Introduction
The use of radiation fields with well-defined properties is one of the fundamental ways in which we probe matter. For example, pulses of radiofrequency waves with well defined phase relationships are routinely used in nuclear magnetic resonance (NMR) experiments to probe structure and spin dynamics in different molecules. These principles have been adapted to do magnetic resonance imaging (MRI), an important diagnostic technique in medicine. In the optical domain, coherent spectroscopy and imaging represent emerging fields, owing to the difficulty in generating short and stable pulses of light at different frequencies. We have developed methods to give enhanced control over optical radiation fields – tailored phase and amplitude modulated femtosecond laser pulses or phase shifted pulse sequences. We now use these methods to extract dynamical information about the interaction of molecular systems with their complex environment and to enhance biomedical imaging.
Ultrafast Laser Pulse Shaping
Shaping ultrafast pulses in the time domain (as is done in NMR) is hindered by the lack of devices that can alter the electric field of light at such high speeds. Instead, one can take advantage of the linear properties of Fourier decomposition to shape pulses in the frequency domain. Since a light pulse contains different frequencies of light, these colors may be spatially separated by diffracting the pulse with a prism or grating. Once the colors are spatially resolved, they may be manipulated by a variety of means to generate a tailored pulse. An acousto-optical crystal driven by radiofrequency waves may be used to diffract the light once again in a completely controllable manner. The resulting mode is spatially recompressed to give a tailored light pulse that may be used in spectroscopic or imaging experiments.
Nonlinear Coherent Spectroscopy
Our group uses pulse shaping techniques to perform nonlinear coherent spectroscopy in a variety of ways. One way is to create a train of pulses in which each individual pulse looks much like a simple Gaussian laser pulse, but where successive pulses are modulated to differ in amplitude and phase. This modulation pattern then gets altered by the nonlinear interaction of the light with the materials to be studied, for example by the process of two-photon absorption. When we use pulse trains of multiple colors the modulation pattern can be transferred from one beam to the other, which lets us study processes like sum-frequency absorption and excited state absorption. Another complementary way to perform nonlinear coherent spectroscopy is to substantially alter the spectrum of each individual pulse from its initial Gaussian shape. The spectrum changes in a characteristic way when the pulse passes through a given nonlinear medium. If we choose the initial spectrum appropriately we can detect these nonlinear shape changes sensitively and selectively. We have used this technique to simultaneously detect two-photon absorption and self-phase modulation. A hybrid approach (and the one that most closely resembles NMR spectroscopy) is to generate a sequence of pulses with varying inter-pulse delays and inter-pulse phase shifts. Changing the delay/phase sequence allows us to perform 2D optical spectroscopy, the optical analog to 2D NMR spectroscopy. We have investigated a variety of atomic and molecular systems using this technique.
Biomedical Imaging
Probing the structure and dynamics of molecules using tailored light pulses is an important task in its own right, although the ultimate goal of our studies is to apply these pulses to tissue in order to improve biomedical imaging. The main focus of our work is the application of these methods for in vivo tissue imaging whereby fundamentally new contrast mechanisms may be explored that are sensitive to structure and function. We expect our technique to possess higher sensitivity than existing optical microscopy methods, leading to an increase in the achievable imaging depth. Here we describe the advantages that our techniques offer for imaging in highly scattering tissue. Using our pulse train modulation technique we are now able to non-invasively obtain images of melanin, which may greatly aid in detecting the early onset of melanoma. We are also able to image oxy- and deoxy-hemoglobin and thereby determine the blood oxygenation in tissue non-invasively. Using the spectral reshaping technique we are also able to detect intrinsic nonlinear signatures of neuronal activation. This promising new technique may substantially enhance functional neuronal imaging.
