Institute for Space Systems Operations * 2001 Annual Report * 72-76

III-Nitrides-on-Silicon Chips for Detection of Live Bacteria in Enclosed Space Environments

Abdelhak Bensaoula, Ph.D., Research Professor; David Starikov, Ph.D., Senior Research Scientist C. Boney, Ph.D., ISSO Post-doctoral Fellow, Texas Center of Superconductivity at the University of Houston (TcSUH)

Portable inexpensive instruments for real-time detection of live microorganisms are needed in several aspects of industrial and human life activities. The most important areas include environmental analysis of indoor air and potable water, the food industry, fossil energy plants, and waste water management. Many space applications can be critical in the real-time detection of live microorganisms in small enclosed environments such as in space missions. Such testing is directed first of all towards protection of the crew and other personnel participating in the mission against airborne and foodborne illnesses. In addition, prevention of biofouling and microbially-induced corrosion can save billions of dollars¹ in replacement and repairs of the costly equipment used in these missions.

Among several methods of bacteria detection that include mass-spectrometry, antibodies, DNA probes, and infrared sensors, employment of optical sensing based on fluorescence for bacteria detection is relatively new and was proposed for the first time in 1998.² Since then, the technique has developed into a reliable method of bacteria detection in aqueous and aerosolized samples. Moreover, portable sensor designs that employ solid-state light sources are disclosed in various patents.^3,4 These methods are based on the employment of SYTO 13 and SYTOX green fluorescent nucleic acid stains as molecular recognition elements and fluorescent reporters. The method is based on the observed increase in fluorescence quantum yields in such stains when they link to a bacteria.

Bacteria type identification is possible through employment of nucleic acid structures that can form complexes only with a specific bacteria. Such a method has already been employed in biochips used in microscopy and cell biology.⁵ The bacteria concentration is detected as a fluorescence intensity change from the SYTO 13 or SYTOX nucleic acid at ~500 nm when excited with a commercial GaN blue (470 nm) LEDs and detected by a Photo Multiplier Tube (PMT) equipped with a band pass optical filter.

In this project, we demonstrated the above method of bacteria detection using UV, blue, and green LEDs and filtered broad-band silicon photodetectors. A mini-grant project directed toward the development of III-Nitride Growth on Si wafers was previously funded by ISSO. As part of that grant, we demonstrated device quality III-V nitride-based semiconductor structures growth on commercial silicon wafers.^6,7 The results from that project combined with current achievements will make possible integration of all fluorescence sensor components on a single silicon chip and employment of fluorescence as the main sensing mechanism.

Technical Plan and Equipment

Experimental Activity
The prototype sensor for detection of live bacteria is shown in Fig. 1. The UV, blue, and green LEDs used in this simulator were set up at 90º to three Si photodiodes filtered by long-pass Roscolux optical filter films from Edmunds Industrial Optics (for the blue and green LEDs) and a 6H-SiC wafer from Cree (for the UV LED) to separate the LED and the photodetector spectral bands, while still allowing a maximum photoresponse in a range next to the LED band.

Figure 1. Working sensor prototype device for measurements of live bacteria populations.

The plasmid pGFPuv (Clontech) which expresses the green fluorescent protein (GFP) (excitation at 385 nm, emission at 508 nm) was transformed into Escherichia coli (E. coli) JM109 high efficiency competent cells using a standard heat-shock method, generating strain GFPuv. Transformants were plated onto ampicillin LB agar plates and grown overnight at 37ºC. A fresh 50 ml liquid culture (with 100 mg/ml ampicillin) was initiated from an LB plate and grown at 37ºC and 150 rpm in a shaking incubator for 24-48 hours. Two additional E. coli strains carrying plasmids which encode green and red fluorescent proteins (Clontech), respectively, were also used in this study. Strain WM611 carries plasmid pKEN, which expresses GFPmut1, a brighter green fluorescent protein variant¹⁰ (excitation at 475 nm; emission at 512 nm). Strain DsRed (Clontech) carries a plasmid which encodes a red fluorescent protein, RFP (excitation at 558 nm; emission at 583 nm). These two E. coli strains were grown in 50 mls LB broth with 100mg/ml ampicillin at 37ºC and 150 rpm in a shaking incubator for 24-48 hours (maximum color development was seen after 48 hrs). After 48 hours, the 50 ml cultures were monitored for fluorescence using the simulator. Cell densities were measured immediately after the growth with a spectrophotometer at 600 nm wavelength and cell concentration was calculated using the formula: N_cells/mL = 0.1 OD_600nm × 10⁸. The solution was then diluted two, three, four times, etc., with water. The first set of concentrations (OD method) was obtained by division of the number of cells determined by the OD measurement by two, three, four, etc., depending on the dilution number. A second set of concentrations (dilution method) was obtained for the same solutions by OD measurements of each diluted solution.

Discussion
Fabrication of the optoelectronic sensor structures on optically opaque Si substrates requires employment of side emission from the LED structure. The fluorescence is usually detected at 90&degree; to the incident (excitation) light to avoid direct illumination of the photodetector by the LED. One of the important goals of the sensor prototype fabrication and testing was verification of the device concept geometry with components configured at 90°.

Fluorescence detection has a wide range of applications in biomedical research, including microbial monitoring, detection of gene expression, and quantification of nucleic acid concentration. We, therefore, conducted measurements on live bacterial populations using the sensor prototype. The concentration of the bacteria cells was measured as a function of the dilution and growth time.

In order to investigate the concept of using spectrally tuned opto-couples to characterize analyte mixtures, we performed measurements using strains WM611 and DsRed mixed at different ratios.

Results
Normalized spectral characteristics measured from the prototype components (Fig. 2) display the three narrow and three wide bands which will allow fluorescence excitation and fluorescence emission measurement.

Figure 2. Spectral characteristics of the prototype components. Spectra of each excitation LED are shown together with the photoresponse of the filtered Si detector used for fluorescence measurements.

Fluorescence signals were measured as a function of cell concentration as determined by dilution OD methods (Fig. 3). Measurements of all three types of analytes indicate a close to linear dependence and correlation with the GFP quantum yield values. The fluorescence signal was also measured from strain WM611 as a function of growth time (Fig. 4). This dependence is close to linear, as well, with respect to optical density and photocurrent dependence on the concentration.

Figure 3. Concentration of E. coli measured with the sensor prototype using different fluorochromes.

Figure 4. E. coli cells concentration measured with the sensor prototype as a function of growth time.

In order to investigate the concept of using spectrally tuned opto-couples to characterize analyte mixtures, we performed measurements using strains WM611 and DsRed mixed at different ratios (Fig. 5). The E. coli WM611 (expressing GFPmut1) cell suspension had a higher cell density than the E. coli Ds Red (expressing RFP) suspension, which explains why OD values change at different ratios. The signal changed in the same dynamic range for both solutions (similar slopes). The higher absolute values of the photocurrent measured from the fluorescence of the red protein is due to the higher sensitivity of the Si photodetector in the red band.

Figure 5. Concentration of E. coli cells labeled with two different fluorochromes measured with the sensor prototype in mixtures at different ratios.

Fabrication and testing of the sensor prototype proved the concept of 90-degree component configuration. The sensor allowed characterization of bacteria mixtures using spectrally tuned opto-couples. Miniaturized sensors based on this technology will be invaluable in a variety of applications, including spacecraft monitoring, environmental, biomedical, and petroleum related problems.

Acknowledgements
This work was performed under support from NASA (collaborative agreement to SVEC), the National Space Biomedical Research Institute (cooperative agreement NCC 9-58), the Texas Space Grant Consortium, and the Texas Higher Education Coordinating Board. We would like to thank our collaborators from the Department of Biology and Biochemistry, Prof. G. Fox and Maia Larios-Sanz, for their help in sample preparation and characterization.

References
¹Mixed Cultures in Biotechnology. Eds. G. Zeikus and E. A. Johnston, New York: McGraw-Hill, 1991. 341-72.
²R. Mitchell. U.S. Patent 5,809.185.
³H. Chuang, P. Macuch, and M. B. Tabacco. "Optical Sensors for Detection of Bacterial. 1. General Concepts and Initial Development," Anal. Chem. 73 (2001): 462-66.
⁴A.-C. Chang, J. B. Gillespie, and M. B. Tabacco. "Enhanced Detection of Live Bacteria Using a Dendrimer Thin Film in an Optical Biosensor," Anal. Chem. 73 (2001): 467-70.
⁵S. K. Moore. "Making Chips," IEEE Spectrum (2001): 54-59.
⁶D. Starikov, I. Berishev, N. Badi, N. Medelci, J.-W. Um, A. Tempez, and A. Bensaoula. "GaN-Based Diode Structures for Optoelectronic Applications in the Near-UV Range of the Spectrum," 47th Int'l American Vacuum Society Symp., Seattle, WA, 1999.
⁷C. A. Trans, Osinski, R. F. Karlicek, Jr., and I. Berishev. "Growth of InGaN/GaN Multiple Quantum-Well Blue Light-Emitting Diodes on Silicon by Metalorganic Vapor Phase Epitaxy," Appl. Phys. Lett. 75 (1999): 1494-96.

Publications
Starikov, D., C. Boney, I. Berishev, I. C. Hernandez, and A. Bensaoula. "Radio-Frequency Molecular Beam Epitaxy Growth of III Nitrides for Microsensor Applications," J. Vac. Sci.Tech. B 19 (2001): 1404-08.
Starikov, D., C. Boney, N. Medelci, J.-W. Um, A. Bensaoula, M. Larios Sanz, and G. E. Fox. "Experimental Simulation of Integrated Optoelectronic Sensors Based on III Nitrides," JVST B (Sept./Oct. 2001). (Accepted for publication.)

Theses or student reports
Um, J.-W. "Integrated Optoelectronic Sensor Based on GaN and its Simulation," M.S. degree, Dept. of Chemistry, 1999-2001.

Presentations
Bensaoula, A., D. Starikov, and C. Boney. "III-N Multifunctional Chemical Sensors on Si Substrates," Invited talk, Barcelona Technical Institute-University of Barcelona, Spain, Aug. 9 2001.
Starikov, D., C. Boney, J.-W. Um, and A. Bensaoula. "Integrated Multifunctional Sensors Based on III-Nitrides," NanoSpace 2001, Galveston, TX, March 13-16 2001.
Starikov, D., C. Boney, J.-W. Um, N. Medelci, and A. Bensaoula. "Development of Integrated Optoelectronic Sensors Based on III Nitrides," 48th American Vacuum Society Int'l Symp., San Francisco, CA, Oct. 28-Nov. 2, 2001.

Proposals and funding
"Nitride-Based Structures for Implantable Glucose Sensors." Principal Investigator: D. Starikov; NIH, Sept. 1, 2002-Aug. 31, 2004, $200,000; pending.
"Self-Aligned Multi-Color Photodetectors Based on III-Nitrides for Advanced Flame/Fire Detection." Principal Investigator: D. Starikov; Texas Higher Education Coordinating Board, Jan. 1, 2002-Dec. 31, 2003, $175,000.

PDF (769K)
Table of Contents

Institute for Space Systems Operations - 2001 Annual Report
Copyright © 2002