Axil: That article is exceedingly difficult to read. It's 2 pages embedded into 969 pages of PDF and page 120 is blank. For the sake of others, so they do not have to try 6 times to load the page, I have copied what I could, but it does not contain page 120... because it is blank.
http://www.lanl.gov/science-innovation/_assets/FY11-Annual-Report.pdf was successfully mounted in the instrument chamber, was fired up, and was shown to work extremely well, capable of producing a tightly focused beam of electron pulses (~1 mm spot size, Figure 1b) controllable in intensity from 5-150 picoamps, and at voltages from ~8 to 21 kilovolts. In the second quarter of FY11, we focused on optimization of the luminescence collection pathway, using a simple drop-cast film of cadmium selenide/zinc sulfide core/shell nanocrystals to produce enough cathodoluminescence signal to couple the instrument to a Hamamatsu streak camera for collection of time- and spectrally-resolved emission. The optimization of the collection pathway and synchronization were far from trivial, but as of March, the instrument has been reliably producing beautiful streak camera traces of nanocrystal emission produced by electrons. Further analysis has brought to our attention a number of subtle yet extremely important details about the operation of this streak camera, including how to avoid saturation effects from the instrumental gain, and how to account for the effect of the reverse sweep of the oscillating streak bias. ** While our appreciation of the complexities of the instrument was still evolving, the third quarter of FY11 was devoted to collection of a database of cathodoluminescence traces from a size-diverse series of cadmium selenide/ zinc sulfide (CdSe/ZnS) core/shell nanocrystals. In parallel, we collected time-resolved photoluminescence on the same exact samples, which provide a vital starting point for analysis of features found in the more complex cathodoluminescence traces. A comparison of the two is shown in Figure 2. At this point, we can confidently say that the quality of the data between the two experiments is completely equivalent, which is a significant achievement in its own right. As of now, we have full data sets for four different sizes of CdSe/ZnS nanocrystals, emitting from 530 to 665 nm. *Figure 2. (a-c) Time-resolved photoluminescence for a sample of CdSe/ZnS nanocrystals emitting at 630 nm. In (a), only single-exciton dynamics are observed. As power is increased (b), a faster biexciton peak emerges at the same energy. Finally, at very high powers (c), a very fast triexciton peak is seen at shorter wavelengths. (d) A cathodoluminescence trace of the same sample shows strong contributions from the multiexciton states. When these contributions are deconvolved, hints of charged “trion” emission emerge.* Through careful analysis over the final quarter of FY11, we have both fully characterized the time resolution of our instrument, and developed the methodology for identifying and deconvolving the contributions of several types of excited states to the cathodoluminescence decay traces. A side-by-side comparison of traces recorded at several beam currents revealed that decay lifetimes associated with a given size of nanocrystal showed no dependence on the beam current, as expected (Figure 3a). However, the apparent rise-time of the signal depended strongly on the current, but not at all on the size of the nanocrystals. Effective pulse widths were extracted from the rise times measured for a range of currents and sizes. When plotted together, the dependence of the pulse widths on current becomes clear (Figure 3b). Extrapolating to 0 current, we see that our nominal response time for cathodoluminescence is 9 ps, which compares favorably to the measurements using the streak camera with laser excitation (~6 ps). We believe that the additional rise stems from a combination of Coulombic pulse spread and possible bulk charging effects within the film itself during measurements. Ongoing experiments are aimed at minimizing the importance of these effects. *Figure 3. (a) Although rise time slows with increasing current, the decay lifetimes of a single nanocrystal sample do not vary, and are superimposable at later times. (b) An analysis of the effect of current on lifetime for several nanocrystal sizes reveals a consistent trend that implies an instrument response time of only 9 ps, which is further broadened by pulse effects that can be processed out after the measurement.* At the present time, we are using the rise-time measurements and laser-based photoluminescence experiments to deconvolve the chief features in our cathodluminescence traces. Preliminarily, we report contributions from a range of excited states, including single exciton, biexciton and tri- and higher multi-exciton states (these last states being indistinguishable by our method). In addition, we have evidence of a substantial signal from charged nanocrystal states, i.e., excited states with an imbalance in electrons and holes resulting in a net charge on the nanocrystal, the simplest being a “trion” (two electrons and one hole). Currently, we are refining a model for inferring the fraction of charged nanocrystals in the whole population. *Future Work* The exact fraction of nanocrystals that are charged by the initial high-energy excitation is the subject of current analysis. However, it is fairly clear that it is significant. Since multiple states with a charge imbalance decay by radiative recombination (slow) and non-radiative Auger recombination (fast) to eventually produce non-emissive ground-state charged nanocrystals, this result yields a possible reason why nanocrystal-based scintillators have underperformed. This suggests two new immediate goals. The first is to show that the electron excitation method we employ does not preferentially result in more charged nanocrystals than would be produced by a gamma-ray of similar energy. This will be the subject of near term studies based on varying the sample form factor, by which we hope to modify the ability of potentially excess charges to be captured by or flow out of the sample. The second is to study nanocrystals or composites that should exhibit modified charging probability (that is, materials that are easier or harder to ionize). Intentional manipulation of the fraction of charged nanocrystals would be the first step towards true optimization of the gamma-to-photon transduction process. *Conclusion* Now that we have optimized and benchmarked the performance of our cathodoluminescence instrument, we have been able to collect the first substantial set of decay traces on a size series of CdSe/ZnS NCs, which is the material system most studied for gamma scintillation. In conjunction with photoluminescence studies, we have been able to identify the contributions of a number of familiar excited states to the decay dynamics. Most surprising so far is the not-insignificant presence of charged nanocrystals. While more complete and quantitative analysis is under way, this observation already suggests the next logical steps for our studies, which will focus on the determining the source and possibly manipulating the fraction of charged states. If this turns out to be the dominant factor in the performance of semiconductor nanocrystals as gamma scintillators, our study will be the first ever to suggest an active pathway toward realizing the latent potential of these materials. This could very well reinvigorate worldwide efforts in developing cheap, rugged replacements for single-crystal scintillators based on this fascinating class of materials.
