Ultrashort-Pulse Lasers for Space Applications
Ultrashort-pulse lasers exhibit exotic, fantastic characteristics. Aerospace scientists and engineers are researching diverse applications that can take advantage of the broad spectrum and high power delivered by these devices.
Ultrashort-pulse lasers (USPLs) are defined by the duration of the pulses they emit, which range from a few femtoseconds (10−15 of a second) to a few picoseconds (10−12 of a second). Because of their pulse duration, USPLs have two novel characteristics: the laser spectrum they produce is broad (up to or even greater than 100 nanometers), and the power in the pulses can be high.
The broad laser spectrum is a consequence of the Fourier-transform relationship between time and frequency, and each pulse results from the coherent superposition of many frequencies. The high-peak power results from temporal confinement of the laser energy. A laser operating with a 50-femtosecond pulse and a 100-megahertz pulse rate will have a peak power that is 200,000 times higher than a continuous-wave laser operating at the same average power.
The Aerospace Corporation is researching applications that have current or anticipated high value to military and civilian aerospace interests. For example, it is easy to envision USPLs as high-data-rate transmitters in free-space optical communications. Using 50-femtosecond pulses, a message containing 10,000 bits could occupy as little as one nanosecond in time. Alternately, even a small amount of energy in an ultrashort pulse generates high power. One millijoule in a 30-femtosecond pulse represents 30 gigawatts of instantaneous power. Because of these characteristics, USPLs could potentially support a variety of military and civilian applications, such as the transmission of large amounts of data or high power to distant locations.
The penetration of USPL technology into military applications such as hyperspectral sensing or secure optical communications will be, to a significant extent, determined by the capacity to control how ultrashort pulses interact with their operational environment, so that the desired functionalities can be realized where and when they are wanted. It is not clear how or if such control may be achieved outside the laboratory; but in well-controlled laboratory environments, the use of USPLs has been growing. In particular, these devices have enabled diagnostic tools for quantitative characterization of material properties, performance testing and calibration of prototype devices, evaluation of device vulnerabilities, and exploration of new device architectures and operating schemes. USPLs are certainly exhibiting the potential to serve well-established and emerging space technologies.
Radiation Hardness for Space Electronics
In the early 1990s, microscopy using USPLs was investigated as a means to simulate the effects of space radiation on semiconductor devices in satellite payloads. Space radiation can give rise to so-called single-event effects, which can degrade the on-orbit performance of electronic devices or render them inoperable. Single-event effects can be triggered when an ionized particle penetrates an electronic component, leaving an ionized trail that can cause current or voltage transients that disrupt normal operation. Scientists have devoted significant effort to evaluating the susceptibility of devices to radiation-induced anomalies and determining their suitability for space missions.
Aerospace and the U.S. Naval Research Laboratory pioneered the use of USPLs to simulate radiation-induced current and voltage transients by generating conduction-band carriers from the absorption of photons having energy equal to or greater than the semiconductor bandgap. The bandgap defines the energy required to elevate electrons from the low-energy valence band, in which they are tightly bound to their nucleus, to the higher-energy conduction band, in which they are mobile, like electrons in a metal. The technique is a variation of laser microscopy using an integrated circuit mounted so that the pinout signals can be monitored by equipment that records the device’s response to laser illumination.
The development of laser techniques for testing single-event effects has significantly expanded and improved the capabilities needed to evaluate payload electronics for space missions. Previously, such evaluations required testing at accelerator facilities, where parts can be subjected to regulated exposures of particles known to be prevalent on common orbits.
While accelerator testing establishes a “gold standard” to qualify parts, it is limited by the type of information it can provide and by its cost and availability. Accelerators are fairly extravagant multiple-user facilities that must be scheduled months in advance and can impose many usage restrictions. They mostly provide information about whole-device susceptibility to specific particles and energies. The spatial resolution needed to identify sensitive microscopic substructures is difficult to achieve because of the large cross section and relatively low flux density of particle beams. Accelerator testing is also typically destructive—the device being tested often fails irreversibly from catastrophic material damage, making iterative and recursive tests unlikely or impossible. Such recursive tests are needed to analyze failure mechanisms and support quick design modifications that explore mitigation alternatives.
Although USPL testing of single-event effects cannot be used as a qualification standard, it is highly valued as a screening tool prior to accelerator testing. It significantly reduces the time and costs of final qualification and has several diagnostic advantages over particle testing. For example, the spatial resolution of laser testing is on the order of a fraction of the laser wavelength, enabling raster-scan images of susceptible spots. The generation of carriers via laser excitation is fast relative to their lifetimes and closely matches the temporal characteristics of a particle strike—but with precise information on event initiation (the timing of accelerator-induced events is often chaotic and not well correlated with any clock). Unlike particle testing, laser testing can be completely nondestructive—experiments can be recursive with little or no analysis latency. This testing of mechanisms at specific circuit nodes can be investigated and characterized to provide empirical data that can be used to benchmark device performance simulations and support fast design-and-fabrication cycles that explore improvements to the radiation hardness of devices.
The laser testing techniques developed in parallel at The Aerospace Corporation and the Naval Research Laboratory used picosecond dye-laser pulses at wavelengths between 600 and 800 nanometers, which are above bandgap (i.e., they have a photon energy level higher than the energy gap) in silicon and gallium arsenide and are therefore suitable for testing most integrated circuit technologies. Laser testing at wavelengths shorter than bandgap is predicated on linear optical absorption and requires an unobstructed line of sight to the device or circuit node being tested. In practice, the line of sight is often obstructed by the ubiquitous metal interconnections of modern integrated circuits as well as by the device mounting and packaging, which prevents above-bandgap light from reaching the device. Despite this limitation, above-bandgap laser testing has established the correspondence between laser and particle test results, thereby validating this technique.
The Next Generation
In addition to the line-of-sight problem, above-bandgap testing only probes near the surface—carriers are typically deposited just a few microns deep. Because modern micro- and nanoelectronic fabrication technology is moving rapidly toward higher feature densities and multiple layers in monolithic and heterogeneous architectures, buried circuit nodes may not be accessible or resolvable by above-bandgap, linear laser-testing techniques. This problem may, however, be resolved by invoking the more extravagant performance properties of a newer generation of USPLs.
At the same time that picosecond dye-lasers were being applied to the development of techniques for testing single-event effects, a generational change in USPL technology was occurring, represented by the emergence of solid-state laser materials such as titanium-doped sapphire. These materials supported dramatically shorter pulses (in the range of 5 to 100 femtoseconds), while a new amplification scheme allowed for much higher pulse energy at much higher efficiency than dye-laser systems. The new femtosecond solid-state laser technology supports wavelength and energy conversion techniques that enable relatively simple access to wavelength sources between about 500 and 3000 nanometers, as well as new sources of coherent ultraviolet and far-infrared radiation, x-rays, and pulsed electron beams.
The ready availability of tunable femtosecond laser sources in the shortwave infrared below the bandgap of silicon permitted the development of nonlinear optical techniques for microelectronics testing and measurement. In the late 1990s, Bell Laboratories introduced two-photon optical beam-induced current as a functional imaging technique. In 2002, laser single-event effects techniques based on nonlinear absorption were developed at the Naval Research Laboratory. The critical distinction between the linear and nonlinear techniques is the mechanism of material interaction with the laser pulse. Semiconductors are essentially transparent to wavelengths below their bandgap, but can be induced to absorb two (or more) subbandgap photons simultaneously if the pulse irradiance is high and the sum of their energy exceeds the bandgap.
This can be easily exploited with USPLs equipped with a fast convergence objective. In fact, two-photon absorption and the resultant generation of conduction-band carriers occurs only at the beam focus where the irradiance is high. Using nonlinear absorption, USPL testing can overcome line-of-sight limitations by addressing the circuit nodes of a device through its substrate. Additionally, a three-dimensional capability can be obtained by controlling the depth of the beam focus in the part, with the result being volumetric images of single-event effect susceptibility or other performance attributes.
In 1998, before the Naval Research Laboratory started developing nonlinear techniques to solve the line-of-sight problem, Aerospace launched an alternative approach based on work pioneered at the University of California, San Diego. In this approach, high-energy femtosecond lasers were used to generate hard x-ray pulses at photon energies sufficient to penetrate the interconnection metallizations of integrated microelectronics. The technique would use x-ray photons capable of penetrating the obstructions and launching conduction-band carriers in the semiconductor, which is in contrast to nonlinear absorption techniques where the obstruction was evaded. The USPL mediates x-ray generation through an energetic plasma initiated by laser ablation of a metal target. Hot electrons from the laser-induced plasma interact with the target to produce K-band x-rays. If sufficient x-rays can be generated, collected, and focused, a laser capability for x-ray testing of single-event effects could be developed.
Since 2005, Aerospace’s research and development in USPL applications has been balanced to include nonlinear optical and x-ray approaches to single-event effects testing and has also broadened to address wider applications of core techniques from laser spectroscopy to testing and measurement, as well as reliability assessments and performance analyses of integrated electronic and photonic device technologies. Aerospace has also closely monitored research and development in the external peer community, where USPL technology has been actively explored for a much wider range of applications, some of which would take advantage of the extreme bandwidth of these lasers. Examples include high-capacity and secure optical communications, all-optical time and frequency standards that could be used at the core of communication and navigation systems, and adaptive ultrawideband waveform generators and signal analyzers. Other applications being investigated in the United States and abroad involve exercising the high power and irradiance of USPLs in a variety of active remote-sensing or situational-awareness schemes.
Because many of these techniques are relatively new and unoptimized and have potentially high value as performance and reliability diagnostics for emerging microelectronic and integrated photonic technologies, the scope of Aerospace research has expanded to address the physics of these techniques and the investigation of their diagnostic potential. For example, the long-term value of nonlinear optical techniques in testing single-event effects and more general device performance diagnostics will be determined by the ability to control the way that ultrashort laser pulses propagate through and interact with the materials and structures of devices undergoing testing.
This kind of control requires a detailed knowledge of the linear and nonlinear optical properties of device materials and their responses to excitation by ultrashort laser pulses. While approximate information from theoretical models and previous experiments is available, updated experimental measurements are needed to advance the theory and support the development of practical diagnostic tools. Aerospace is addressing these needs with quantitative USPL measurements of nonlinear refraction and nonlinear absorption in elemental and compound semiconductors using techniques adapted from peer literature, as these properties determine how a focused USPL beam will propagate through a device structure and where carriers will be generated.
USPLs are also ideally suited to time-resolved probes of other photonic material properties important to new technology development. The lifetimes and relaxation dynamics of conduction-band carriers establish fundamental limits on the performance of semiconductor electronic and photonic devices. Carrier lifetimes are quantified by the decay of the luminescence emitted when electrons and holes recombine across the bandgap (an interband transition), while intraband carrier relaxation can be probed by time-resolved absorption spectroscopy. USPLs provide a means of instantly generating excited electron-hole pairs and a “clock” to measure the recombination rate and evolution of the excited electron-hole population.
Two salient examples of photonic technologies that immediately benefit from carrier lifetime diagnostics are diode lasers and solar photovoltaic cells. Semiconductor quantum-well lasers are at the core of any solid-state laser system that will be used for on-orbit optical communications or satellite-based active remote sensing. These are in continual development to improve their efficiency and noise properties, and to establish new wavelengths of operation. The materials and junction structures of photovoltaic cells are similarly in constant development. One path being explored involves the use of semiconductor quantum dots, in which the nanoscale size of the material structure (the “dot”) alters the electronic structure and behavior of the material. Quantum-dot solar cells may enhance the action spectrum and efficiency of current multijunction devices, which could significantly affect the power budgets of all satellite programs. In both of these technologies, carrier lifetimes are a critical performance indicator.
Aerospace has built a time-resolved luminescence spectrometer capable of measuring lifetimes as short as tens of picoseconds and as long as a microsecond. Additionally, a time-resolved absorption probe for measuring intraband carrier dynamics has also been configured and tested. Photovoltaic devices enhanced by quantum dots represent a technology development topic for which USPL-based diagnostics provide critical support. For example, the phenomenon of carrier multiplication in quantum dots could significantly enhance solar-cell efficiency if it can be controlled in a device configuration, but there is still some controversy about the reality and efficacy of carrier multiplication.
Carrier multiplication in quantum dots is believed to result from the optical excitation of “hot” carriers into the upper conduction-band levels by photons having energy at least greater than twice the bandgap. Such highly excited carriers can “cool” by a process in which excess conduction band carrier energy greater than the bandgap energy is coupled to the generation of additional electron-hole pairs, with the result that multiple conduction-band carriers are generated and available for electrical work from the absorption of a single photon. Some recent measurements suggest that this process is much more efficient in quantum dots than in bulk semiconductor material; however, these results are controversial, and more research is needed to validate these claims.
Aerospace is working with the National Renewable Energy Laboratory to make quantitative measurements of carrier dynamics and yields in quantum dots excited by photons greater than twice the bandgap to resolve the uncertainties surrounding carrier multiplication and establish screening diagnostics for feedback to material optimization and performance evaluations of prototype devices. The Rochester Institute of Technology is designing and fabricating multijunction solar cells containing quantum-dot layer structures intended to extend the absorption spectrum of a junction and increase the use of the solar spectrum. In support of the Rochester program, Aerospace will conduct USPL studies of the optical and photonic properties of quantum-dot test structures and junctions.
Aerospace is also investigating the use of USPLs to evaluate the spatial resolution of sensors in focal-plane arrays. This performance parameter is measured by the modulation-transfer function, which quantifies the spatial frequencies that the focal-plane sensors can distinguish by measuring the array’s response to laser illumination. For visible and near-infrared focal-plane arrays, visible continuous-wave lasers at a set of discrete above-bandgap wavelengths are typically used to measure a single pixel point-spread function, which can be used to generate the two-dimensional modulation-transfer function for the array. The lasers can be focused to a spot smaller than the pixel size, which allows electronic structures within the pixel to be correlated to the measured point-spread and modulation-transfer functions. By taking multiple measurements with different lasers, researchers can build up the sensor performance over the entire wavelength range of operation.
This technique has some disadvantages that can compromise the modulation-transfer function diagnostic. For example, the spot size, particularly at long wavelengths, begins to approach the size of a single pixel. Additionally, the above-bandgap excitation cannot probe the pixel response in three dimensions, which could be important in CMOS (complementary metal-oxide semiconductor) active pixel designs where the pixels can have complex, multilayered layouts with very high aspect ratios.
One way of potentially resolving these problems is to use the nonlinear optical technique for carrier injection. For USPL sources, the area over which carriers are deposited by nonlinear absorption contracts according to the order of the nonlinear interaction, and the laser focusing geometry can vary the depth at which carriers are generated in a pixel. For example, in the case of two-photon absorption, carrier injection is proportional to the square of the beam irradiance, and the effective beam area for carrier injection should contract by relative to that of linear absorption. Furthermore, carrier injection occurs only at the beam focus where the irradiance is sufficient for multiphoton absorption. The development of the below-bandgap modulation-transfer function diagnostic shares nearly all the optimization criteria of nonlinear absorption laser single-event effects testing. However, there is a need for detailed information on the nonlinear optical properties of the materials from which these devices are made, along with advanced laser microscopy techniques to exploit these properties—and, in many cases, a way to manipulate and control the laser-material interactions.
There is a distinct advantage in being able to manipulate the temporal shape of the laser pulse or the order of wavelengths within the pulse, and a pulse-shaper subsystem capable of these manipulations is in development. Because USPLs possess a wavelength bandwidth determined by the Fourier transform of the temporal pulse, a 30-femtosecond pulse at a center wavelength of 800 nanometers will have approximately 35 nanometers of bandwidth. The pulse-shaper decomposes the bandwidth of a single USPL into discrete spectral “bins” that can be independently delayed or attenuated, and then recombines these bins to generate a new single pulse with a modified temporal shape, a specific time-ordering of wavelengths, or a series of pulses (a pulse “burst”). The pulse-shaper makes it possible to investigate alternative ways of presenting USPL energy to a material for the purpose of optimizing a desired effect or investigating the material’s response to USPL illumination. For example, in the nonlinear optical techniques under development for testing microelectronics and measuring the modulation transfer function of focal-plane arrays, temporal shaping of ultrashort pulses may enable the generation of conduction-band carriers in areas significantly smaller than the diffraction-limited spot size of the laser and allow for enhanced control over carrier deposition depth.
Another approach to improving the spatial resolution of a laser probe involves controlling the spatial characteristics of the laser beam through the curvature of the pulse spatial wavefront. Aerospace is investigating the use of adaptive optics for this purpose. Control over the effective area in which a USPL interacts with a material or device structure is critically important to the diagnostic utility of USPLs in the emerging nanoscale material and device technologies, where the length and separation of device structures is approaching a few tens of nanometers, and for which there is a shortage of nondestructive measurement tools capable of resolving the structures or their time/frequency performance characteristics.
Several other areas of research on USPL applications exist where control over the spatiotemporal formatting of the laser is known to be critical and where adaptive control of such formatting based on active feedback may dramatically enhance capability. In the x-ray technique discussed earlier for testing single-event effects, the results have shown that enhancement of x-ray yield is a complicated function of experimental and environmental parameters that significantly affects the evolution of pulse characteristics along the propagation path. A USPL with pulse-shaping capability can “precompensate” the laser pulse to invert temporal distortions and ensure that the highest-peak power is delivered where it can have its greatest effect—in this case, the highest x-ray yield. Additionally, the temporal format in which the laser energy is delivered to the metal target may have a significant effect on the generation of a plasma with optimal thermophysical characteristics for x-ray generation.
Laser remote sensing and free-space optical communications using USPLs are also likely to benefit from adaptive spatiotemporal formatting. These potential space applications receive a lot of attention, in part because USPLs have been shown capable of some unique signal-propagation characteristics. For example, parameter regimes have been identified in which the laser beam propagates without spatial diffraction (beam expansion) and in which the beam is able to penetrate obscurants such as clouds, fog, and suspended particles.
The prospects of USPL-based hyperspectral schemes for meteorological lidar and active remote sensing for chemical, biological, and hazardous-material detection or target characterization will be significantly determined by the ability (or lack thereof) to manipulate and control these propagation characteristics, which appear to arise from a balance of linear and nonlinear interactions with the atmosphere along the beam propagation path. Because these interactions are not fully understood, the fragility of this balance and the ability to control the interaction of the laser pulses with the atmosphere and the target in remote-sensing schemes represents a critical risk element that will have to be retired to realize a practical technology. The ability to control these interactions and accommodate fluctuating atmospheric conditions via adaptive pulse-shaping would have utility and benefit, but necessarily requires a detailed understanding of the underlying physics that is unlikely to emerge without a continuing research commitment.
The Aerospace research program in USPL applications is an interdepartmental team effort. The contributions and participation of David Cardoza, Kevin Gaab, Nathan Wells, Stephen LaLumondiere, and Paul Belden have been critical to the establishment and progress of the program. The author also wishes to acknowledge the support and encouragement of Steven Beck, Dean Marvin, Steven Moss, Gina Galasso, Bernardo Jaduszliwer, Rami Razouk, and Sherrie Zacharius.
- M. Beard and R. Ellingson, “Quantum Dot Solar Photovoltaics. Multiple Exciton Generation in Semiconductor Nanocrystals: Toward Efficient Solar Energy Conversion,” Laser and Photonics Reviews, Vol. 2, pp. 377–399 (2008).
- T. Guo, C. Spielmann, B. Walker, et al., “USPL X-ray Generation: Generation of Hard X Rays by Ultrafast Terawatt Lasers,” Review of Scientific Instruments, Vol. 74, pp. 41–47 (2001).
- J. Kasparian and J. Wolf, “USPL Remote Sensing and Atmospheric Propagation: Physics and Applications of Atmospheric Nonlinear Optics and Filamentation,” Optics Express, Vol. 16, pp. 466–492 (2008).
- D. McMorrow, W. Lotshaw, J. Mellinger, et al., “Subbandgap Laser-Induced Single Event Effects: Carrier Generation via Two-Photon Absorption,” IEEE Transactions on Nuclear Sciences, Vol. 49, pp. 3002–3008 (2002).
- J. Mellinger, S. Buchner, D. McMorrow, et al., “Pulsed Laser SEE Testing: Critical Evaluation of the Pulsed Laser Method for Single Event Effects Testing and Fundamental Studies,” IEEE Transactions on Nuclear Sciences, Vol. 41, pp. 2574–2584 (1994).
- S. Moss, S. LaLumondiere, J. Scarpulla, et al., “Correlation of Picosecond Laser-Induced Latchup and Energetic Particle-Induced Latchup in CMOS Test Structures,” IEEE Transactions on Nuclear Sciences, Vol. 42, pp. 1948–1956 (1995).
- A. Redo-Sanchez and X-C. Zhang, “USPLs in THz Technology and Electron Microscopy: Terahertz Science and Technology Trends,” IEEE Journal of Selected Topics in Quantum Electronics, Vol. 14, pp. 260–269 (2008).
- A. Rundquist, C. Durfee, Z. Chang, et al., “Femtosecond Ti:sapphire USPLs: Ultrafast Laser and Amplifier Sources,” Applied Physics B, Vol. 65, pp. 61–174 (1997).
- E. Van Stryland, M. Sheik-Bahae, A. Said, et al., “Material Property Characterization: Characterization of Nonlinear Absorption and Refraction,” Progress in Crystal Growth and Characterization, Vol. 27, pp. 279–311 (1993).
- A. Weiner, “USPL Pulse Shaping: Femtosecond Pulse Shaping Using Spatial Light Modulators,” Review of Scientific Instruments, Vol. 71, pp. 1929–1960 (2000).
- C. Xu and W. Denk, “Nonlinear Optical Laser SEE: Two-photon Optical Beam Induced Current Imaging Through the Backside of Integrated Circuits,” Applied Physics Letters, Vol. 71, pp. 2578–2581 (1997).
- A. Yurtsever and A. Zewail, “4D Nanoscale Diffraction Observed by Convergent-Beam Ultrafast Electron Diffraction,” Science, Vol. 326, pp. 708–712 (2009).