Feng, Yinghua
(2013)
SENSING SOLUBLE ORGANIC COMPOUNDS WITH MICROBIAL FUEL CELLS.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
![[img]](http://d-scholarship.pitt.edu/16460/1.hassmallThumbnailVersion/FENGYH_Dissertation2012.pdf)  Preview |
|
PDF (This is the final version, and I have made the corrections basing on the requirements.)
Primary Text
Download (3MB)
| Preview
|
Abstract
Water quality is central to the social, economic, and ecological well-beings, so it becomes vital to monitor aquatic ecosystems. In recent years, multifarious biosensors have demonstrated great potential to support environmental analysis and water quality monitoring. As one type of biosensors, microbial fuel cells (MFCs) have been investigated and shown good operational capabilities. However, the response patterns during MFC-based biosensing process have not been characterized.
This study explored the start-up, operation, and data analysis associated with an air-cathode MFC system. Electrical signals were generated in response to the injections of synthetic water and field samples. The highest coefficient of determination in laboratory testing was produced when the peak area (PA) was correlated with influent COD concentrations, which is the approach that has not been previously reported. However, the peaks obtained in field testing of the MFC were smaller in size and with longer cycle time, and the samples with lower COD produced smaller peak areas (PAs) and peak heights (PHs). Higher coefficients of determination (0.99 for synthetic water and 0.95 for field samples) were obtained the artificial neural network (ANN) model was used for COD determination. Furthermore, the use of ANN permitted accurate identification of acetate, butyrate, glucose and corn starch.
This study also revealed that addition of BES (2-bromoethane sulfonic acid) increased the magnitude of peak area (PA) and columbic efficiency (CE) by inhibiting the activity of methanogens when glucose was used as the primary substrate. A revised ANN was utilized to interpret the low concentration peaks and the result showed that ANN processing expanded detection limits (the lowest linear detectable COD) of MFC biosensor from 20mg/l to a below 5mg /l.
Another properly-trained mathematical model, time series analysis (TSA, at f=0.2) successfully predicted the temporal current trends in properly functioning MFCs, and in a device that was gradually failing.
This study was the first MFC biosensing effort to propose peak area as an appropriate response metric and the first to integrate ANNs and TSA model into MFC-based biosensing. This study is expected to provide a template for future MFC-based biosensing efforts.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
31 January 2013 |
| Date Type: |
Publication |
| Defense Date: |
13 November 2012 |
| Approval Date: |
31 January 2013 |
| Submission Date: |
16 November 2012 |
| Access Restriction: |
1 year -- Restrict access to University of Pittsburgh for a period of 1 year. |
| Number of Pages: |
132 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
Swanson School of Engineering > Civil and Environmental Engineering |
| Degree: |
PhD - Doctor of Philosophy |
| Thesis Type: |
Doctoral Dissertation |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
microbial fuel cells, biosensing, artificial neural networks(ANNs), time series analysis(TSA) |
| Date Deposited: |
31 Jan 2013 21:10 |
| Last Modified: |
15 Nov 2016 14:07 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/16460 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}