Ultrafast Science and Energy

Mason Jiang
December 6, 2014

Submitted as coursework for PH240, Stanford University, Fall 2014


Fig. 1: A typical pump-probe experiment setup.

The ongoing search for new and more efficient techniques to generate energy has become an increasingly complicated mission despite the progress made. Improving biofuels, for instance, entails a thorough understanding of all chemical reactions and processes involved, no matter how complex. Likewise, seeking the best substances for efficient energy conversion and transport such as in solar panels and thermoelectric materials means knowing exactly how they will act internally when performing desired tasks.

In particular, the difficulties here are that these chemical processes and internal actions frequently occur at an ultrafast timescale on the order of nanoseconds (10-9 seconds) to femtoseconds (10-15 seconds). This is basically the time it takes for the movement of atoms and molecules as they split, combine, or dart around under the influence of a variety of interactive forces. As expected, without suitable technology, it has been a nontrivial assignment to study these extremely short-timescale phenomena. Luckily, though, the advent and growth of a field called ultrafast science has made it possible to examine these types of processes with high temporal resolution under an assortment of clever methods. Advances in the field have made it a crucial cornerstone in various energy-related investigations today.

What Is Ultrafast Science?

At its heart, ultrafast science is the study of phenomena that happen at the aforementioned timescales and typically with photons as the main tools. To gain detailed information about ultrafast processes, though, the photons used need to have special temporal properties. Imagine the task of trying to measure the change in reflectivity of a particular material caused by the vibration of its atoms with a picosecond (10-12) scale period. A pulse of photons is fired towards the material, reflected at an angle, and collected by a detector to perform the measurement. If the pulse lasts for longer than the picosecond period, then instead of examining the reflectivity change due to a single atomic vibration, the effects of multiple vibration periods are gathered. Thus, to ensure that only the effects of a single atomic vibration period are observed, the pulse must be shorter than the period. This is akin to capturing an action with a camera: if the shutter speed is slower than the events of the action then a blurry photograph will result. A fast enough shutter speed is just like a short enough pulse.

This principle is the basis for the majority of techniques in ultrafast science attempting to learn details about various fast-moving processes. Each method provides its own unique information and applications. The following gives an overview of some of these often used techniques and their applications to energy research.

Pump-Probe Experiments

Pump-probe type experiments are very common in ultrafast science as they provide insight into how a system/material reacts as a function of time (probe) due to an excitation (pump). Usually, the pump itself is a temporally short, intense pulse of laser light that excites a sample of interest electronically, thermally, or mechanically. Whichever the case, the sample's equilibrium structure is disturbed and consequently its optical absorption characteristics altered. The probe is another temporally short pulse of light that is not intense enough to excite the sample, but is able to reflect off of the sample some time after the arrival of the pump pulse and be captured by a detector. This reflected pulse contains information about the absorption of the sample and thus details about how it has been distorted structurally. By continuously varying the delay time between the pump and probe pulses and measuring the material absorption at each point, a so- called time-resolved analysis is performed diagramming how a particular sample reacts to an impetus over time.

Fig. 1 provides a simple pump-probe schematic in which a pulsed laser beam is split into two different paths by a beamsplitter. While path 1 is a relatively unobstructed one, path 2 features a set of mirrors sitting on a delay stage that can either shorten or lengthen the path as it moves. The result of this is that after the two paths recombine collinearly, there are two pulses separated by some delay time Δτ. As explained above, pulse 1 acts as the pump and pulse 2 is the probe. Depending on where the delay stage is during a measurement, pulses 1 and 2 could arrive at the sample at the same time (if the path lengths are exactly the same) or one could arrive after the other (if the path lengths are different). For ultrafast time resolution, the paths only need to be offset by a couple inches or less. In fact, an inch in path length difference is equivalent to a Δτ of merely 85 picoseconds given that the pulses travel at the speed of light. Despite the relative straightforwardness of the experiment, setups like this can provide a plethora of ultrafast-scale information and are useful for many applications including energy.

One type of energy application is the investigation of solar cell structures and materials. Researchers working with devices meant to improve solar energy extraction efficiencies frequently need to characterize how quickly their devices can generate free charge carriers at p-n junctions as a result of photoexcitation. Since the generation of these carriers usually occurs on a picosecond timescale and involves light-induced excitation, ultrafast pump-probe experiments are ideal for these performance gauges. An example of such a study is the use of time-resolved pump- probe spectroscopy by a group from the Eindhoven University of Technology in the Netherlands to examine excitons in thin-film photovoltaic devices made of organic materials. [1] In this experiment, the group has fabricated novel organic/polymer-based photovoltaic devices (nc-ZnO as the organic material and MDMO-PPV as the polymer) and as part of its analysis, wants to understand the carrier generation dynamics across the nc-ZnO:MDMO-PPV interface. Similar to the experimental steps described above, the group first excites the devices using intense 200-fs pulses (pump) to produce strongly bound electron-hole pairs called excitons.

Then, as the excitons dissolve at the nc-ZnO:MDMO-PPV interfaces to form free charge carriers, non-intrusive probe pulses are reflected off the devices, first arriving at exactly the same time as the pump pulses and then gradually coming at delay steps of picosecond-scale after the pump pulses. By collecting a complete set of the outgoing intensities of the probe pulses as a function of delay time, the group is able to discern exactly when the excitons are separated into carriers, which changes the absorption characteristics of the device, and when the device returns to equilibrium. In the end, the group found an exciton decay time of approximately one picosecond, which is indicative of an ultrafast electron transfer and decent performance for a photovoltaic device. In addition to photovoltaic characterizations, ultrafast pump-probe experiments have a function in any energy applications that require the understanding of sub-ns dynamics from photo- induced effects.

Two-Dimensional Infrared Spectroscopy

Another useful practice in ultrafast science is two-dimensional infrared spectroscopy (2D IR spectroscopy), which is only made possible by the ability to generate femtosecond-scale pulses. In many systems and materials, there is often frequency coupling between structural vibration modes, called phonons, which can either be harmonic or anharmonic in nature. These couplings definitively come into play when complex chemical exchanges, energy transfers, and other dynamical processes occur. However, most available methods to inspect these types of activities, such as 2D nuclear magnetic resonance (NMR), do not have enough time resolution since the coupling dynamics happen at femtosecond scales. 2D IR spectroscopy possesses the necessary time resolution (down to <50 fs) and thus is a tremendously functional tool. [2]

Fig. 2: A schematic of the 2D IR spectroscopy pulse sequence. Here τ is the coherence time, T is the waiting time, and t is the detection time.

The basic goal of this technique is to collect a frequency correlation spectrum using a three-pulse sequence of infrared light. Fig. 2 shows this pulse sequence and the various relevant time parameters. The experiment begins with the arrival of pulse 1, which sets all of the vibration modes of the sample of interest oscillating at their natural frequencies and thus creates coherence between them. In other words, the pulse manufactures an in-phase superposition of all of the modes. The time τ between pulse 1 and the arrival of pulse 2 is called the coherence time in which the system is in a nonequilibrium state and the once in-phase modes quickly de-phase and lose coherence in a phenomenon called free induction decay. From this, it is expected that all coherence information such as phase relationships is lost, but pulse 2 serves to preserve it by generating populations of particular oscillation modes to be studied, letting certain phonons resonate and suppressing others. The time T between pulse 2 and the arrival of pulse 3 is called the waiting time and here population dynamics are allowed to occur including interactions between selected phonons and various couplings. After T, pulse 3 appears to incite a state of coherence once again, which occurs after a rephasing time t. Finally, the sample emits an echo signal at the frequency of the coherent state, which is detected and measured.

In performing this experiment, the key variable parameters are the coherence time τ and the waiting time T. The decision of when to introduce pulse 2 and thus how long τ should be effectively chooses which phonon modes will have populations during the waiting period T. In addition, the decision of how long T should be directly affects the sort of dynamics that will be studied since certain interactions will occur on a longer timescale than others. In typical 2D IR experiments, T is kept constant as τ is scanned and multiple T's are investigated. The end results are frequency versus frequency plots for each T, with one axis denoting the coherence frequency of the initial coherent state and the other axis showing the coherence frequency of the final coherent state. By analyzing how these two coherence frequencies correlate and seeing whether they are the same (diagonal peaks) or different (cross peaks), the coupling between available phonon modes in a material or system can be learned. The inspection of the shapes and sizes of these peaks also reveals the nature of various activities such as the interaction of vibration modes with their environments (useful for chemical solutions) and how different chemical process pathways evolve.

A very good example of how ultrafast 2D IR spectroscopy has been used for energy applications is in the study of coupling dynamics in photosynthesis processes. In photosynthetic systems, it is known that electronic couplings are responsible for the conversion of light into energy since they monitor how energy is transported from the pigments that capture light to the conversion areas. [3] Thus, in the quest to completely unravel how photosynthesis works at the atomic scale to learn more about nature's most efficient energy generation process and apply it to man-made energy applications, these electronic coupling dynamics must be well understood and characterized. An instance of such an attempt to describe the couplings in detail is a study of the Fenna-Matthews-Olson (FMO) photosynthetic light-harvesting protein by a group from the University of California, Berkeley. [3] In its study, the group uses exactly the 2D IR spectroscopy technique illustrated above with the FMO proteins as samples and produces plots of coherence frequency before the waiting time T versus coherence frequency after T for various values of T, bound to a couple hundred femtosecond scale. The results produce several diagonal peaks, but more importantly, some cross peaks, which identify the suspected photosynthetic electronic couplings. From analyzing and following these cross peaks with T from 0 to 1000 femtoseconds, the group is able to obtain information about all of the ultrafast energy transfers that occur including which pathways are taken and which molecules see energy flow and when that happens. The spatial and temporal insight gained is crucial to all photosynthesis-related research and is only made available by ultrafast 2D IR spectroscopy.

X-ray Diffraction

One last frequently used method in ultrafast science with plenty of energy applications is actually a very established one in X-ray diffraction. Introductory physics teaches a well-known equation called Bragg's law, which applies to X-rays scattering off of crystals:

2dsinθ = nλ

In this equation, d is the spacing between consecutive diffracting crystal planes, θ is the incident angle, n is an integer, and λ is the wavelength of the scattering X-rays. Fig. 3 depicts this common scheme. This law is the primary basis behind all X-ray diffraction experiments and with this simple geometric relationship, much can be learned about a particular material's crystal structure. As an elementary example, if a material of unknown crystal lattice spacing is studied with X-rays of known wavelength scattering off specified incident angles, then the resulting diffraction pattern collected in combination with Bragg's law will reveal that information.

Fig. 3: An X-ray diffration scheme with X-rays scattering off of a crystal.

Although X-ray diffraction is a classical physics concept, it has also found widespread use in the ultrafast science community. Ultrafast X-ray diffraction is basically the same as typical X-ray diffraction, but with the use of femtosecond X-ray pulses, which provide access to ultrafast time resolution. This may seem trivial at first since non-X-ray pulses are also available on a femtosecond scale such as those used for tabletop pump-probe experiments as discussed earlier. However, the key difference is that X-rays have wavelengths on the angstrom (10-10 meters) scale, which just happens to be at the same size as atoms. This means that ultrafast X-rays possess the unique ability to probe frenetic atomic movements and processes.

Currently, there are only a handful of X-ray facilities around the world that have the proper combination of X-ray intensity, energy, and pulse width desired by researchers who want to perform time-resolved ultrafast experiments on materials. These include powerful synchrotron facilities such as the Advanced Photon Source (APS) at Argonne National Lab and BESSY in Germany along with free electron laser facilities like the Linac Coherent Light Source (LCLS) at SLAC National Lab. At these locations, ultrafast experiment setups actually mimic other well-known ultrafast techniques such as the pump-probe experiments described earlier. A common arrangement, for instance, uses laser or electrical pulses as pumps and takes advantage of X-ray pulses as probes. The rest of the experiment is very familiar, with the use of a delay stage for achieving time resolution and a detector sensitive to X-ray scattering.

Given this, many energy applications are available. One particular matter to look at as a case is the characterization of thermoelectric devices, which can convert heat into useful electrical energy. The goal for these types of apparatuses is to diffuse heat away as quickly as possible so that electrical energy can be generated faster. Normally, this occurs at rates of a couple hundred picoseconds so the research of finding the right materials and designs for these devices falls right into the realm of ultrafast science. Specifically, this sort of investigation is ideal for ultrafast X-ray diffraction since the designs for thermoelectric devices usually involve atomic-scale features and activities and thus need X-ray wavelengths to properly probe.

As an example, one group from Germany recently looked at heat transport through an exotically-fabricated superlattice designed to diffuse heat quickly and efficiently. [4] They actually used the X-rays from the aforementioned BESSY synchrotron to probe their samples after generating heat in the sample superlattices using laser pulses. Just like with the pump-probe experiments, the group used a movable stage to impose a time delay between the pump pulses and the probe pulses. As a result, X-ray diffraction images were collected as a function of time. The analytic principles used by the group were no more difficult than applying Bragg's law for crystal scattering with a constant wavelength λ. Basically, when the thermoelectric device was in a heating phase, the superlattice expanded, increasing its lattice spacing d. Consequently, θ decreased and the position of the Bragg reflection from the superlattice on the detector was shifted. On the other hand, when the device entered a cooling phase, the superlattice contracted its spacing and its Bragg reflection shifted in the other direction on the detector. By tracking the detector position shifts of the superlattice Bragg reflection and plotting them against time delay, the group was able to determine how quickly its devices would heat up and then transport the heat away. From this analysis, the results were that the constructed superlattices transported heat away completely on a scale of tens of nanoseconds, which is pretty decent for heat transfer.


As shown through various effective techniques, ultrafast science is a very functional and versatile field for researching applications for energy. Moving forward with energy research and expanding into the use of more innovative methods to generate energy such as thermoelectric and solar devices, many materials and designs will have to be characterized to determine optimal performance and efficiency. And since the attributes needed to fully understand these materials often manifest themselves in ultrafast processes, the field of ultrafast science will become increasingly important. So far, the applications from this field are mostly related to characterization and benchmarking, but in the future they will expand into a more serious and intricate understanding of all atomic motions in various chemical and physical activities. When this occurs, it may be possible for ideas such as artificial photosynthesis, which would revolutionize energy research.

© Mason Jiang. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.


[1] W. Beek et al., "Hybrid Zinc Oxide Conjugated Polymer Bulk Heterojunction Solar cells," J. Phys. Chem. B 109, 9505 (2005).

[2] J. Zheng, K. Kwak, and M. D. Fayer, "Ultrafast 2D IR Vibrational Echo Spectroscopy," Acc. Chem. Res. 40, 75 (2007).

[3] T. Brixner et al., "Two-Dimensional Spectroscopy of Electronic Couplings in Photosynthesis," Nature 434, 625 (2005).

[4] R. Shayduk et al., "Nanoscale Heat Transport Studied by High-Resolution Time-Resolved X-Ray Diffraction," New J. Phys. 13, 093032 (2011).