to play the dominant role in distorting the charge density profile at larger pulse energies. The simulation results are validated with experimentally measured beam sizes for different focusing conditions. [ABSTRACT FROM PUBLISHER]