We have developed a new imaging method, ultrafast optical wide field microscopy, capable of rapidly acquiring wide field images of nearly any sample in a non-contact manner with high spatial and temporal resolution. Time-resolved images of the photoinduced changes in transmission for a patterned semiconductor thin film and a single silicon nanowire after optical excitation are captured using a two-dimensional smart pixel array detector. These images represent the time-dependent carrier dynamics with high sensitivity, femtosecond time resolution and sub-micrometer spatial resolution.
© 2013 OSA
Optical microscopy is a well-established tool for examining small features in many physical, chemical, and biological systems [1, 2]. However, nearly all microscopic measurements have been performed without high temporal resolution, which would add critical insight into the properties of many systems. In fact, present-day microscopic imaging is nearly always performed in a time-integrated mode, such that the full information contained in the time-domain profile is under-utilized. Furthermore, the recent explosion of interest in nanoscale materials [3–5] has made it even more important to capture carrier dynamics with high spatial and temporal resolution. Ultrafast optical spectroscopy is the only technique capable of providing femtosecond temporal resolution over a broad energy range [6, 7]. Therefore, the marriage of wide field optical microscopy and ultrafast optical spectroscopy stands to revolutionize the study of a number of systems, ranging from living cells to metamaterials to high-temperature superconductors.
Previous efforts to obtain simultaneous sub-micrometer spatial resolution and sub-picosecond temporal resolution in wide field optical imaging experiments were limited to measurements of large (>1%) pump-induced reflectivity changes, due to the inability to combine standard lock-in amplification techniques used for detecting small pump-induced changes with CCD detection [8, 9]. Alternatives such as the combination of laser scanning microscopy with pump-probe spectroscopy suffer from long data acquisition times due to the need to scan both the time delay and the sample position to obtain an image, making it substantially more difficult and time-consuming to perform temperature-and-wavelength dependent pump-probe measurements. However, time-resolved optical scanning microscopy has still been used to shed light on spatiotemporally varying dynamics in various systems, including semiconductor nanowires [10–12], graphene , GaAs , and Si . Finally, other approaches to imaging spatiotemporal dynamics have recently been demonstrated, including femtosecond Kerr-gated wide field fluorescence microscopy  and serial time-encoded amplified microscopy (STEAM) . Although the first of these is capable of tracking dynamics with femtosecond time resolution and diffraction-limited spatial resolution, its applicability is limited to imaging luminescent systems (and therefore will not apply to non-luminescent systems such as metals, superconductors, etc.). In addition, STEAM is able to acquire images in real time at 6.1 MHz frame rates, but in the initial demonstration its temporal resolution was limited to 440 ps.
Here, we utilize a two-dimensional (2D) smart pixel array detector to create an ultrafast optical wide field microscope, capable of rapidly acquiring wide field microscopic images with high sensitivity, femtosecond time resolution and sub-micrometer spatial resolution. Time-resolved images of a gold-patterned amorphous silicon film and single silicon nanowire (NW) are acquired to demonstrate the validity of this novel concept. Importantly, these images were obtained with two different femtosecond laser systems, demonstrating the versatility of this technique. Overall, ultrafast optical wide-field microscopy (UOWFM) is a very general technique that is applicable to any material that can be imaged with a conventional optical microscope, and can thus be expected to have many future applications in a variety of physical, chemical, and biological systems.
2. Experimental setup for ultrafast optical wide-field microscopy
An attractive feature of UOWFM is that it can be easily implemented in nearly any existing optical pump-probe setup, largely independent of the specific femtosecond laser source. A generic UOWFM system in Fig. 1 is based on a conventional pump-probe system, which begins by splitting the output of a femtosecond laser system into pump and probe beams. The beams then are focused onto the sample with a lens and imaged onto a two-dimensional (2D) smart pixel detector array (provided by Heliotis AG) using a 50X (0.50 NA) objective and a zoom lens with variable magnification. This 2D detector array, originally developed for optical coherence tomography , was previously used to perform time-and-wavelength-resolved spectroscopy , but has never been used for ultrafast optical imaging, to the best of our knowledge. The 144x90 pixel silicon detector array used here performs real-time amplitude demodulation on each pixel and can achieve sensitivities comparable to those attainable through lock-in amplification, making it possible to detect small signals in optical pump-probe experiments . The quantum efficiency of the detector ranges from 40 to 60%, for the probe wavelengths used here, and the noise-equivalent power (NEP) is ~4x10−3 W/m2Hz. This allows us to calculate that each pixel was not photon shot-noise limited in our experiments, influencing the sensitivity. The integration time in our experiments was 17 ms; this parameter can be varied from 10 to 655 ms as needed.
Conventional pump-probe experiments can also be performed in this generic setup by replacing the 2D smart pixel detector with a typical single pixel detector. In addition, this setup can operate as a standard optical microscope by blocking the pump beam and modulating the probe beam. Finally, we note that UOWFM can be performed in either transmission or reflection; the experiments described below were performed in transmission as this was optimal for the specific samples being examined.
Our general experimental procedure is to first block the pump beam, modulate the probe beam with an optical chopper, and record the resulting image on the 2D detector array (as in a conventional optical microscope). We then use a flip mirror to direct the probe onto the single pixel detector, modulate the pump, and take a standard pump-probe scan. Finally, we flip the mirror back to direct the probe to the 2D detector array while still modulating the pump, allowing us to acquire time-resolved microscopic images by scanning the delay stage. We can thus provide several different types of data from a single experimental setup: static optical microscopic images of the sample, time-resolved photoinduced changes in transmission or reflection, and time-resolved wide field images with high spatial and temporal resolution. The ultimate temporal resolution of our UOWFM system is primarily determined by the pump and probe pulse durations, while the ultimate spatial resolution is primarily determined by the probe wavelength and focusing objective.
3. Ultrafast optical wide field microscopy on a gold patterned amorphous Si film
For the first experimental demonstration of UOWFM, we examined spatiotemporal dynamics in a gold patterned amorphous Si film (described in more detail in ref .), chosen to benchmark the performance of our system for its relatively large photoinduced change in transmission (ΔT/T) and microscopic spatial features. In those experiments, the setup was based on a 100 kHz regeneratively amplified Ti:sapphire laser system which produced 50 femtosecond, 10 μJ pulses at 800 nm that concurrently pumped two optical parametric amplifiers, enabling measurements with independently tunable pump and probe wavelengths from 400 nm to 2.4 μm (Fig. 2(a) ) [20, 21]. A 400 nm pump beam and a 550 nm probe beam were non-collinearly focused onto the sample in transmission with standard lenses, as in a conventional pump-probe setup, and a mechanical chopper modulated the pump beam at 4 kHz. After the sample, the probe beam was imaged onto the 2D smart pixel detector array by a 50X (0.50 NA) objective and a zoom lens with variable magnification. A scanning electron microscopy (SEM) image of the photolithography mask used to pattern our sample is shown in Fig. 2(b). Following the experimental procedure described in section 2, we used conventional optical microscopy to obtain an optical image of the Au patterned amorphous Si sample used in this work (Fig. 2(c)); we note that some of the features appear blurred due to imperfections in the photolithography process. An image taken with the 550 nm probe beam at the 2D detector array without pump excitation (Fig. 2(d)) compared well with that in Fig. 2(b), demonstrating that our system operates with imaging performance comparable to that of a conventional optical microscope.
The probe beam was then focused onto a single-pixel InGaAs detector and the chopper was moved to the pump beam to perform a standard pump-probe measurement. The resulting time-resolved signal (Fig. 3(a) ) revealed carrier relaxation dynamics at 550 nm in our sample after 400 nm excitation. To demonstrate UOWFM, we then redirected the probe beam to the 2D smart pixel detector and acquired images at different time delays between the pump and probe pulses (Figs. 3(b)-3(e)). As expected, there was no signal at negative time delays. At a pump-probe time delay of t = 0 (Fig. 3(b)), an image became visible (Fig. 3(c)), indicating that photoexcited carriers in the amorphous Si film modified the probe transmission. The photoinduced transmission change (proportional to the image intensity) was maximum from t = 1-10 ps (Fig. 3(d)), as anticipated from Fig. 3(a), and began to decay thereafter. The image had nearly disappeared by t = 150 ps, indicating that a significant fraction of the photoexcited carriers have returned to equilibrium (Fig. 3(e)). Overall, the timescale for carrier relaxation measured using UOWFM corresponded well with that from the single pixel detector (Fig. 3(a)), verifying that UOWFM can record carrier dynamics while also providing information regarding any spatial variations of the dynamics. A movie depicting images acquired at many different time delays provides additional detail (Fig. 3(f)).
We note that the 2D array detector in our UOWFM system performs with sensitivity comparable or better to that of the single pixel detector in these experiments. This can be seen by considering that the photoinduced change in transmission measured by the single pixel detector is spread over ~6000 pixels on the 2D array detector, so each pixel in the 2D array detector detects a significantly smaller ΔT signal that is comparable to the noise level of our single pixel detector, as measured with a lock-in amplifier. We thus estimate that our 2D array detector should be able to detect ΔT/T signals with similar magnitudes (potentially smaller) as those measurable with the single pixel detector, although further studies must be done to quantify this, since it depends on the particular experimental conditions and detector settings. In addition, the sensitivity could be further improved by, e.g., averaging several images at a given time delay, optimizing the detector and/or modulation frequency for this application, and achieving shot-noise-limited performance. Finally, it is also worth noting that each image was acquired in <1 second, demonstrating the rapid image acquisition capability of this technique, especially since each image represents ΔT signals acquired over a relatively large 50 μm by 50 μm area.
4. Ultrafast optical wide field microscopy on a single Si nanowire
Ultrafast optical wide field microscopy will be particularly useful for studying spatiotemporal dynamics in nanoscale materials, which are of great importance for many applications, such as nanoscale transistors [22, 23] and photovoltaics . We therefore explored the use of UOWFM to map carrier dynamics in single Si NWs, aided by our extensive experience with these nanomaterials [11, 12, 21]. To perform UOWFM measurements on single NWs, a microscope objective-based time-resolved optical spectroscopy system  was used with the 2D smart pixel array detector (Fig. 4(a) ). In this system, a 840 nm femtosecond Ti:sapphire laser oscillator was used to generate both pump and probe beams, with the probe power < 10% of the pump power. The 420 nm pump beam was produced through second harmonic generation in a BBO crystal. The pump (5 μm spot size) and probe (2 μm spot size) beams were collinearly focused at the sample position by using a 20X (0.4 NA) microscopic objective lens, with polarization parallel to the NW axis. This system thus enabled us to effectively control the position and direction of both beams on a single NW, and allowed us to selectively measure its optical properties without any complications from inhomogeneous broadening, as is often encountered when studying NW ensembles [11, 21]. The pump beam was then modulated at 4 kHz by a mechanical chopper. The sample studied here consisted of a  oriented Si NW fabricated by e-beam lithography and Si deep reactive ion etching, which had a core Si diameter of 250 nm with a 75 nm SiO2 shell layer . The Si/SiO2 NW was then transferred to a sapphire substrate, on which it lies flat, which allowed us to image it as in a conventional microscope (Fig. 4(b)).
Initially, while chopping the pump beam, the transmitted probe beam was focused onto a single pixel detector for a standard pump-probe measurement. The measured ΔT/T signal (Fig. 4(c)) can be fit with a decay time constant of τ = 217 ps, which is due to carrier diffusion, carrier trapping, and recombination, described in more detail in ref . We note that our core-shell NW had a longer decay time constant as compared to the values from other Si NWs with the same diameter, owing to the growth-seed (Au)-free-fabrication method as well as surface passivation from the oxide shell .
In order to measure the spatial resolution of our system, we performed UOWFM on the Si NW in static imaging mode without pump excitation, i.e. the transmitted probe beam was chopped and directed to the 2D smart pixel detector with the pump beam blocked (Fig. 5(a) ) In this mode, the probe beam diameter was enlarged (> 5 μm) to show the whole shape of the SiNW. For this sample, the surrounding sapphire substrate was also transparent at this wavelength in the present geometry; in contrast, the amorphous Si film was surrounded by opaque gold patterns in the previous experiment. Thus, the directly transmitted probe beam around the NW was also projected to the 2D smart pixel detector, blurring the image. In addition, the image appeared dark at the NW position and bright everywhere else in Fig. 5(a), since the direct transmission through the substrate was much larger than the transmission through the NW.
To selectively image the single Si NW without a background signal from direct transmission through the substrate, which is normally much bigger than the signal from the Si NW, an additional measurement step was necessary. This entailed recording the probe image on a blank region of the sapphire substrate (Fig. 5(b)) in the same manner. Subtracting the image in Fig. 5(a) from the image in Fig. 5(b) then allowed us to isolate the image from the NW region (Fig. 5(c)). Furthermore, we obtained the spatial resolution of our UOWFM system (~770 nm (Fig. 5(d))) from measuring the cross section of the NW image in Fig. 5(c). Using this method, background-free microscopic images could be obtained, even for samples with transparent substrates.
To obtain time-resolved optical images on the single Si NW, our UOWFM experiments were performed in the same manner as our experiments on the Au patterned amorphous Si film, i.e. the transmitted probe beam was directed to the 2D smart pixel detector, which allowed us to capture a time-dependent image of the single NW. We note that since the time-dependent images were obtained while modulating the pump beam and detecting the probe beam, as in a standard pump-probe experiment, it is unnecessary to follow the procedure described above to obtain static images of a sample on a transparent substrate (since there is no photoinduced change in the probe transmission through the substrate).
Images of the Si NW at different time delays between pump and probe, with both beams overlapped on the center of the NW, are shown in Figs. 5(d)-5(f). At t = 0, a bright image was observed (Fig. 5(e)), indicating that photoexcited carriers in the Si/SiO2 NW modify the probe transmission, followed by a decay in the image intensity, as seen for t = 100 ps in Fig. 5(f). At t = 400 ps the brightness was much weaker, but still observable, which was due to the surface passivation provided by the SiO2 shell (Fig. 5(g)); in unpassivated NWs the ΔT/T signal returns to zero within ~200 ps . Finally, we note that since the pump and probe spot diameters were much smaller than the length of the NW along its axis (~9 μm), only the center of the NW could be photoexcited and detected, which caused the brightest region in each image to be circular (Figs. 5(e)-5(g)). In general, optimization of this system will require careful consideration of the pump and probe beam profiles, both to obtain diffraction-limited spatial resolution as well as to optimize the system for a given sample.
We emphasize that this imaging technique, which combines time-resolved optical spectroscopy with a 2D smart pixel array detector, works for both small samples with sub-micron spatial features (which will be very useful in nanomaterials research) as well as relatively large targets (possessing spatial features on the order of microns), with high temporal and spatial resolution as well as rapid image acquisition times in both cases. In essence, UOWFM makes it possible to look through an optical microscope and record the evolution of the sample properties in time after photoexcitation with femtosecond time resolution.
In conclusion, we have demonstrated ultrafast optical wide field microscopy with high sensitivity for the first time. This novel technique combines the non-contact nature and spatial resolution of conventional optical microscopy with the temporal resolution of ultrafast spectroscopy to rapidly and sensitively acquire spatially- and temporally-resolved images of nearly any sample. Time-resolved optical images of a single Si NW provide valuable insight into carrier dynamics with sub-micron optical resolution, while time-resolved images of a Au-patterned Si film demonstrate our ability to capture information over a large area within a short acquisition time. Future work will include further optimization of the 2D array detector for this unique application, as well as the use of this technique to track space- and time-varying processes in a variety of biological, chemical, and physical systems. A particularly exciting prospect is to combine UOM with approaches for increasing the spatial resolution in a wide field microscopy experiment, such as imaging interferometric microscopy  or non-invasive optical nanoscopy . Both of these techniques can provide sub-diffraction limited spatial resolution, which could make optical wide field imaging with simultaneous sub-100 fs temporal resolution and sub-100 nm spatial resolution possible.
This work was supported by Los Alamos National Laboratory under the auspices of the U.S. Department of Energy, under contract No. DE-AC52-06NA25396. Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000. This research was performed at the Center for Integrated Nanotechnologies (CINT), a U.S. Department of Energy (DOE), Office of Basic Energy Sciences (BES) user facility, and also partially supported by the NNSA’s Laboratory Directed Research and Development Program. We acknowledge Patrick Lambelet of Heliotis AG for assistance in setting up and optimizing the 2D smart pixel array detector and Prashanth Upadhya for assistance in the initial experiments.
References and links
1. S. Inoue and K. R. Spring, Video Microscopy: the Fundamentals (Plenum Press, 1997).
2. A. Neumann, Y. Kuznetsova, and S. R. J. Brueck, “Microscopy,” in Optical Techniques for Solid-State Materials Characterization, eds. R. P. Prasankumar and A. J. Taylor, (CRC Press, 2011).
3. P. Cherukuri, S. M. Bachilo, S. H. Litovsky, and R. B. Weisman, “Near-infrared fluorescence microscopy of single-walled carbon nanotubes in phagocytic cells,” J. Am. Chem. Soc. 126(48), 15638–15639 (2004). [CrossRef] [PubMed]
4. M. Böhmler, Z. Wang, A. Myalitsin, A. Mews, and A. Hartschuh, “Optical imaging of CdSe nanowires with nanoscale resolution,” Angew. Chem. Int. Ed. Engl. 50(48), 11536–11538 (2011). [CrossRef] [PubMed]
5. G. Beaune, B. Dubertret, O. Clément, C. Vayssettes, V. Cabuil, and C. Ménager, “Giant vesicles containing magnetic nanoparticles and quantum dots: feasibility and tracking by fiber confocal fluorescence microscopy,” Angew. Chem. Int. Ed. Engl. 119(28), 5517–5520 (2007). [CrossRef] [PubMed]
6. D. J. Hilton, “Ultrafast pump-probe spectroscopy,” in Optical Techniques for Solid-State Materials Characterization, R. P. Prasankumar and A. J. Taylor, eds. (CRC Press, 2011).
7. R. D. Averitt and A. J. Taylor, “Ultrafast optical and far-infrared quasiparticle dynamics in correlated electron materials,” J. Phys. Condens. Matter 14(50), R1357–R1390 (2002). [CrossRef]
8. V. V. Temnov, K. Sokolowski-Tinten, P. Zhou, and D. von der Linde, “Femtosecond time-resolved interferometric microscopy,” Appl. Phys., A Mater. Sci. Process. 78(4), 483–489 (2004). [CrossRef]
9. M. C. Downer, R. L. Fork, and C. V. Shank, “Femtosecond imaging of melting and evaporation at a photoexcited silicon surface,” J. Opt. Soc. Am. B 2(4), 595–599 (1985). [CrossRef]
10. C. R. Carey, Y. Yu, M. Kuno, and G. V. Hartland, “Ultrafast transient absorption measurements of charge carrier dynamics in single II-VI nanowires,” J. Phys. Chem. C 113(44), 19077–19081 (2009). [CrossRef]
11. M. A. Seo, S. A. Dayeh, P. C. Upadhya, J. Martinez, B. S. Swartzentruber, S. T. Picraux, A. J. Taylor, and R. P. Prasankumar, “Understanding ultrafast carrier dynamics in single quasi-one-dimensional Si nanowires,” Appl. Phys. Lett. 100(7), 071104 (2012). [CrossRef]
12. M. A. Seo, J. Yoo, S. A. Dayeh, S. T. Picraux, A. J. Taylor, and R. P. Prasankumar, “Mapping carrier diffusion in single silicon core-shell nanowires with ultrafast optical microscopy,” Nano Lett. 12(12), 6334–6338 (2012). [CrossRef] [PubMed]
13. L. Huang, G. V. Hartland, L.-Q. Chu, R. M. Luxmi, R. M. Feenstra, C. Lian, K. Tahy, and H. Xing, “Ultrafast transient absorption microscopy studies of carrier dynamics in epitaxial graphene,” Nano Lett. 10(4), 1308–1313 (2010). [CrossRef] [PubMed]
14. Y. Fujii, K. Horiuchi, F. Kannari, M. Hase, and M. Kitajima, “Optical imaging of defect density distribution in ion-implanted GaAs using ultrafast carrier dynamics,” Jpn. J. Appl. Phys. 43(1), 184–185 (2004). [CrossRef]
15. A. Othonos and C. Christofides, “Spatial dependence of ultrafast carrier recombination centers of phosphorus-implanted and annealed silicon wafers,” Appl. Phys. Lett. 81(5), 856 (2002). [CrossRef]
18. S. Bourquin, P. Seitz, and R. P. Salathé, “Two-dimensional smart detector array for interferometric applications,” Electron. Lett. 37(15), 975–976 (2001). [CrossRef]
19. S. Bourquin, R. P. Prasankumar, F. X. Kärtner, J. G. Fujimoto, T. Lasser, and R. P. Salathé, “High-speed femtosecond pump-probe spectroscopy with a smart pixel detector array,” Opt. Lett. 28(17), 1588–1590 (2003). [CrossRef] [PubMed]
20. K. M. Dani, Z. Ku, P. C. Upadhya, R. P. Prasankumar, S. R. J. Brueck, and A. J. Taylor, “Subpicosecond optical switching with a negative index metamaterial,” Nano Lett. 9(10), 3565–3569 (2009). [CrossRef] [PubMed]
21. R. P. Prasankumar, P. C. Upadhya, and A. J. Taylor, “Ultrafast carrier dynamics in semiconductor nanowires,” Phys. Status Solidi B 246(9), 1973–1995 (2009). [CrossRef]
22. M. T. Bjork, J. Knoch, H. Schmid, H. Riel, and W. Riess, “Silicon nanowire tunneling field-effect transistors,” Appl. Phys. Lett. 92(19), 193504 (2008). [CrossRef]
23. S. A. Dayeh, C. Soci, P. K. L. Yu, E. T. Yu, and D. Wang, “Influence of surface states on the extraction of transport parameters from InAs nanowire field effect transistors,” Appl. Phys. Lett. 90(16), 162112 (2007). [CrossRef]
25. Y. Cotte, F. Toy, P. Jourdain, N. Pavillon, D. Boss, P. Magistretti, P. Marquet, and C. Depeursinge, “Marker-free phase nanoscopy,” Nat. Photonics 7(2), 113–117 (2013). [CrossRef]