The exquisite selectivity and sensitivity of SHG for the detection of oriented biopolymers (e.g., collagen networks) has opened up exciting opportunities in label-free biological imaging. We have been working to develop a predictive framework for bridging local structure and orientation to macroscopic polarization-dependent measurements in biological assemblies.

We have developed a perturbation theory approach to predict vibrational and electronic second order NLO property of protein secondary structure (Acc. Chem. Res., 2007; J. Phys. Chem. B, 2006; Chem. Phys. Lett., 2004).  In brief, the second order NLO tensors of a protein can be recovered through relatively simple models of exciton coupling between adjacent chromophores.  
More recently, we have extended this basic framework to higher-order nonlinear optical effects, including two-photon absorption (Chem. Phys. Lett., 2008) and coherent anti-Stokes Raman (J. Phys. Chem. B, 2008) of biopolymers. Unlike one-photon phenomena such as circular dichroism, polarization dependent chiral sensitivity is electric dipole allowed in higher order nonlinear optics. This sensitivity suggests that NLO-based measurements can potentially serve as alternative tools for protein secondary structure determination, with ultrasmall volume (fL) requirements and producing dichroic ratios (circular-linear difference) approaching 100%.

In collaboration with the Indiana University Scientific Data Analysis Lab, we have developed NLOPredict, a software program capable of predicting polarization dependent, second order NLO effects for thin films of small molecules and proteins.  Given the molecular tensor, NLOPredict is capable of predicting experimental intensities in the Jones coordinate system.  Efforts are currently underway to adapt NLOPredict for predicting and visualizing higher and lower order optical processes and for predicting experimental intensities from bulk and crystalline systems.  Another software program, NLOAnalyze, is currently in the early planning stages.  When completed, NLOAnalyze will compliment NLOPredict by determining likely crystal structures for a sample based on input of experimental measurements. 

Copyright © 2010, Simpson Research Group, Department of Chemistry, Purdue University, all rights reserved
560 Oval Drive, West Lafayette, IN 47907