Analytical Support Services

Charles R. Hurburgh, Jr., Professor, Agricultural and Biosystems Engineering; Connie L. Hardy, Extension Specialist, Value Added Agriculture; and Glen R. Rippke, Laboratory Manager, Agricultural and Biosystems Engineering


Bioprocessing has created the need for processors and users to measure properties of feedstocks and outputs. Examples are ethanol yield of corn, amino acid levels of corn, composition of various soybean meal formulations and of and increasing number of ethanol coproducts. The Value Added Agriculture Program, the Grain Quality Initiative and the Grain Quality Laboratory have organized a new program by which external clients can obtain assistance in measurement and quality control issues. Iowa State researchers, liscencees of Iowa State germplasm, and other public research organizations can also obtain instrument calibrations and support based on long term research in the Grain Quality Laboratory.


To provide rapid analytical support for bioprocess industries and biotechnology researchers.


Traditional PLS and robust regression methods were used to calibrate Perten DA7200 near infrared reflectance units for measurement of moisture, protein, oil, and fiber in as received soybean meal, with all processing methods combined in the same calibration. A world wide pool of about 500 samples was used for the calibration.  Traditional PLS methods gave calibrations that transferred to another unit within the error criteria of the AACC Method 39-00, but robust methods yielded standard deviations between units of less than 0.1 percentage point for all factors. A combination of constituent’s matrix was used to choose 50 new samples for validation, from a set of about 300 samples provided by a feed mill.  Standard errors for validation were 0.25, 0.58, 0.28, and 0.32 for moisture, protein, oil and fiber, respectively.  The uniqueness of this project was the demonstration of the advantage of robust methods, on an instrument type that is more difficult to transfer calibrations, and the combination of all meal types into one calibration.

An NIRS-based ethanol yield calibration using laboratory fermentation reference data from Illinois Crop Improvement Association was compared to a multiple regression against combinations of NIRS-predicted values of protein, oil, starch, and density.  Near infrared (NIR) spectra from a FOSS Infratec 1229 Grain Analyzer (FOSS Group, was obtained for 249 corn samples.   The calibration and the calculation approaches had nearly equal statistics (r2 =0.8; std dev = 0.03 gallons/bushel), but when validated against 55 new corn samples, the calculation maintained accuracy while the calibration did not.  The calculation approach requires fewer samples for updates, is easier to use in practice, is not limited to one NIRS unit, and is more accurate.  Application of the calculation to four years of variety trial data showed a 0.6 gal/bu (std dev=0.1 gal/bu) range across typical corn received by ethanol plants.

Whole soybean fatty acid contents were measured by near infrared spectroscopy. Three calibration algorithms—partial least squares (PLS), artificial neural networks (ANN), and least squares support vector machines (LS-SVM)—were implemented. Three different validation strategies using independent sets and part of calibration samples as validation sets were created. There was a significant improvement of the prediction precision of all fatty acids measured on relative concentration of oil compared with previous literature using PLS (standard error of prediction of 0.85, 0.42, 1.64, 1.67, and 0.90% for palmitic, stearic, oleic, linoleic and linolenic acids respectively). ANN and LS-SVM methods performed significantly better than PLS for palmitic, oleic and linolenic acids. Calibration models developed on relative concentrations (% of oil) were compared to prediction models created on absolute fatty acid concentration (% of weight) and corrected to relative concentration by multiplying by the predicted oil content. While models were easier to develop in absolute concentration (higher coefficients of determination), the multiplication of errors with the total oil content model resulted in no net precision improvement.


An ingredient quality control program was set up for a major Iowa feed company.  In the first 3 months of operation, the start-up and instrumentation costs have been recovered.  Ingredient quality has improved.

There is now an ethanol yield ranking procedure in the public domain.  It is being used by plant genetics researchers, and is being tested for supply management by at least one ethanol plant.

Modified fatty acid soybeans of all types can now be identified with a rapid NIRS test.  The limits of accuracy have been determined, which show that NIRS is effective at initial screening but not effective at detailed component analysis, in this case.  Additional mathematical capability in instrument software would improve accuracy, because the most accurate equations cannot be used in present instrument configurations.


180 Other ANR Programs

Page last updated: February 19, 2009
Page maintained by Linda Schultz,