Source: PRAEVIUM RESEARCH, INC. submitted to
DENSE MULTI-WAVELENGTH LASER CLUSTERS FOR NEAR INFRARED AGRICULTURAL SPECTROSCOPY
Sponsoring Institution
National Institute of Food and Agriculture
Project Status
COMPLETE
Funding Source
Reporting Frequency
Annual
Accession No.
0204091
Grant No.
2005-33610-16158
Cumulative Award Amt.
(N/A)
Proposal No.
2005-03218
Multistate No.
(N/A)
Project Start Date
Sep 1, 2005
Project End Date
Aug 31, 2008
Grant Year
2005
Program Code
[8.5]- (N/A)
Recipient Organization
PRAEVIUM RESEARCH, INC.
5276 HOLLISTER AVE
SANTA BARBARA,CA 93111
Performing Department
(N/A)
Non Technical Summary
Although Near Infrared Spectroscopy (NIRs) is a powerful non-invasive technique for identifying constituent components in a variety of agricultural products, current NIRs instrumentation is bulky and expensive, and requires long measurement times. This limits the use of NIRs in remote settings and high-speed sorting applications. This project seeks to apply miniaturized semiconductor laser technology to create hand-held NIRs instruments which are portable and inexpensive, and have measurement times sufficiently short for high-speed sorting applications.
Animal Health Component
100%
Research Effort Categories
Basic
(N/A)
Applied
100%
Developmental
(N/A)
Classification

Knowledge Area (KA)Subject of Investigation (SOI)Field of Science (FOS)Percent
71150102020100%
Goals / Objectives
Near Infrared Spectroscopy (NIRs) is widely employed in agriculture as a simple, non-invasive technique for identifying constituent components in a variety of agricultural products. Current NIRs instrumentation, however, relies on filtered white light source technology, which leads to large, power consumptive systems with critical intra-system mechanical alignments, low signal to noise ratio, and slow measurement times. This restricts the use of NIRs in remote areas or high-speed sorting applications. This phase II research effort seeks to replace filtered white light sources with multi-wavelength semiconductor laser cluster technology, enabling a portable high-speed NIRs system which covers a substantial portion of the near infrared range, with no dispersive elements or moving parts. Phase II obectives are to develop a laser cluster-based NIRs instrument employing approximately 8 semiconductor laser wavelengths in a compact, hand-held configuration. This laser-based NIRs instrument will be validated in one or more of the following areas: characterization of toxins in corn, characterization of wheat kernels, and characterization of insects. This research effort will combine the semiconductor laser expertise of Praevium Research, Inc. with the NIRs expertise of collaborators at the Grain Marketing and Production Research Center of the USDA/ARS.
Project Methods
This research will seek to miniature semiconductor laser technology by exploring a combination of advanced semiconductor materials growth, processing and integration techniques, along with new approaches to packaging semiconductor lasers to create the smallest and most practical multi-wavelength semiconductor source possible. This work will proceed in collaboration with researchers at the Grain Marketing and Production Research Center of the USDA/ARS, who will help to specify optimal wavelengths and optimal sample illumination geometry, based on their previous experience with white light measurements. Once the instrument has been assembled, it will be used to characterize wheat, insects, and/or corn, and to compare results with those obtained from traditional white light measurements. This will, for the first time, determine whether laser technology, with its improved signal to noise ratio and measurement speed, can improve confidence limits or detection thresholds in the above mentioned application areas, and whether it opens the door to high-speed sorting applications which were previously inaccessible with traditional NIRs instrumentation.

Progress 09/01/05 to 08/31/08

Outputs
OUTPUTS: The research has been devoted to developing a portable laser-cluster based near infrared spectroscopy instrument, and validating that instrument in the assessment of two areas of agricultural relevance: characterization of protein, moisture, waxy character, and hardness in wheat kernels, and characterization of agriculturally relevant insects. Commercial NIR instruments today use filtered white light sources, which suffer from low signal to noise ratio (SNR), power instability, low scan speeds, hardware expense and complexity, and lack of source modulation capability. The laser based approach uses an array of semiconductor laser light sources, which have excellent stability, high power, and single frequency operation, eliminating the need for detector arrays or gratings to provide wavelength filtering. The multi-wavelength array approach provides a discrete comb of wavelengths instead of a continuous spectrum, which is sufficient for many applications. The high SNR and compactness of the laser based approach enable portable sorting instruments and new measurement modes. This work was carried out at Praevium Research, Inc, which developed the laser cluster and spectrometer, and also at the Grain Marketing and Production Research Center (GMPRC) in Manhattan, KS, which performed spectral measurements of wheat kernels, tsetse flies, and mosquitos, and subsequent statistical analysis. This effort resulted in the successful development of multi-wavelength laser cluster consisting of 8 wavelengths in the 700 to 1700nm range, housed in a commercially available butterfly package of approximately 20X10X10mm size, with a multi-mode optical fiber exiting the package and directing radiation to samples of interest. This light source in turn enabled the spectral measurements of several samples in transmission mode. These measurements and subsequent statistical analysis demonstrated that the Praevium laser source generally provided measurement accuracy as good or better than the leading competing white light instrument, at a fraction of the size and cost, making it an attractive commercial product opportunity. These measurements were performed on stationary samples. The results obtained set the stage for full development of a future high-speed sorting system complete with drive electronics and mechanical sample handling. These results are being disseminated through use of this new instrument at the USDA ARS GMPRC lab in Manhattan, KS, which is continually evaluating this instrument in measuring new quantities, and disseminating that information within the USDA. PARTICIPANTS: The primary individuals on this project were Drs. V. Jayaraman, E. Hall, and D. Leonard from Praevium Research, and Dr. Floyd E. Dowell from GMPRC of the USDA ARS in Manhattan, KS. A significant portion of this effort was the CRADA collaboration between Praevium and the GMPRC. TARGET AUDIENCES: The targe audience for this work consists of any agricultural researchers employing near infrared spectroscopy. PROJECT MODIFICATIONS: Not relevant to this project.

Impacts
The development of a compact portable high-performance spectrometer based on inexpensive semiconductor laser technology will increase the use of spectroscopy as a rapid non-invasive analysis tool for a variety of agricultural products. This will ultimately increase the quality, throughput, safety, and classification accuracy of agricultural products. Although Near Infrared spectroscopy has been in use in agriculture for some time, the bulk and expense of the instrumentation has prohibited widespread use. A number of potential commercial applications emerge for this instrument in the areas of high speed sorting and field portable use. High speed sorting of single seeds is a natural application, as this is the area in which this instrument has been validated. Other sorting applications include differential grape harvesting for wine production, which is critical to increase the quality and global competitiveness of the U.S. wine industry, rapid characterization of wood strips to aid in tree breeding, and in insect control through rapid spectroscopic selection. Spectroscopic tools for these and other applications must be field portable and inexpensive.

Publications

  • No publications reported this period


Progress 09/01/05 to 09/01/06

Outputs
The ultimate goal of this effort is to create a miniaturized spectrometer based on a dense multi-wavelength cluster of semiconductor lasers, and to evaluate the performance of that spectrometer in the characterization of one or more of the following: wheat kernels, corn kernels, and insects. This laser-based spectrometer has signal to noise, speed, portability, and compactness advantages over traditional instruments based on filtered white light sources. The first year of the effort was devoted to developing the laser-based instrument, and the second year will be devoted to evaluating its effectiveness in sorting wheat kernels based on protein and moisture content. This effort involves a CRADA collaboration with the grain marketing and production research center (GMPRC) in Manhattan, Kansas, of the USDA ARS. The first year of this effort has focused on developing the multi-wavelength laser cluster, and we have succeeded in demonstrating a single-chip cluster of 12 wavelengths in the range of 700-1700nm. Typical laser output powers are 5-40 milliwatts, and typical spectral widths are approximately 1-2 nm. This cluster of 12 wavelengths exceeds the original phase II target of 8 wavelengths. Wavelengths were chosen based on beta coefficient analysis of existing white light spectra of wheat kernels. This 12-channel chip was enabled partly by polymer waveguide technology developed at Praevium which allows multiple discrete semiconductor laser beams to be routed to an area that is substantially smaller than the wheat kernel. The lasers and waveguide occupy an area of about 6X6 millimeters, enabling a very compact near infrared spectroscopic source. In addition to developing the source, collaboration with GMPRC researchers has established transmission mode as the sample illumination/detection geometry for the first wheat sorting experiments. That is, laser light transmitted through the wheat kernels will be measured, and an "underfill" geometry where the beam illuminates an area smaller than the kernel will be employed. The high power of the laser-based approach enables transmission measurements, which are much more difficult with filtered white light sources. In the next 2-3 months, packaging of the multi-wavelength laser array will be completed, and laser based spectrometer will be handed to GMPRC collaborators for wheat kernel sorting experiments. We expect initial sorting accuracy numbers by Spring of 2007.

Impacts
The development of a compact portable high-performance spectrometer based on inexpensive semiconductor laser technology will increase the use of spectroscopy as a rapid non-invasive analysis tool for a variety of agricultural products. This will ultimately increase the quality, throughput, safety, and classification accuracy of agricultural products. Although Near Infrared spectroscopy has been in use in agriculture for some time, the bulk and expense of the instrumentation has prohibited widespread use. The current focus is on characterizing wheat kernels, enabling rapid sorting of individual kernels on the basis of protein or moisture content. This would be of economic benefit to grain producers and seed growers. A related application is detection of toxins in corn kernels.

Publications

  • No publications reported this period