 Currently, there's a lot of interest in replacing petroleum transportation fuels with clean and renewable alternatives. One option is to convert lignocellulosic materials or plant biomass, non-edible plants, into biofuels. Currently, one of the most promising conversion techniques is within a biochemical conversion platform. Now within this platform, the first step is to convert the biomass into simple sugars like glucose and xylose that can then be fermented into biofuels. And usually, this bioconversion platform consists of two processes, a low severity pretreatment followed by enzymatic hydrolysis. In our work, we used Fourier Transform Infrared Spectroscopy, or FTIR for short, to predict sugars released in enzymatic hydrolysis. We took six biomass substrates that ranged from corn stover to switchgrass to mixed hardwood, and each substrate underwent four levels of alkaline pretreatment to generate 24 pretreated substrates with a wide range of digestibilities. Each of these 24 substrates none underwent enzymatic hydrolysis with cellulase enzymes. In addition, we took FTIR spectra of each of the pretreated substrates. And I should note that taking an FTIR spectra takes about 30 seconds. It's quick and easy and requires a really small sample size. We measured the sugars released from enzymatic hydrolysis with HPLC, and we then correlated the FTIR spectra to the sugars using a multivariate statistical technique called PLS regression. We were happy to find out that the PLS regression model could accurately predict the sugars released in enzymatic hydrolysis. In addition, we were able to look at the regression coefficients and see that the regions of the spectra that were deemed important by the model for sugar production had all been attributed previously to known chemical bonds and functional groups in lignocellulose. So this told us that the model was rooted in reality and not some sort of mathematical artifact. So what's the next step? First, the model needs to be extended to predict overall sugar productions from combined pretreatment and enzymatic hydrolysis. That's already underway. And after that, the model should be extended to include additional pretreatment types, both acidic and other alkaline pretreatments, as well as other enzyme mixtures and enzyme loadings. Now, this is really exciting because if this model works for other pretreatments and enzyme mixtures, it has the potential to replace a lot of the wet chemistry techniques that practitioners spend so much time on. In addition, if we construct these models carefully with enzymes with known specificities and pretreatments, we may be able to actually elucidate some of the mechanism by which lignocellulose resists enzymatic hydrolysis.