Wednesday, November 7, 2007 - 11:15 AM
318-7

Microscale Colorimetric Analysis of Several Chemical Species in Soil and Water Samples Using Automated Digital Image Analysis.

Douglas Soldat and Phillip Barak. University of Wisconsin-Madison, Dept. of Soil Science, 1525 Observatory Drive, Madison, WI 53706-1299

Colorimetric analyses are used ubiquitously in turfgrass research. Over the past 15 years, several colorimetric reactions for soil and environmental water sample analysis have been modified for use with microplates. Advantages of method miniaturization include a reduction in reagents required, reduced waste stream, and increased sample throughput. However, the widespread use of these microscale techniques is likely limited by the cost of plate readers. Our objective was to determine if an inexpensive transparency scanner and a public domain software package could replace a microplate reader for analysis of several common colorimetric methods for soil and water analysis. Standard curves for colorimetric analysis methods of nitrate, ammonium, urea, phosphate, potassium, bromide, and chloride were prepared and analyzed at the appropriate wavelength using a microplate reader. Immediately following the reading a digital image of the plate was created using a 48-bit color transparency scanner. Each pixel of the digital image contains a numerical representation of light intensity for three color channels [red (R), green (G), and blue (B)]. A macro written for ImageJ, a public domain software package, was used to export the RGB data from the center of each microplate well into a spreadsheet. The raw RGB values were log transformed to confirm with the Beer-Lambert law. The complimentary color channel(s) of each colorimetric method was highly linearly correlated with the absorbance reading from the microplate reader (r2 = 0.98 – 0.99, see figure). These results suggest that several common colorimetric reactions can be accurately and inexpensively analyzed using a transparency scanner in lieu of a microplate reader.