Degree Type

Thesis

Date of Award

2008

Degree Name

Master of Science

Department

Agricultural and Biosystems Engineering

First Advisor

Robert P. Anex

Abstract

Typical methods for determining ethanol production from biomass feedstocks involve the use of High Performance Liquid Chromatography (HPLC) or Gas Chromatography (GC). Such methods require expensive instruments and the time required to process a large number of samples can delay experimental campaigns and process development. The object of this study was to develop a simple, high-throughput, low-cost ethanol assay using CO2 as a surrogate for ethanol production during fermentation. A chemi-visual sensor was developed based on visually measuring color change due to pH in a buffered indicator solution separated from the fermentation chamber by a CO2-permeable membrane. Carbon dioxide was introduced into the fermentation chamber of the chemi-visual sensor while the pH and red-green-blue (RGB) color values of the phenol red indicator solution were recorded. A CCD camera (WebCam) and image analysis software package developed in Matlaby was used to record the RGB values of the chemi-visual solution at each CO2 loading. Calibration curves were developed for the following relationships: CO2 vs. pH, pH vs. RGB, and CO2 vs. RGB. The chemi-visual sensor solution was used to monitor CO2 production in a series of glucose fermentations. The CCD camera recorded the RGB signal and samples of the fermentation broth were taken throughout the experiments. The use of green signal change in the chemi-visual solution as a predictor for ethanol production can account for approximately 92% of the change in actual ethanol content for real-time ethanol production values. Multiple fermentations were conducted in order to calibrate the chemi-visual sensor and to characterize the accuracy of ethanol predictions. It was determined that it would be most appropriate to use this sensor as a predictor of final ethanol production values since dynamic effects of fermentation kinetics, gas transfer, and green signal variability make predictions of real-time ethanol values less reliable.

Copyright Owner

Steven Bly

Language

en

Date Available

2012-04-30

File Format

application/pdf

File Size

78 pages

Share

COinS